1. health and fitness
2. disease

The American Journal of Cardiology
Copyright © 2005 Elsevier Inc. All rights reserved
Volume 95, Issue 2, Pages 161-310 (15 January 2005)
1. g
c
d
e
f
Editorial board • EDITORIAL BOARD
Page A3
Coronary Artery Disease
2. g
c
d
e
f
3. g
c
d
e
f
4. g
c
d
e
f
5. g
c
d
e
f
Impact of sirolimus-eluting stent on the outcome of patients with
chronic total occlusions • ARTICLE
Pages 161-166
Sunao Nakamura, Tamil Selvan Muthusamy, Jang-Ho Bae, Yeo Hans
Cahyadi, Wasan Udayachalerm and Damras Tresukosol
Impact of baseline renal function on mortality after percutaneous
coronary intervention with sirolimus-eluting stents or bare metal
stents • ARTICLE
Pages 167-172
Pedro A. Lemos, Chourmouzios A. Arampatzis, Angela Hoye, Joost
Daemen, Andrew T.L. Ong, Francesco Saia, Willem J. van der Giessen,
Eugene P. McFadden, Georgios Sianos, Pieter C. Smits et al.
Determinants of 30-day adverse events following saphenous vein graft
intervention with and without a distal occlusion embolic protection
device • ARTICLE
Pages 173-177
Gregory R. Giugliano, Richard E. Kuntz, Jeffrey J. Popma, Donald E.
Cutlip, Donald S. Baim and Saphenous Vein Graft Angioplasty Free of
Emboli Randomized (SAFER) Trial Investigators
Artificial neural network modeling of stress single-photon emission
computed tomographic imaging for detecting extensive coronary
artery disease • ARTICLE
Pages 178-181
J. Scott Allison, Jaekyeong Heo and Ami E. Iskandrian
6. g
c
d
e
f
Incremental prognostic power of single-photon emission computed
tomographic myocardial perfusion imaging in patients with known or
suspected coronary artery disease • ARTICLE
Pages 182-188
Salvador Borges-Neto, Linda K. Shaw, Robert H. Tuttle, John H.
Alexander, William T. Smith IV, Marianna Chambless, R. Edward
Coleman, Robert A. Harrington and Robert M. Califf
Preventive Cardiology
7. g
c
d
e
f
8. g
c
d
e
f
Comparative effects of rosuvastatin and gemfibrozil on glucose,
insulin, and lipid metabolism in insulin-resistant, nondiabetic patients
with combined dyslipidemia • ARTICLE
Pages 189-193
Cindy Lamendola, Fahim Abbasi, James W. Chu, Howard Hutchinson,
Valerie Cain, Elizabeth Leary, Tracey McLaughlin, Evan Stein and Gerald
Reaven
Relation between atherogenic dyslipidemia and the Adult Treatment
Program-III definition of metabolic syndrome (Genetic Epidemiology
of Metabolic Syndrome Project) • ARTICLE
Pages 194-198
Diego F. Wyszynski, Dawn M. Waterworth, Philip J. Barter, Jonathan
Cohen, Y. Antero Kesäniemi, Robert W. Mahley, Ruth McPherson,
Gérard Waeber, Thomas P. Bersot, Sanjay S. Sharma et al.
Miscellaneous
9. g
c
d
e
f
New predictors of outcome in idiopathic pulmonary arterial
hypertension • ARTICLE
Pages 199-203
Steven M. Kawut, Evelyn M. Horn, Ketevan K. Berekashvili, Robert P.
Garofano, Rochelle L. Goldsmith, Allison C. Widlitz, Erika B.
Rosenzweig, Diane Kerstein and Robyn J. Barst
Interview
10. g
c
d
e
f
Anthony Nicholas Demaria, MD: A conversation with the editor •
DISCUSSION
Pages 204-223
Brief Reports
11. g
c
d
e
f
12. g
c
d
e
f
13. g
c
d
e
f
14. g
c
d
e
f
15. g
c
d
e
f
16. g
c
d
e
f
Relation of impaired Thrombolysis In Myocardial Infarction
myocardial perfusion grades to residual thrombus following the
restoration of epicardial patency in ST-elevation myocardial
infarction • SHORT COMMUNICATION
Pages 224-227
Ajay J. Kirtane, Aaron Weisbord, Dimitrios Karmpaliotis, Sabina A.
Murphy, Robert P. Giugliano, Christopher P. Cannon, Elliott M. Antman,
E. Magnus Ohman, Matthew T. Roe, Eugene Braunwald et al.
Angiography and revascularization in patients with heart failure
following fibrinolytic therapy for ST-elevation acute myocardial
infarction • SHORT COMMUNICATION
Pages 228-233
Amir Kashani, C. Michael Gibson, Sabina A. Murphy, Marc S. Sabatine,
David A. Morrow, Elliott M. Antman and Robert P. Giugliano
Impact of ST-segment depression resolution on mortality after
successful mechanical reperfusion in patients with ST-segment
elevation acute myocardial infarction • SHORT COMMUNICATION
Pages 234-236
Giuseppe De Luca, Arthur C. Maas, Arnoud W.J. van't Hof, Jan Paul
Ottervanger, Jan C.A. Hoorntje, A.T. Marcel Gosselink, Jan-Henk E.
Dambrink, Menko-Jan de Boer and Harry Suryapranata
Effects of tirofiban and statins on high-sensitivity C-reactive protein,
interleukin-6, and soluble CD40 ligand following percutaneous
coronary interventions in patients with stable coronary artery
disease • SHORT COMMUNICATION
Pages 236-240
Rabih R. Azar, Georges Badaoui, Antoine Sarkis, Roland Kassab, Elie
Salamé, Samira Klaymé, Roger Naman and Mirna Germanos
Value of preprocedure multislice computed tomographic coronary
angiography to predict the outcome of percutaneous recanalization
of chronic total occlusions • SHORT COMMUNICATION
Pages 240-243
Nico R. Mollet, Angela Hoye, Pedro A. Lemos, Filippo Cademartiri,
Georgios Sianos, Eugene P. McFadden, Gabriel P. Krestin, Patrick W.
Serruys and Pim J. de Feyter
Angiographic and clinical outcomes of polytetrafluoroethylenecovered stent use in significant coronary perforations • SHORT
COMMUNICATION
Pages 244-246
Hung Ly, Jean-Pierre S. Awaida, Jacques Lespérance and Luc Bilodeau
17. g
c
d
e
f
Role of postoperative vasopressor use in occurrence of atrial
fibrillation after coronary artery bypass grafting • SHORT
COMMUNICATION
Pages 247-249
Vikrant Salaria, Nirav J. Mehta, Syed Abdul-Aziz, Syed M. Mohiuddin
and Ijaz A. Khan
18. g
c
d
e
f
19. g
c
d
e
f
Reaching recommended lipid and blood pressure targets with
amlodipine/atorvastatin combination in patients with coronary heart
disease • SHORT COMMUNICATION
Pages 249-253
Jean-Francois Dorval, Todd Anderson, Jean Buithieu, Sammy Chan,
Stuart Hutchison, Thao Huynh, Jean Jobin, Eva Lonn, Paul Poirier,
Lawrence Title et al.
Benefits of niacin by glycemic status in patients with healed
myocardial infarction (from the Coronary Drug Project) • SHORT
COMMUNICATION
Pages 254-257
Paul L. Canner, Curt D. Furberg, Michael L. Terrin and Mark E.
McGovern
20. g
c
d
e
f
Occurrence of acute myocardial infarction in Worcester,
Massachusetts, before, during, and after the terrorists attacks in New
York City and Washington, DC, on 11 September 2001 • SHORT
COMMUNICATION
Pages 258-260
Robert J. Goldberg, Frederick Spencer, Darleen Lessard, Jorge
Yarzebski, Craig Lareau and Joel M. Gore
21. g
c
d
e
f
Characteristics and outcomes of patients with acute myocardial
infarction and angiographically normal coronary arteries • SHORT
COMMUNICATION
Pages 261-263
Alf Inge Larsen, P. Diane Galbraith, William A. Ghali, Colleen M.
Norris, Michelle M. Graham, Merril L. Knudtson and APPROACH
Investigators
22. g
c
d
e
f
Long-term safety and efficacy of high-dose atorvastatin treatment in
patients with familial hypercholesterolemia • SHORT
COMMUNICATION
Pages 264-266
Sanne van Wissen, Tineke J. Smilde, Mieke D. Trip, Anton F.H.
Stalenhoef and John J.P. Kastelein
23. g
c
d
e
f
24. g
c
d
e
f
25. g
c
d
e
f
26. g
c
d
e
f
Effect of medical and surgical weight loss on endothelial vasomotor
function in obese patients • SHORT COMMUNICATION
Pages 266-268
Noyan Gokce, Joseph A. Vita, Marie McDonnell, Armour R. Forse,
Nawfal Istfan, Maria Stoeckl, Izabella Lipinska, John F. Keaney, Jr and
Caroline M. Apovian
Prevalence of adequate control of increased serum low-density
lipoprotein cholesterol in self-pay or Medicare patients versus
Medicaid or private insurance patients followed in a University
General Medicine Clinic • SHORT COMMUNICATION
Pages 269-270
Raja Varma, Wilbert S. Aronow, Glenn Gandelman and Christopher
Zammit
Clinical variables predicting inappropriate use of implantable
cardioverter-defibrillator in patients with coronary heart disease or
nonischemic dilated cardiomyopathy • SHORT COMMUNICATION
Pages 271-274
Dominic A.M.J. Theuns, A. Peter J. Klootwijk, Maarten L. Simoons and
Luc J. Jordaens
Design of the SHock Inhibition Evaluation with Azimilide (SHIELD)
study: A novel method to assess antiarrhythmic drug effect in
patients with an implantable cardioverter-defibrillator • SHORT
COMMUNICATION
Pages 274-276
Craig M. Pratt, Paul Dorian, Hussein R. Al-Khalidi, Jose M. Brum,
Martin Borggrefe, Daljit S. Tatla, Johannes Brachmann, Robert J.
Myerburg, David S. Cannom, Michael J. Holroyde et al.
27. g
c
d
e
f
28. g
c
d
e
f
Refeeding normalizes the QT rate dependence of female anorexic
patients • SHORT COMMUNICATION
Pages 277-280
Frédéric Roche, Jean-Claude Barthélémy, Norbert Mayaud, Vincent
Pichot, David Duverney, Natacha Germain, François Lang and Bruno
Estour
Left ventricular structure and function in sedentary and physically
active subjects with left ventricular hypertrophy (the LIFE Study) •
SHORT COMMUNICATION
Pages 280-283
Kurt Boman, Mona Olofsson, Björn Dahlöf, Eva Gerdts, Markku S.
Nieminen, Vasilios Papademetriou, Kristian Wachtell and Richard B.
Devereux
29. g
c
d
e
f
30. g
c
d
e
f
Comparison of angiotensin-converting enzyme inhibitors in the
treatment of congestive heart failure • SHORT COMMUNICATION
Pages 283-286
Karen Tu, Muhammad Mamdani, Alex Kopp and Douglas Lee
Simplified peak power reserve in patients with an implantable
cardioverter-defibrillator and advanced heart failure • SHORT
COMMUNICATION
Pages 286-288
William T. Katsiyiannis, Alan D. Waggoner, Benico Barzilai, Brian F.
Gage, Jose M. Sanchez, Joseph G. Rogers, Bruce D. Lindsay and Marye
J. Gleva
31. g
c
d
e
f
32. g
c
d
e
f
33. g
c
d
e
f
34. g
c
d
e
f
35. g
c
d
e
f
Effect of hormone therapy on mortality rates among women with
heart failure and coronary artery disease • SHORT COMMUNICATION
Pages 289-291
Kirsten Bibbins-Domingo, Feng Lin, Eric Vittinghoff, Elizabeth BarrettConnor, Stephen B. Hulley, Deborah Grady and Michael G. Shlipak
Aortic dimensions in patients with bicuspid aortic valve without
significant valve dysfunction • SHORT COMMUNICATION
Pages 292-294
Moreno Cecconi, Marcello Manfrin, Alessandra Moraca, Raffaele
Zanoli, Pier Luigi Colonna, Maria Grazia Bettuzzi, Stefano Moretti,
Domenico Gabrielli and Gian Piero Perna
Is transesophageal echocardiography overused in the diagnosis of
infective endocarditis? • SHORT COMMUNICATION
Pages 295-297
Molly Thangaroopan and Jonathan B. Choy
Symptomatic patients have similar outcomes compared with
asymptomatic patients after carotid artery stenting with emboli
protection • SHORT COMMUNICATION
Pages 297-300
Michael H. Yen, David S. Lee, Samir Kapadia, Ravish Sachar, Deepak L.
Bhatt, Christopher T. Bajzer and Jay S. Yadav
Prognostic importance of isolated T-wave abnormalities • SHORT
COMMUNICATION
Pages 300-304
Takuya Yamazaki, Jonathan Myers and Victor F. Froelicher
36. g
c
d
e
f
37. g
c
d
e
f
Relation between effects of adenosine on brachial artery reactivity
and perfusion pattern in patients with known or suspected coronary
artery disease • SHORT COMMUNICATION
Pages 304-307
Deval Mehta, Gurpreet Baweja, Rajesh Venkataraman, Gilbert J. Zoghbi,
Thein Htay, Jaekyeong Heo, Navin C. Nanda and Ami E. Iskandrian
Usefulness of real-time three-dimensional echocardiography for
reliable measurement of cardiac output in patients with ischemic or
idiopathic dilated cardiomyopathy • SHORT COMMUNICATION
Pages 308-310
Sean M. Fleming, Barry Cumberledge, Christoph Kiesewetter, Gareth
Parry and Antoinette Kenny
EDITOR IN CHIEF
William C. Roberts,
ASSOCIATE EDITORS
Baylor Heart & Vascular Hospital
Baylor University Medical Center
Wadley Tower No. 457
3600 Gaston Avenue
Dallas, Texas 75246
(214)826-8252
Fax: (214)826-2855
CARDIOVASCULAR
MEDICINE
In Adults
Jonathan Abrams
Robert J. Adolph
Joseph S. Alpert
Martin A. Alpert
Ezra A. Amsterdam
Jeffrey Anderson
William F. Armstrong
Richard W. Asinger
Donald S. Baim
Gary John Balady
Thomas M. Bashore
Eric Bates
George A. Beller
David G. Benditt
Peter C. Block
William E. Boden
Monty M. Bodenheimer
Robert O. Bonow
Jeffrey S. Borer
Harisios Boudoulas
Martial G. Bourassa
Eugene Braunwald
Jeffrey A. Brinker
Bruce R. Brodie
Alfred E. Buxton
Michael E. Cain
Richard O. Cannon III
Samuel Ward Casscells
Agustin Castellanos
Bernard R. Chaitman
Kanu Chatterjee
John S. Child
Robert J. Cody
Lawrence S. Cohen
Marc Cohen
C. Richard Conti
Michael H. Crawford
Robert F. DeBusk
Prakash C. Deedwania
Gregory J. Dehmer
Efthymios N. Deliargyris
James A. de Lemos
Anthony N. DeMaria
Pablo Denes
Nicholas L. DePace
Richard B. Devereux
George A. Diamond
John P. DiMarco
Michael J. Domanski
Gerald Dorros
John S. Douglas, Jr.
Pamela S. Douglas
Eric J. Eichhorn
Mark J. Eisenberg
Uri Elkayam
Thomas C. Andrews
Paul A. Grayburn
MD
ASSISTANT EDITORS
Calixto A. Romero
Robert L. Rosenthal
Carlos E. Velasco
EDITORIAL BOARD
Kenneth A. Ellenbogen
Myrvin H. Ellestad
Stephen G. Ellis
Toby R. Engel
Andrew E. Epstein
N. A. Mark Estes, III
Michael Ezekowitz
Rodney H. Falk
John A. Farmer
David P. Faxon
Robert L. Feldman
Ted Feldman
Jack Ferlinz
Jerome L. Fleg
Gerald F. Fletcher
Nancy C. Flowers
James S. Forrester
Joseph A. Franciosa
Gary S. Francis
Victor F. Froelicher, Jr.
W. Bruce Fye
William H. Gaasch
William Ganz
Julius M. Gardin
Bernard J. Gersh
Mihai Gheorghiade
Raymond Gibbons
D. Luke Glancy
Stephen P. Glasser
Michael R. Gold
Samuel Z. Goldhaber
Robert E. Goldstein
Sidney Goldstein
Steven A. Goldstein
J. Anthony Gomes
Antonio M. Gotto, Jr.
K. Lance Gould
Donald C. Harrison
Richard H. Helfant
Gary V. Heller
Philip D. Henry
L. David Hillis
David R. Holmes, Jr.
William G. Hundley
Abdulmassih S. Iskandrian
Allan S. Jaffe
William B. Kannel
Norman M. Kaplan
Joel S. Karliner
John A. Kastor
Sanjiv Kaul
Harold L. Kennedy
Kenneth M. Kent
Richard E. Kerber
Dean J. Kereiakes
Morton J. Kern
Spencer B. King III
Robert E. Kleiger
George J. Klein
Lloyd W. Klein
Paul Kligfield
Robert A. Kloner
John B. Kostis
Morris N. Kotler
Charles Landau
Richard L. Lange
Carl J. Lavie
Carl V. Leier
Joseph Lindsay, Jr.
Gregory Y.H. Lip
Joseph Loscalzo
G.B. John Mancini
Francis E. Marchlinski
Frank I. Marcus
Barry J. Maron
Randolph P. Martin
Attilo Maseri
Dean T. Mason
Michael D. McGoon
Raymond G. McKay
Jawahar L. Mehta
Richard S. Meltzer
Franz H. Messerli
Eric L. Michelson
Alan B. Miller
Gary S. Mintz
Fred Morady
Arthur J. Moss
James E. Muller
Robert J. Myerburg
Gerald B. Naccarelli
Navin C. Nanda
Christopher O’Connor
Robert A. O’Rourke
Erik Magnus Ohman
Antonio Pacifico
Richard L. Page
Eugene R. Passamani
Richard C. Pasternak
Alan S. Pearlman
Carl J. Pepine
Joseph K. Perloff
Ileana Pina
Bertram Pitt
Philip J. Podrid
Arshed A. Quyyumi
Charles E. Rackley
C. Venkata Ram
Nathaniel Reichek
Robert Roberts
William J. Rogers
Melvin M. Scheinman
Nelson B. Schiller
David J. Schneider
John S. Schroeder
Pravin M. Shah
Prediman K. Shah
Jamshid Shirani
Robert J. Siegel
Marc A. Silver
Mark E. Silverman
Ross J. Simpson, Jr.
Steven N. Singh
Sidney C. Smith, Jr.
Burton E. Sobel
John C. Somberg
Peter C. Spittell
David H. Spodick
Lynne W. Stevenson
John R. Stratton
Jonathan M. Tobis
Eric J. Topol
Byron F. Vandenberg
Hector O. Ventura
George W. Vetrovec
Ronald G. Victor
Robert A. Vogel
Frans J. Wackers
David D. Waters
Nanette K. Wenger
William B. White
Robert Wilensky
James T. Willerson
John R. Wilson
Miguel Zabalgoitia
Barry L. Zaret
Douglas P. Zipes
In Infants and Children
Hugh D. Allen
Bruce S. Alpert
Arthur Garson, Jr.
Stanley J. Goldberg
Thomas P. Graham, Jr.
Warren G. Guntheroth
Howard P. Gutgesell
John D. Kugler
James E. Lock
John W. Moore
Lowell W. Perry
David J. Sahn
Richard M. Schieken
A. Rebecca Snider
CARDIOVASCULAR SURGERY
Eugene H. Blackstone
Lawrence I. Bonchek
Lawrence H. Cohn
John A. Elefteriades
Thomas L. Spray
RELATED SPECIALISTS
L. Maximilian Buja
Barry A. Franklin
Charles B. Higgins
Jeffrey E. Saffitz
Renu Virmani
Redford B. Williams
A3
Impact of Sirolimus-Eluting Stent on the
Outcome of Patients With Chronic
Total Occlusions
Sunao Nakamura, MD, PhD, Tamil Selvan Muthusamy, MD, Jang-Ho Bae, MD,
Yeo Hans Cahyadi, MD, Wasan Udayachalerm, MD, and Damras Tresukosol, MD
Several randomized trials have demonstrated that stent
implantation after successful recanalization of long-term
total occlusions decreases restenosis and reocclusion
rates. The sirolimus-eluting stent (SES) has recently
proved its efficacy to decrease restenosis in selected
patients. However, the efficacy of SES implantation in
patients who have chronic total occlusions is currently
unknown. Therefore, we investigated procedural and 6and 12-month angiographic outcomes (analyzed by
quantitative coronary angiography) and left ventricular
function in 60 patients who received SESs and 120
patients who received bare metal stents (BMSs). Minimum luminal diameter did not differ immediately after
recanalization (SES group 3.04 ⴞ 0.50 mm vs BMS
group 3.12 ⴞ 0.48 mm). After 6 months, the SES group
still had significantly larger luminal diameters (3.04 ⴞ
0.44 mm vs 1.94 ⴞ 0.98 mm) and significantly lower
restenosis and reocclusion rates (2% and 0%, respectively) than did the BMS group (32% and 6%, respectively). Late loss was significantly smaller in the SES
group than in the BMS group. At follow-up, the SES
group had fewer cardiac events, including target lesion
revascularization (p <0.001), than did the BMS group.
In conclusion, SES implantation after recanalization of
chronic total occlusion provides a better clinical outcome
with less restenosis and target lesion revascularization
after 6 months than does BMSs. 䊚2005 by Excerpta
Medica Inc.
(Am J Cardiol 2005;95:161–166)
he long-term success after percutaneous coronary
intervention for chronic total occlusions (CTOs) is
T
hampered by high rates of subacute reocclusion and
tation in patients who have CTO is currently unknown. Therefore, we performed a prospective, multicenter, observational study in patients who had CTO
and underwent elective SES implantation or conventional bare metal stent (BMS) implantation to evaluate
success rates, in-hospital and clinical outcomes, and
their correlation with predefined clinical and angiographic variables.
late restenosis. After successful recanalization of total
occlusion (provided the vessel remains patent), patients report relief from symptoms and develop improvement of left ventricular function, fewer cardiac
events, and fewer requirements for bypass surgery.1,2
In addition, several randomized trials have demonstrated that stent implantation decreases restenosis and
reocclusion rates3–9 and confers a long-term survival
advantage.10 However, restenosis remains the major
complication limiting late outcome after percutaneous
coronary intervention. The efficacy of the sirolimuseluting stent (SES; Cypher, Cordis Corp., Johnson &
Johnson, Miami, Florida) has recently been proved to
decrease restenosis in selected populations.11–13 Importantly, by maintaining all mechanical properties,
the late benefit observed with the SES has been accomplished without compromising the excellent procedural and acute results obtained with conventional
metallic stents. However, the efficacy of SES implanFrom the Department of Cardiology, New Tokyo Hospital, Matsudo,
Japan; the Department of Cardiology, Damansara Specialist Hospital,
Kuala Lumpur, Malaysia; the Department of Cardiology, Konyang
University Hospital, Daejeon, Korea; the Department of Cardiology,
Husada Hospital, Jakarta, Indonesia; the Department of Cardiology,
King Chulalongkorn Memorial Hospital, Bangkok, Thailand; and the
Department of Cardiology, Siriaj Hospital, Bangkok, Thailand. Manuscript received May 7, 2004; revised manuscript received and accepted August 31, 2004.
Address for reprints: Sunao Nakamura, MD, PhD, Department of
Cardiology, New Tokyo Hospital, 473-1 Nemoto, Matsudo-shi,
Chiba 271-0077, Japan. E-mail: boss@ryukyu.ne.jp.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
METHODS
Study design: From March 2002 to March 2003, a
multicenter, prospective, observational clinical study
was conducted in 6 Asian heart centers. All 180 consecutive patients who had CTO successfully recanalized with percutaneous coronary intervention and
who underwent coronary stenting were enrolled in this
study. Among them, 60 patients (60 CTOs in total)
underwent SES implantation and were enrolled in the
SES group, and 120 patients (120 CTOs in total)
underwent BMS implantation (Medtronic S670,
Medtronic Vascular, Santa Rosa, California; MultiLink Tristar, Guidant, Temecula, California; BX Velocity, Cordis Corp.) and were enrolled in the BMS
group. Patients who had a previous revascularization
procedure (percutaneous coronary intervention or coronary artery bypass graft) or myocardial infarction ⱕ3
months of the procedure and patients who had an
estimated CTO duration ⬍90 days were excluded.
Lesion-related criteria for exclusion were a reference
diameter ⬍2.5 mm by visual estimation, an angiographically visible thrombus adjacent to the occlusion
site, significant left main coronary artery disease, and
totally occlusive diffuse in-stent restenosis. CTO was
defined as complete interruption of the vessel with
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.088
161
TABLE 1 Baseline Characteristics of Patients Who Received
Bare Metal Stent or Sirolimus-eluting Stent
Age (yrs)
Men
Hypertension
Hyperlipidemia
Diabetes mellitus
Current smoker
Previous myocardial infarction
Previous coronary bypass
Ejection fraction (%)
Estimated duration of
occlusion (mo)
No. of narrowed coronary
arteries
1
2
3
Occluded coronary arteries
Left anterior descending
Left circumflex
Right
BMS
(n ⫽ 120)
SES
(n ⫽ 60)
68.6 ⫾ 7.8
75 (63%)
72 (60%)
40 (33%)
39 (33%)
59 (49%)
42 (35%)
10 (8%)
55.8 ⫾ 6.8
7.8 ⫾ 6.6
69.5 ⫾ 8.8
42 (70%)
35 (58%)
21 (35%)
20 (33%)
30 (50%)
22 (37%)
7 (11%)
51.8 ⫾ 7.7
8.8 ⫾ 7.8
20 (17%)
62 (52%)
38 (32%)
13 (22%)
33 (55%)
14 (23%)
54%
27%
33%
52%
28%
28%
Values are mean ⫾ SD or numbers of patients (percentages).
TABLE 2 In-hospital Clinical Outcome
Angiographic success
Procedural success
Major adverse cardiac event
Death
Q-wave myocardial infarction
Urgent coronary bypass
Urgent percutaneous coronary
intervention
Minor complications
Vessel perforation
Cardiac tamponade
Vascular hematoma
BMS
(n ⫽ 120)
SES
(n ⫽ 60)
120 (100%)
119 (99%)
60 (100%)
60 (100%)
0
1
0
1
0
0
0
0
0
0
2 (2%)
0
0
1 (2%)
Thrombolysis In Myocardial Infarction flow grade 0
that was ⱖ3 months old. Occlusion age was estimated
from the date of myocardial infarction on the distribution of the occluded vessel, abrupt worsening of
angina pectoris, or information provided by sequential
angiograms when they were available. All patients
gave written, informed consent to the procedure. In
addition, at the time of initial stent implantation, patients were informed about the need for 6- and 12month angiographic and late clinical follow-up.
Stent implantation procedure: Premedication consisted of 200 mg/day of aspirin orally and 150 mg of
clopidogrel starting ⱖ24 hours before percutaneous coronary intervention. A bolus of 10,000 U of heparin was
administered after sheath insertion, with repeat bolus
given as needed to maintain an activated clotting time of
⬎250 seconds. All patients underwent percutaneous coronary intervention by a standard over-the-wire technique
using an exchange catheter (Excelsior, Scimed, Boston
Scientific Corporation, Natick, Massachusetts) in which
the femoral approach was used. The technique has been
162 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
TABLE 3 Angiographic Characteristics (Baseline)
Reference diameter (proximal)
(mm)
Minimum luminal diameter
after stenting (mm)
Angiographic calcification
Side branch: present
Lesion length (mm)
Chronic total occlusion length
(mm)
Stent length (mm)
Maximum inflation pressure
(atm)
Stent overlap
BMS
(n ⫽ 120)
SES
(n ⫽ 60)
3.32 ⫾ 0.50
3.12 ⫾ 0.60
3.12 ⫾ 0.48
3.04 ⫾ 0.50
57%
32%
32.5 ⫾ 7.8
23.6 ⫾ 6.6
68%
30%
35.0 ⫾ 8.8
24.8 ⫾ 5.5
30.8 ⫾ 8.8
15.9 ⫾ 4.4
36.5 ⫾ 2.2
16.8 ⫾ 3.0
43%
47%
Values are mean ⫾ SD or percentages.
modified over time using second-generation guidewires,
including Cross-It (Guidant, Santa Clara, California) and
Conquest Pro (Asahi Intecc Co., Ltd., Nagoya, Japan), or
use of sophisticated techniques for crossing the occluded
artery.14 After the lesion was crossed by the guidewire
and successfully recanalized with standard angioplasty
that resulted in a reference diameter ⱖ2.5 mm on on-line
quantitative coronary angiography, BMSs and SESs
were implanted in patients on the BMS and SES registries, respectively. Stent deployment was performed using an initial inflation of 16 atm in the 2 groups. Stent
expansion was completed by additional inflation (maximum 20 atm) using other short balloons until good
angiographic results were obtained.
Definitions: CTO was defined as a lesion that exhibited Thrombolysis In Myocardial Infarction flow
grade 0 in a native coronary artery. Angiographic
technical success was defined as the ability to cross
the occluded segment with wire and balloon and successfully open the artery with ⬍40% residual stenosis
from all views. Procedural success was defined as
angiographic success with no in-hospital major adverse cardiac events. Major adverse cardiac events
were defined as the occurrence of death, Q-wave myocardial infarction, or urgent revascularization. Urgent
revascularization was classified by operators who
cared for patients and required repeat percutaneous
coronary intervention of the target vessel during the
same admission or coronary artery bypass grafting,
including bypass of the target vessel.
Follow-up: Coronary events during and after the
procedure were recorded. Clinical follow-up was performed in the cardiology clinic 1, 3, 6, 9, and 12
months after the procedure and included physical examination, stress testing, and documenting a history of
cardiac complaints for all patients. Six and 12 months
after recanalization, patients were hospitalized for repeat angiography as much as possible (follow-up angiographic rates of 67% at 6 months and 75% at 12
months in the SES group and 97% at 6 months in the
BMS group).
Angiographic analysis: Angiographic analysis was
performed at a central core laboratory by 2 experiJANUARY 15, 2005
period, 120 of 130 screened patients
who had CTO (92%) were successSES
fully recanalized, implanted with
BMS at
BMS (BX Velocity 59%, Multi-Link
6 months
6 months
12 months
Tristar 22%, Medtronic AVE 19%),
(n ⫽ 120)
(n ⫽ 60)
(n ⫽ 60)
p Value
and enrolled in the multicenter BMS
Clinical follow-up
120/120
60/60
60/60
NS
registry.
Angiographic follow-up
116/120
40/60
45/60
0.05
Baseline clinical characteristics: BaseReference diameter (mm)
3.24 ⫾ 0.48 3.14 ⫾ 0.55 3.16 ⫾ 0.50
NS
Minimum luminal diameter (mm) 1.94 ⫾ 0.98 3.04 ⫾ 0.44 3.00 ⫾ 0.50
0.001
line characteristics of the 2 groups
Late loss (mm)
1.36 ⫾ 0.88 0.08 ⫾ 0.10 0.12 ⫾ 0.10
0.001
are listed in Table 1. There was no
Loss index
0.41 ⫾ 0.48 0.03 ⫾ 0.02 0.04 ⫾ 0.03
0.001
significant difference in baseline
Restenosis rate
32%
2%
2%
0.001
clinical and angiographic characterReocclusion rate
6%
0%
0%
0.001
Target vessel revascularization
30%
3%
3%
0.001
istics between groups. Most patients
Target lesion revascularization
23%
2%
2%
0.001
presented with multivessel disease
(BMS 83%, SES 78%). About 33%
Values are numbers of patients, mean ⫾ SD, or percentages.
of patients had a previous myocardial infarction (in the region of the
myocardium supplied by the target
enced operators who were not involved in the stenting vessel; BMS 35%, SES 37%). Viability of the myoprocedure. Angiograms were quantitatively analyzed cardium supplied by the target vessel was not invesby a computer-assisted cardiovascular measurement tigated by protocol, but in 70% of patients, the pressystem (MEDIS, Nuenen, The Netherlands) that used ence of normal or only decreased regional systolic
semiautomated edge detection. Minimum luminal di- wall motion of that area evaluated by angiography or
ameter, reference diameter, and percent diameter ste- echocardiography suggested the presence of a viable
nosis were measured before and after the procedure myocardium.
and at 6- and 12-month follow-up; results from the
In-hospital outcomes: Table 2 lists in-hospital
single “worst” view were recorded.
events. Angiographic success was obtained in 100%
Study end points: The primary end point in this of the 2 groups. Procedural success was obtained in
study was the occurrence of major adverse cardiac 100% of the SES group but was 99% of the BMS
events, defined as cardiac death, any myocardial in- group because of 1 patient who developed subacute
farction, or target lesion revascularization with repeat thrombosis. This patient was treated with repeat perpercutaneous coronary intervention or coronary artery cutaneous coronary intervention and eventually develbypass graft after SES implantation versus BMS im- oped Q-wave myocardial infarction. Bleeding compliplantation. Death was considered cardiac unless oth- cations in the 2 groups were also not significantly
erwise demonstrated. Myocardial infarction was doc- different.
umented by the presence of a new pathologic Q wave
Angiographic results: Baseline and follow-up anon an electrocardiogram, according to the Minnesota giographic results are presented in Tables 3 and 4.
Code (F-11), or an increase in creatinine kinase to ⬎2 There were no significant differences between the
times the upper limit of normal. Anginal symptoms at groups in terms of reference diameter, minimum lufollow-up and angiographic restenosis after SES im- minal diameter after stenting, lesion length, CTO
plantation versus BMS implantation were secondary length, and stent length. Angiographic follow-up at 6
end points. These symptoms were graded according to months was performed in 67% of the SES group and
the classification of the Canadian Cardiovascular in 97% of the BMS group. In the SES group, 75% of
Society.
patients consented to 12-month follow-up angiograStatistical analysis: Categorical values are ex- phy. At follow-up after 6 and 12 months, the SES
pressed as absolute numbers and percent values. Con- group had significantly larger minimum luminal ditinuous variables are expressed as mean ⫾ SD. Dif- ameters because of a significantly smaller late loss.
ferences between groups were assessed with Target lesion revascularization and target vessel recontinuity-adjusted chi-square test or Fisher’s exact vascularization were also significantly lower in the
test for categorical variables and with a 2-tailed un- SES group. The restenosis rates at follow-up, defined
paired Student’s t test or Mann-Whitney U statistical as ⬎50% stenosis, were 32% in the BMS group and
test for continuous variables. Event-free survival rates 2% in the SES group (p ⬍0.001). Reocclusion rates
were estimated with the Kaplan-Meier method, with were 6% in the BMS group and 0% in the SES group
the difference between groups assessed with a log- (p ⬍0.001). Figure 1 shows the cumulative distriburank test of significance. A p value ⬍0.05 was con- tion curve of changing minimum luminal diameter of
sidered statistically significant.
the BMS and SES groups. The cumulative distribution
curve of the SES group did not differ from baseline
(after stenting) to follow-up.
RESULTS
Patients: From March 2002 to March 2003, 60 of
Events and clinical outcome: Long-term clinical out66 screened patients who had CTO (91%) were suc- come is presented in Table 5. Clinical follow-up data
cessfully recanalized, implanted with SES, and en- were available from 100% of patients in the 2 groups.
rolled in the multicenter SES registry. In the same Three nonfatal myocardial infarctions were observed,
TABLE 4 Angiographic Characteristics and Clinical Outcome (Follow-up)
CORONARY ARTERY DISEASE/SIROLIMUS-ELUTING STENT IN CTO
163
FIGURE 2. Elective implantation of the Cypher SES after successful recanalization improved left ventricular (LV) ejection fraction
significantly.
FIGURE 1. Cumulative distribution curves of mininum luminal diameter (MLD) assigned to (A) BMS and (B) SES implantations.
Binary restenosis rates of the BMS and SES groups were 27.5%
and 2.2%, respectively. mo f/u ⴝ months follow-up.
TABLE 5 Long-term Clinical Outcome (12 months)
Major adverse cardiac event
Death
Myocardial infarction
Coronary bypass
Repeat percutaneous
coronary intervention
Any event
BMS
(n ⫽ 120)
SES
(n ⫽ 60)
p Value
0
3 (3%)
7 (6%)
44 (37%)
0
0
0
2 (3%)
NS
NS
0.01
0.001
50 (42%)
2 (3%)
0.001
and 7 patients in the BMS group underwent coronary
artery bypass grafting. Repeat percutaneous coronary
intervention was performed in 37% of patients in the
BMS group versus only 3% of patients in the SES
group. Ten patients in the BMS group and 1 patient in
the SES group showed no involvement of the study
vessel and were treated by progression of the disease
in other vessels; the other 28% of patients in the BMS
group and 1 patient in the SES group underwent target
lesion revascularization procedures. Overall, surgical
or percutaneous target lesion revascularization after 6
months was performed in 42% of patients in the BMS
group and in 3% of patients in the SES group. Stress
164 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
testing and left ventricular function data analysis (Figure 2) showed a strong tendency toward a better
outcome in patients who received SESs. Patients in
the SES group showed a significantly lower percentage of positive results from stress testing at follow-up
than did those in the BMS group (3% vs 37%). This
difference was statistically significant. Estimated
probabilities of freedom from major adverse cardiac
events at 6 months were 92% in the BMS group and
100% in the SES group, and probabilities of freedom
from major adverse cardiac events and repeat percutaneous coronary intervention were 58% in the BMS
group and 97% in the SES group (p ⬍0.001). This
difference was essentially determined by a higher
estimated 6-month survival rate without target lesion
revascularization in the SES group. Kaplan-Meier
analyses of freedom from major adverse cardiac
events and freedom from major adverse cardiac events
and repeat percutaneous coronary intervention are
shown in Figure 3.
DISCUSSION
In the present study, we report the first clinical
experience with the Cypher SES for the management
of CTO. The major findings of this study were as
follows. First, implantation of SES was feasible and
safe for the management of CTO; no cases of subacute
stent thrombosis were noted. Second, late loss 6
months after SES implantation seemingly decreased
compared with late loss in the BMS group. Third, the
angiographic restenosis rate was 2% in the SES group
compared with 32% in the BMS group. Fourth, the
target lesion revascularization rate of 2% in the SES
group was surprisingly lower than the rate of 23% in
the BMS group.
Several clinical studies have documented the challenge of controlling CTO because of the lower success
rate and high incidence of restenosis (30% to
60%).15,16 The recent availability of new specialized
guidewires or more sophisticated techniques for crossing occluded arteries has contributed to the higher
success rate of CTO management.17–21 In addition,
JANUARY 15, 2005
enotic effect after management of CTO without the
attendant risk of stent thrombosis observed with other
effective treatment modalities. Long-term follow-up will
be required to determine whether the inhibition of restenosis observed in this study is sustained. Randomized
controlled trials will be necessary.
Acknowledgment: We are indebted to Shigeru
Saito, MD, Shonan Kamakura General Hospital, Japan for valuable comments. We also thank Ritsuko
Otaka for diligent work in organizing and typing the
manuscript.
1. Heyder M, Engel HJ, Hormann E. Angioplasty of chronic coronary artery
FIGURE 3. Kaplan-Meier analysis of survival rates free of (A) major adverse cardiac events (MACE) and (B) MACE ⴙ target lesion
revascularization (TLR) in patients assigned to BMS implantation
(solid line) or SES implantation (dotted line).
randomized studies have demonstrated that stent implantation decreases restenosis and reocclusion
rates.3–9 However, the remaining restenosis is mainly
attributed to reoccurrence of major adverse cardiac
events. Therefore, solving the problem of restenosis is
apparently the key to improving the long-term outcome in these patients.
In this study, notwithstanding the challenging population treated, we found strikingly similar results in
terms of suppression of neointimal proliferation to that
reported previously in low-risk patient populations.11,22,23 Acute procedural, in-hospital, and 12month outcomes were favorable. At 6- and 12-month
angiographic follow-up, only 1 patient required target
lesion revascularization due to having a gap between 2
stents. In the remaining patient, late luminal loss averaged 0.08 mm. This was extremely low compared
with the BMS group and provides an extremely low
incidence of restenosis.
This was a small observational study. However, the
lack of stent thrombosis, myocardial infarction, or death
and only 1 case of repeat percutaneous coronary intervention indicate that antiproliferative and anti-inflammatory actions of SES may provide a long-lasting antirest-
occlusions: impact on functional status and left ventricular performance. Z
Kardiol 1996;85:125–132.
2. Bell MR, Berger PB, Bresnahan JF, Reeder GS, Bailey KR, Holmes DR Jr.
Initial and long-term outcome of 354 patients after coronary balloon angioplasty
of total coronary occlusions. Circulation 1992;85:1003–1011.
3. Sirnes PA, Golf S, Myreng Y, Molstad P, Emanuelsson H, Albertsson P,
Brekke M, Mangschau A, Endresen K, Kjekshus J. Stenting In Chronic Coronary
Occlusion (SICCO): a randomized, controlled trial of adding stent implantation
after successful angioplasty. J Am Coll Cardiol 1996;28:1444 –1451.
4. Hancock J, Thomas MR, Holmberg S, Wainwright RJ, Jewitt DE. Randomized
trial of elective stenting after successful percutaneous transluminal coronary
angioplasty of occluded coronary arteries. Heart 1998;79:18 –23.
5. Rubartelli P, Niccoli L, Verna E, Giachero C, Zimarino M, Fontanelli A,
Vassanelli C, Campolo L, Martuscelli E, Tommasini G. Stent implantation versus
balloon angioplasty in chronic coronary occlusions: results from the GISSOC
trial. Gruppo Italiano di Studio sullo Stent nelle Occlusioni Coronariche. J Am
Coll Cardiol 1998;32:90 –96.
6. Buller CE, Dzavik V, Carere RG, Mancini GBJ, Barbeau G, Lazzam C,
Anderson TJ, Knudtson ML, Marquis JF, Suzuki T, et al. Primary stenting versus
balloon angioplasty in occluded coronary arteries: the Total Occlusion Study of
Canada (TOSCA). Circulation 1999;100:236 –242.
7. Sievert H, Rohde S, Utech A, Schulze R, Scherer D, Merle H, Ensslen R,
Schrader R, Spies H, Fach A. Stent or angioplasty after recanalization of chronic
coronary occlusions? (The SARECCO Trial). Am J Cardiol 1999;84:386 –390.
8. Hoher M, Wohrle J, Grebe OC, Kochs M, Osterhues HH, Hombach V,
Buchwald AB. A randomized trial of elective stenting after balloon recanalization
of chronic total occlusions. J Am Coll Cardiol 1999;34:722–729.
9. Lotan C, Rozenman Y, Hendler A, Turgeman Y, Ayzenberg O, Beyar R,
Krakover R, Rosenfeld T, Gotsman MS. Stents in total occlusion for restenosis
prevention. The multicentre randomized STOP study. The Israeli Working Group
for Interventional Cardiology. Eur Heart J Med 2000;21:1960 –1966.
10. Olivari Z, Rubartelli P, Piscione F, Ettori F, Fontanelli A, Salemme L,
Giachero C, Di Mario C, Gabrielli G, Spedicato L, Bedogni F. Immediate results
and one-year clinical outcome after percutaneous coronary interventions in
chronic total occlusions. J Am Coll Cardiol 2003;41:1672–1678.
11. Morice MC, Serruys PW, Sousa JE, Fajadet J, Hayashi EB, Perin M,
Colombo A, Schuler G, Barragan P, Guagliumi G, et al. A randomized comparison of a sirolimus-eluting stent with a standard stent for coronary revascularization. N Engl J Med 2002;346:1773–1780.
12. Degertekin M, Serruys PW, Foley DP, Tanabe K, Regar E, Vos J, Smits PC,
van der Giessen WJ, van den Brand M, de Feyter P, Popma JJ. Persistent
inhibition of neointimal hyperplasia after Sirolimus-eluting stent implantation:
long-term (up to 2 years) clinical, angiographic, and intravascular ultrasound
follow-up. Circulation 2002;106:1610 –1613.
13. Rensing BJ, Vos J, Smits PC, Foley DP, van den Brand MJ, van der Giessen
WJ, de Feijter PJ, Serruys PW. Coronary restenosis elimination with a sirolimus
eluting stent: first European human experience with 6-month angiographic and
intravascular ultrasonic follow-up. Eur Heart J 2001;22:2125–2130.
14. Lefevre T, Louvard Y, Loubeyre C, Dumas P, Piechaud JF, Krol M,
Benslimane A, Premchand RK, Morice MC. A randomized study comparing two
guidewire strategies for angioplasty of chronic total coronary occlusion. Am J
Cardiol 2000;85:1144 –1147.
15. Violaris AG, Melkert R, Serruys PW. Long-term luminal renarrowing after
successful elective coronary angioplasty of total occlusions. A quantitative angiographic analysis. Circulation 1995;91:2140 –2150.
16. Berger PB, Holmes DR Jr, Ohman EM, O’Hanesian MA, Murphy JG,
Schwartz RS, Serruys PW, Faxon DP. Restenosis, reocclusion and adverse
cardiovascular events after successful balloon angioplasty of occluded versus
nonoccluded coronary arteries. Results from the Multicenter American Research
Trial With Cilazapril After Angioplasty to Prevent Transluminal Coronary Obstruction and Restenosis (MARCATOR). J Am Coll Cardiol 1996;27:1–7.
CORONARY ARTERY DISEASE/SIROLIMUS-ELUTING STENT IN CTO
165
17. Allemann Y, Kaufmann UP, Meyer BJ, Gunnes P, Urban P, Mulhauser B,
Dorsaz PA, Meier B. Magnum wire for percutaneous coronary balloon angioplasty in 800 total chronic occlusions. Am J Cardiol 1997;80:634 – 637.
18. Reimers B, Camassa N, Di Mario C, Akiyama T, Di Francesco L, Finci L,
Colombo A. Mechanical recanalization of total coronary occlusions with the use
of a new guide wire. Am Heart J 1998;135:726 –731.
19. Kahler J, Koster R, Brockhoff C, Reimers J, Baldus S, Terres W, Meinertz T,
Hamm CW. Initial experience with a hydrophilic-coated guidewire for recanalization
of chronic coronary occlusions. Catheter Cardiovasc Interv 2000;49:45–50.
20. Serruys PW, Hamburger JN, Koolen JJ, Fajadet J, Haude M, Klues H, SeabraGomes R, Corcos T, Hamm C, Pizzuli L, et al. Total occlusion trial with angioplasty
by using laser guidewire. The TOTAL trial. Eur Heart J 2000;21:1797–1805.
166 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
21. Piscione F, Galasso G, Maione AG, Pisani A, Golino P, Leosco D, Indolfi C,
Chiariello M. Immediate and long term outcome of recanalization of chronic total
coronary occlusions. J Interv Cardiol 2002;15:173–179.
22. Sousa JE, Costa MA, Abizaid A, Abizaid AS, Feres F, Pinto IMF, Seixas AC,
Staico R, Mattos LA, Sousa AGMR, et al. Lack of neointimal proliferation after
implantation of sirolimus-coated stents in human coronary arteries: a quantitative
coronary angiography and three-dimensional intravascular ultrasound study. Circulation 2001;103:192–195.
23. Moses JW, Leon MB, Popma JJ, Fitzgerald PJ, Holmes DR, O’Shaughnessy
C, Caputo RP, Kereiakes DJ, Williams DO, Teirstein PS, et al. Sirolimus-eluting
stents versus standard stents in patients with stenosis in a native coronary artery.
N Engl J Med 2002;349:1315–1323.
JANUARY 15, 2005
Impact of Baseline Renal Function on
Mortality After Percutaneous Coronary
Intervention With Sirolimus-Eluting
Stents or Bare Metal Stents
Pedro A. Lemos, MD, PhD, Chourmouzios A. Arampatzis, MD, PhD,
Angela Hoye, MB, ChB, MRCP, Joost Daemen, Andrew T.L., Ong, MBBS,
Francesco Saia, MD, Willem J. van der Giessen, MD, PhD,
Eugene P. McFadden, MB, ChB, MD, Georgios Sianos, MD, PhD, Pieter C. Smits, MD,
Pim de Feyter, MD, PhD, Sjoerd H. Hofma, MD, Ron T. van Domburg, PhD, and
Patrick W. Serruys, MD, PhD
PhD,
Renal impairment is an important predictor of mortality
after percutaneous coronary intervention and may increase the restenosis rate. However, the relation between restenosis and the risk of death in patients who
have renal impairment remains unclear. We evaluated
the incidences of repeat revascularization and mortality
in patients who had renal impairment and those who
did not and who received sirolimus-eluting stents or
bare stents. A total of 1,080 consecutive patients treated
for 1 year had available data to calculate baseline
creatinine clearance. Patients received bare stents (first 6
months, n ⴝ 543) or sirolimus-eluting stents (last 6
months, n ⴝ 537) and were grouped according to the
presence or absence of renal impairment (creatinine
clearance <60 ml/min). Patients who had renal impairment had a higher mortality rate at 1 year (7.6% vs
2.5%, hazard ratio 3.14, 95% confidence interval 1.68
to 5.88, p <0.01), with no differences in mortality be-
tween patients who received bare stents and those who
received sirolimus-eluting stents (hazard ratio 0.91, 95%
confidence interval 0.49 to 1.68, p ⴝ 0.8). The incidence
of target vessel revascularization decreased significantly
in patients who were treated with sirolimus-eluting
stents and did not have renal impairment (hazard ratio
0.59, 95% confidence interval 0.39 to 0.90, p ⴝ 0.01)
and in those who had decreased renal function (hazard
ratio 0.37, 95% confidence interval 0.15 to 0.90, p ⴝ
0.03). Thus, sirolimus-eluting stents compared with conventional stents decreased clinical restenosis in patients
who had renal impairment. However, this benefit was
not paralleled by a decrease in the risk of death in this
population. It seems unlikely that restenosis could be a
contributing factor that influenced the increased mortality
of patients who had impaired renal function. 䊚2005 by
Excerpta Medica Inc.
(Am J Cardiol 2005;95:167–172)
hronic kidney disease has been shown to strongly
increase the risk of short- and long-term adverse
C
events in patients who have atherosclerotic dis-
clinical trials conducted to date have excluded this
subset of patients who had impaired renal function
and, as a consequence, the effect of SES implantation
on the outcomes of this subset of patients is currently
unknown. The present study evaluated the effect of
baseline renal function on 1-year mortality of patients
who had been treated with conventional bare stents
or SESs.
ease,1–11 and patients who have renal failure have been
reported to have higher mortality rates after successful
percutaneous coronary intervention.10 –12 Neither surgical nor percutaneous revascularization has been
shown to eliminate the increased risk of patients who
have renal impairment.7–12 Sirolimus-eluting stents
(SESs) have proved to markedly decrease neointimal
growth and in-stent restenosis compared with conventional stents, with an impressive decrease in the risk of
subsequent repeat revascularization.13–15 However, all
From the Heart Institute (InCor), University of São Paulo Medical
School, São Paulo, Brazil; and the Erasmus Medical Center, Thoraxcenter, Rotterdam, The Netherlands. This study was supported by the
Erasmus Medical Center, Erasmus University, Rotterdam, The Netherlands, and by an unrestricted institutional grant from Cordis, a Johnson
& Johnson Company, Miami Lakes, Florida. Manuscript received June
23, 2004; revised manuscript received and accepted August 31,
2004.
Address for reprints: Patrick W. Serruys, MD, PhD, Thoraxcenter,
Bd-406, Dr. Molewaterplein 40, 3015-GD Rotterdam, The Netherlands.
E-mail: p.w.j.c.serruys@erasmusmc.nl.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
METHODS
Patient population: Since April 2002, SES implantation (Cypher; Cordis, Johnson & Johnson, Cordis
Europa NV, Roden, The Netherlands) has been
adopted as the default interventional strategy for all
patients who are treated in our institution, as described
elsewhere.16,17 For comparison, a control group was
composed of all consecutive patients who had been
treated with conventional bare stents in the period
before the introduction of SESs.16,17 From October
2001 to October 2002 (6-month enrollment for presirolimus and sirolimus phases), 1,262 consecutive
patients who were not on dialysis were treated with
bare stents or SESs in the 2 study periods. Among
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.089
167
these patients, 1,080 (86%) had preprocedural serum
creatinine measured in our institution and comprised
the present study population (bare stent group, n ⫽
543 patients; SES group, n ⫽ 537 patients). This
protocol was approved by the local ethics committee
and was conducted in accordance with the Declaration
of Helsinki. Written, informed consent was obtained
from every patient.
Interventional procedures: All interventions were
performed according to standard techniques, and the
final strategy was left to the discretion of the operators. Angiographic success was defined as residual
stenosis ⬍30% by visual analysis with Thrombolysis
In Myocardial Infarction grade 3 anterograde flow.
Periprocedural use of a glycoprotein IIb/IIIa inhibitor
was left to the discretion of the operator. All patients
were advised to maintain lifelong use of aspirin. Clopidogrel was prescribed for ⱖ1 month in the bare stent
group. For patients who were treated with SESs, clopidogrel was recommended for 3 months, unless patients had ⱖ1 of the following characteristics (in
which case clopidogrel was maintained for ⱖ6
months): multiple SES implantations (⬎3 stents), total
stented length ⬎36 mm, chronic total occlusion, bifurcations, and in-stent restenosis.
Clinical follow-up and end points: In-hospital clinical
information was retrieved from an electronic database of
patients maintained in our hospital and by review of
hospital records for those who had been discharged to
referring hospitals. Postdischarge survival status was obtained from the municipal civil registries. Repeat revascularization procedures (surgical or percutaneous) and
rehospitalizations were prospectively collected during
follow-up. Patients were directly approached and/or the
referring physicians and institutions contacted whenever
necessary for additional information.
The primary end point of the present study was
all-cause mortality at 1 year. Incidence of target vessel
revascularization was assessed to evaluate the antirestenotic effect of SESs versus bare stents. Target
vessel revascularization was defined as a reintervention (surgical or percutaneous) to treat any lesion
located in the same epicardial vessel treated at the
index procedure.
Renal function evaluation: The closest creatinine
values before the procedure were used to calculate
baseline creatinine clearance according to the formula
proposed by Cockcroft and Gault18: creatinine clearance (milliliters/minute) ⫽ (140 ⫺ age) ⫻ weight
(kilograms) ⫼ 72 ⫻ serum creatinine (milligrams/
deciliter) (⫻ 0.85 for women). Renal impairment was
defined as a calculated creatinine clearance ⬍60 ml/min,
a cut-off value previously proposed by the National
Kidney Foundation’s Kidney Disease Outcome Quality
Initiative Advisory Board to identify patients who have
moderate renal impairment5 and the American Heart
Association’s Councils on Kidney in Cardiovascular
Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention.6
Statistical analysis: Continuous variables are presented as mean ⫾ SD and were compared with Student’s t test. Categorical variables are presented as
168 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
counts and percentages and were compared with Fisher’s exact test. An unadjusted cumulative incidence of
death and target vessel revascularization was evaluated by the Kaplan-Meier method. Cox’s proportional
hazards models were used to examine the effects of
renal impairment, SES implantation, and the interaction between stent type versus renal function on clinical end points. To adjust for baseline differences
between study groups, all variables associated with
the clinical end points at univariate analyses (p ⱕ0.2
for selection) were tested in multivariate analyses to
identify independent predictors of 1-year mortality
(tested variables were gender, acute myocardial infarction at admission, 3-vessel disease, hypercholesterolemia, current smoking status, diabetes, angiographic success, statin prescription, left anterior
descending artery stenting, left main coronary artery
stenting, bypass graft stenting, and renal function impairment) and target vessel revascularization (tested
variables were previous bypass surgery, acute myocardial infarction at admission, 3-vessel disease, hypercholesterolemia, current smoking status, diabetes,
left main coronary artery stenting, bypass graft stenting, number of stents implanted, treatment of in-stent
restenosis, and SES use). The final models were built
by backward stepwise variable selection, with a p
value ⬍0.05 used as a criterion for entry and removal
of variables. All reported p values are 2-tailed, and a
p value ⬍0.05 was considered statistically significant.
RESULTS
Baseline and procedural characteristics: Among 543
patients who had been treated with bare stents, 92
(17%) had renal dysfunction at baseline; among 537
patients who had been treated with SESs, renal dysfunction was present in 94 (18%). Table 1 presents
baseline and procedural characteristics of patients who
had normal renal function or renal impairment, according to the type of stent used. Pooled over the stent
type used, patients who had renal impairment were
older, more frequently women, and more often had
hypertension, previous coronary surgery, 3-vessel disease, bypass graft stenting, and more stents implanted
per procedure. In addition, patients who had a lower
rate of clearance weighed less, were shorter, and were
less likely to smoke and to have received stenting in
the right coronary artery. Overall, in patients who did
or did not have renal impairment, average serum creatinine level was 0.9 ⫾ 0.2 versus 1.3 ⫾ 0.4 mg/dl,
respectively (p ⬍0.01), and average creatinine clearance was 99 ⫾ 27 versus 49 ⫾ 9 mg/dl, respectively
(p ⬍0.01).
Preprocedural baseline characteristics of patients
who had been treated with SESs and those who had
been treated with bare stents were similar, except for
a lower frequency of current smokers and a higher rate
of previous percutaneous intervention and treatment
of restenotic lesions in patients who received SESs
(Table 1). The average creatinine clearance was similar between patients who had been treated with SESs
or with bare stents. In the SES group, use of glycoprotein IIb/IIIa inhibitors was lower and the number of
JANUARY 15, 2005
TABLE 1 Baseline and Procedural Characteristics of Patients With or Without Renal Impairment Treated With Bare Stents or
Sirolimus-eluting Stents
Normal Renal Function
Variable
Men
Age (yrs)
Height (cm)
Weight (kg)†
Hypercholesterolemia or use of lipid-lowering drug
Hypertension
Current smoker†
Diabetes mellitus
Insulin dependent
Non–insulin dependent
Previous myocardial infarction
Previous bypass surgery
Previous percutaneous coronary intervention†
Clinical presentation
Stable angina pectoris
Unstable angina pectoris
Acute myocardial infarction
Coronary vessel disease
1-vessel
2-vessel
3-vessel
Coronary vessel treated
Right coronary artery
Left anterior descending
Left circumflex artery
Left main coronary
Bypass graft
No. of in-stent restenosis (ⱖ1 lesion)†
No. of stents implanted per patient‡
Angiographic success for all lesions
Periprocedural use of glycoprotein IIb/IIIa inhibitor†
Statin use at discharge
ACE inhibitor use at discharge
Clopidogrel prescription (months)‡
Serum creatinine (mg/dl)
Creatinine clearance (ml/min)
Renal Impairment
Bare Stent
(n ⫽ 451)
SES
(n ⫽ 443)
Bare Stent
(n ⫽ 92)
SES
(n ⫽ 94)
79%
59 ⫾ 11
174 ⫾ 8
84 ⫾ 13
56%
36%
37%
14%
4%
10%
39%
7%
21%
73%
59 ⫾ 10
173 ⫾ 9
82 ⫾ 14
58%
39%
31%
18%
5%
13%
33%
9%
27%
48%
72 ⫾ 8
167 ⫾ 10
71 ⫾ 11
51%
49%
20%
20%
9%
11%
39%
25%
25%
50%
72 ⫾ 9
167 ⫾ 9
70 ⫾ 11
62%
50%
17%
20%
6%
14%
31%
21%
31%
47%
34%
19%
50%
35%
16%
55%
35%
10%
38%
47%
15%
49%
36%
16%
48%
34%
19%
36%
29%
35%
33%
31%
36%
39%
57%
34%
4%
3%
4%
1.9 ⫾ 1.1
97%
39%
66%
30%
3.0 ⫾ 2.1
0.9 ⫾ 0.8
101 ⫾ 29
39%
57%
31%
2%
3%
10%
2.1 ⫾ 1.4
97%
17%
65%
25%
4.2 ⫾ 2.0
0.9 ⫾ 0.2
98 ⫾ 25
32%
52%
33%
5%
12%
6%
2.1 ⫾ 1.4
98%
29%
59%
26%
3.0 ⫾ 1.8
1.3 ⫾ 0.4
49 ⫾ 9
28%
63%
33%
5%
11%
13%
2.4 ⫾ 1.6
98%
21%
62%
28%
4.1 ⫾ 2.0
1.3 ⫾ 0.4
50 ⫾ 9
p Value*
⬍0.01
⬍0.01
⬍0.01
⬍0.01
0.9
⬍0.01
⬍0.01
0.2
0.1
0.7
0.8
⬍0.01
0.2
0.1
⬍0.01
0.02
0.9
1.0
0.2
⬍0.01
0.2
⬍0.01
0.8
0.5
0.2
1.0
0.9
⬍0.01
⬍0.01
*Patients who had normal renal function versus patients who had renal impairment and were pooled over stent type group.
†
p ⬍0.05 for bare stents versus SESs pooled over renal function group.
‡
p ⬍0.01 for bare stents versus SESs pooled over renal function group.
Values are mean ⫾ SD or percentages.
ACE ⫽ angiotensin-converting enzyme.
stents implanted per procedure was greater (Table 1).
Clopidogrel use was longer in the SES group, according to the predefined management protocol. There
were no differences with regard to postprocedural
prescription of statins between the SES and bare stent
groups.
One-year mortality: Clinical follow-up data were
available for 1,074 patients (99.4%; median follow-up
period 421 days, interquartile range 391 to 459).
When all patients were pooled together, regardless of
SES or bare stent use, the unadjusted risk of death at
1 year was significantly higher in patients who had
renal impairment than in patients who had normal
renal function (7.6% vs 2.5%, respectively, hazard
ratio 3.14, 95% confidence interval [CI] 1.68 to 5.88,
p ⬍0.01; Figure 1). Similarly, when analyzed separately, baseline renal impairment significantly increased the risk of death in patients who had been
treated with bare stents or SESs (Figure 1). When
evaluated irrespective of renal function, patients who
had been treated with bare stents or SESs had similar
1-year mortality rates (3.6% vs 3.2%, respectively,
hazard ratio 0.91, 95% CI 0.49 to 1.68, p ⫽ 0.8;
Figure 2). The interaction factor for the relation of the
effects of renal impairment and stent type on risk of
death was not significant (p ⫽ 0.7 for interaction).
Independent predictors of mortality at 1 year by
multivariate analysis are presented in Table 2. Use of
SESs had no influence in risk of death at 1 year
(adjusted hazard ratio 1.10, 95% CI 0.59 to 2.07, p ⫽
0.8).
Repeat revascularization: Overall, SES implantation significantly decreased the incidence of target
vessel revascularization at 1 year compared with bare
stent implantation (6.9% vs 13.1%, unadjusted hazard
ratio 0.54, 95% CI 0.37 to 0.79, p ⬍0.01). SES implantation was effective in decreasing the risk of target
vessel revascularization in patients who did not have
CORONARY ARTERY DISEASE/RENAL FUNCTION AFFECTS OUTCOMES AFTER SIROLIMUS STENTS
169
FIGURE 2. Unadjusted 1-year incidence of all-cause death for
patients who had been treated with bare stents versus those
who had been treated with SESs pooled over renal function.
TABLE 2 Independent Multivariate Predictors of One-year
Mortality and Target Vessel Revascularization Rates
Variable
1-yr mortality rate
Women
Renal impairment
Acute myocardial infarction
3-Vessel disease
SES use
1-yr target vessel
revascularization rate
Renal impairment
SES use
Current smoker
Treatment of in-stent restenosis
No. of stents implanted
Hazard
Ratio
95% CI
p
Value
2.00
2.15
3.00
2.75
1.10
1.04–3.85
1.10–4.28
1.54–5.86
1.44–5.22
0.59–2.07
0.04
0.03
⬍0.01
⬍0.01
0.8
1.22
0.43
0.56
3.29
1.23
0.79–1.88
0.29–0.64
0.36–0.88
2.05–5.28
1.09–1.38
0.4
⬍0.01
0.01
⬍0.01
⬍0.01
0.37, 95% CI 0.15 to 0.90, p ⫽ 0.03). At multivariate
analysis, SES use remained an important factor that
decreased the risk of repeat revascularization (adjusted hazard ratio 0.43, 95% CI 0.29 to 0.64,
p ⬍0.01). Importantly, the presence of renal impairment did not significantly influence the risk of target
vessel revascularization (adjusted hazard ratio 1.22,
95% CI 0.79 to 1.88, p ⫽ 0.4). Other independent
predictors of 1-year target vessel revascularization are
presented in Table 2.
FIGURE 1. Unadjusted 1-year incidence of all-cause death after
percutaneous coronary intervention for patients who had renal
impairment and those who did not. Total study population
pooled over stent type (top) and patients who had been treated
with bare stents (middle) or SESs (bottom).
renal impairment (7.2% vs 11.8%, unadjusted hazard
ratio 0.59, 95% CI 0.39 to 0.90, p ⫽ 0.01) and in those
who did (5.6% vs 19.6%, unadjusted hazard ratio
170 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
DISCUSSION
The main finding of the present study was that
impaired renal function significantly increases 1-year
mortality after percutaneous coronary revascularization, regardless of stent type. Despite the clear antirestenotic effect of SESs, which markedly decreased
the incidence of target vessel revascularization compared with bare stents, mortality rates in patients who
had renal dysfunction and those who did not were
similar for the 2 treatment strategies.
JANUARY 15, 2005
Impaired renal function has been shown to negatively influence survival rates after percutaneous intervention.7–12 Although patients who have renal impairment are well known to have an increased
prevalence of associated risk factors, it has been identified in our series and in previous reports as an
important independent predictor of mortality. Several
inflammatory, procoagulant, and atherogenic markers
have been described in patients who have renal impairment,19 –26 which might accelerate disease progression and account for a higher tendency to acute
events. Some of these factors have been associated
with an increased risk of late restenosis.27–29 Accordingly, patients who have end-stage renal failure have
been shown to present with increased levels of fibrinogen and higher rates of restenosis than patients not on
dialysis.29 It has been hypothesized that the high incidence of restenosis may contribute to the increased
mortality rate seen in patients who have renal impairment.10,12 Our results challenge this concept by demonstrating that the striking decrease in clinical restenosis after SES implantation was not paralleled by any
decrease in mortality rate among patients who had
renal impairment.
Although drug-eluting stents did not decrease mortality risk after coronary intervention, the decrease in
restenosis represents an important therapeutic
achievement for the clinical management of patients
who have renal impairment. In our series, SES implantation decreased the risk of repeat revascularization by ⬎50% of that observed with bare stents in
patients who had renal impairment. The marked decrease in the risk of repeat revascularization may shift
the focus of clinical attention after percutaneous intervention from restenosis prevention toward the institution of more aggressive disease-modifying
strategies.
Although the present study has limitations related
to its nonrandomized nature, the bare stent and SES
groups had comparable baseline characteristics. Creatinine clearance was calculated because it has been
shown to correlate well with actual values18 and provide a better estimate of renal function than serum
creatinine alone.30 Complete data for preprocedural
creatinine clearance calculation were available for
86% of patients, which may have introduced a selection bias in the analysis. Nevertheless, patients who
had been excluded from this study due to missing
creatinine clearance data had a 1-year mortality rate of
⬃5.5%, which was intermediate between patients who
had renal impairment and those who did not, indicating that those who had been excluded likely included
similar proportions of patients who had renal impairment and those who did not.
1. Al Suwaidi J, Reddan DN, Williams K, Pieper KS, Harrington RA, Califf RM,
Granger CB, Ohman EM, Holmes DR Jr. Prognostic implications of abnormalities in renal function in patients with acute coronary syndromes. Circulation
2002;106:974 –980.
2. Shlipak MG, Simon JA, Grady D, Lin F, Wenger NK, Furberg CD. Renal
insufficiency and cardiovascular events in postmenopausal women with coronary
heart disease. J Am Coll Cardiol 2001;38:705–711.
3. Mann JF, Gerstein HC, Pogue J, Bosch J, Yusuf S. Renal insufficiency as a
predictor of cardiovascular outcomes and the impact of ramipril: the HOPE
randomized trial. Ann Intern Med 2001;134:629 – 636.
4. Gibson CM, Pinto DS, Murphy SA, Morrow DA, Hobbach HP, Wiviott SD,
Giugliano RP, Cannon CP, Antman EM, Braunwald E. Association of creatinine
and creatinine clearance on presentation in acute myocardial infarction with
subsequent mortality. J Am Coll Cardiol 2003;42:1535–1543.
5. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic
kidney disease: evaluation, classification, and stratification. Am J Kidney Dis
2002;39:S1–S246.
6. Sarnak MJ, Levey AS, Schoolwerth AC, Coresh J, Culleton B, Hamm LL,
McCullough PA, Kasiske BL, Kelepouris E, Klag MJ, et al. Kidney disease as a
risk factor for development of cardiovascular disease: a statement from the
American Heart Association Councils on Kidney in Cardiovascular Disease, High
Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention.
Circulation 2003;108:2154 –2169.
7. Gruberg L, Dangas G, Mehran R, Mintz GS, Kent KM, Pichard AD, Satler LF,
Lansky AJ, Stone GW, Leon MB. Clinical outcome following percutaneous
coronary interventions in patients with chronic renal failure. Catheter Cardiovasc
Interv 2002;55:66 –72.
8. Szczech LA, Best PJ, Crowley E, Brooks MM, Berger PB, Bittner V, Gersh BJ,
Jones R, Califf RM, Ting HH, et al. Outcomes of patients with chronic renal
insufficiency in the bypass angioplasty revascularization investigation. Circulation 2002;105:2253–2258.
9. Dixon SR, O’Neill WW, Sadeghi HM, Stone GW, Brodie B, Cox DA, Garcia
E, Mattos L, Grines LL, Boura JA, et al. Usefulness of creatinine clearance in
predicting early and late death after primary angioplasty for acute myocardial
infarction. Am J Cardiol 2003;91:1454 –1457.
10. Best PJ, Lennon R, Ting HH, Bell MR, Rihal CS, Holmes DR, Berger PB.
The impact of renal insufficiency on clinical outcomes in patients undergoing
percutaneous coronary interventions. J Am Coll Cardiol 2002;39:1113–1119.
11. Rubenstein MH, Harrell LC, Sheynberg BV, Schunkert H, Bazari H, Palacios
IF. Are patients with renal failure good candidates for percutaneous coronary
revascularization in the new device era? Circulation 2000;102:2966 –2972.
12. Herzog CA, Ma JZ, Collins AJ. Comparative survival of dialysis patients in the
United States after coronary angioplasty, coronary artery stenting, and coronary
artery bypass surgery and impact of diabetes. Circulation 2002;106:2207–2211.
13. Moses JW, Leon MB, Popma JJ, Fitzgerald PJ, Holmes DR, O’Shaughnessy
C, Caputo RP, Kereiakes DJ, Williams DO, Teirstein PS, et al. Sirolimus-eluting
stents versus standard stents in patients with stenosis in a native coronary artery.
N Engl J Med 2003;349:1315–1323.
14. Morice MC, Serruys PW, Sousa JE, Fajadet J, Ban Hayashi E, Perin M,
Colombo A, Schuler G, Barragan P, Guagliumi G, et al. A randomized comparison of a sirolimus-eluting stent with a standard stent for coronary revascularization. N Engl J Med 2002;346:1773–1780.
15. Lemos PA, Serruys PW, van Domburg RT, Saia F, Arampatzis CA, Hoye A,
Degertekin M, Tanabe K, Daemen J, Liu TK, et al. Unrestricted utilization of
sirolimus-eluting stents compared with conventional bare stent implantation in
the “real world”: the Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology Hospital (RESEARCH) registry. Circulation 2004;109:190 –195.
16. Lemos PA, Lee C, Degertekin M, Saia F, Tanabe K, Arampatzis CA, Hoye
A, van Duuren M, Sianos G, Smits PC, et al. Early outcome after sirolimuseluting stent implantation in patients with acute coronary syndromes. Insights
from the Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology Hospital
(RESEARCH) registry. J Am Coll Cardiol 2003;41:2093–2099.
17. Lemos PA, Serruys PW, van Domburg RT, Saia F, Arampatzis CA, Hoye A,
Degertekin M, Tanabe K, Daemen J, Liu TKK, et al. Unrestricted utilization of
sirolimus-eluting stents compared to conventional bare stent implantation in the
“real world”: The Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology
Hospital (RESEARCH) registry. Circulation 2004;109:190 –195.
18. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum
creatinine. Nephron 1976;16:31– 41.
19. Shlipak MG, Fried LF, Crump C, Bleyer AJ, Manolio TA, Tracy RP, Furberg
CD, Psaty BM. Elevations of inflammatory and procoagulant biomarkers in
elderly persons with renal insufficiency. Circulation 2003;107:87–92.
20. Henry RM, Kostense PJ, Bos G, Dekker JM, Nijpels G, Heine RJ, Bouter
LM, Stehouwer CD. Mild renal insufficiency is associated with increased cardiovascular mortality: the Hoorn Study. Kidney Int 2002;62:1402–1407.
21. Dzurik R, Spustova V, Janekova K. The prevalence of insulin resistance in
kidney disease patients before the development of renal failure. Nephron 1995;
69:281–285.
22. Sechi LA, Zingaro L, De Carli S, Sechi G, Catena C, Falleti E, Dell’Anna E,
Bartoli E. Increased serum lipoprotein(a) levels in patients with early renal
failure. Ann Intern Med 1998;129:457– 461.
23. Stuveling EM, Hillege HL, Bakker SJ, Gans RO, De Jong PE, De Zeeuw D.
C-reactive protein is associated with renal function abnormalities in a nondiabetic population. Kidney Int 2003;63:654 – 661.
24. Holvoet P, Donck J, Landeloos M, Brouwers E, Luijtens K, Arnout J,
Lesaffre E, Vanrenterghem Y, Collen D. Correlation between oxidized low
density lipoproteins and von Willebrand factor in chronic renal failure. Thromb
Haemost 1996;76:663– 669.
CORONARY ARTERY DISEASE/RENAL FUNCTION AFFECTS OUTCOMES AFTER SIROLIMUS STENTS
171
25. Thambyrajah J, Landray MJ, McGlynn FJ, Jones HJ, Wheeler DC, Townend
JN. Abnormalities of endothelial function in patients with predialysis renal
failure. Heart 2000;83:205–209.
26. Parsons DS, Reaveley DA, Pavitt DV, Brown EA. Relationship of renal
function to homocysteine and lipoprotein(a) levels: the frequency of the combination of both risk factors in chronic renal impairment. Am J Kidney Dis
2002;40:916 –923.
27. Otsuka M, Hayashi Y, Ueda H, Imazu M, Kohno N. Predictive value of
preprocedural fibrinogen concerning coronary stenting. Atherosclerosis 2002;
164:371–378.
172 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
28. Benchimol D, Bonnet J, Benchimol H, Drouillet F, Duplaa C, Couffinhal T,
Desgranges C, Bricaud H. Biological risk factors for restenosis after percutaneous
transluminal coronary angioplasty. Int J Cardiol 1993;38:7–18.
29. Schoebel FC, Gradaus F, Ivens K, Heering P, Jax TW, Grabensee B, Strauer
BE, Leschke M. Restenosis after elective coronary balloon angioplasty in patients
with end stage renal disease: a case-control study using quantitative coronary
angiography. Heart 1997;78:337–342.
30. Bostom AG, Kronenberg F, Ritz E. Predictive performance of renal function
equations for patients with chronic kidney disease and normal serum creatinine
levels. J Am Soc Nephrol 2002;13:2140 –2144.
JANUARY 15, 2005
Determinants of 30-Day Adverse Events
Following Saphenous Vein Graft
Intervention With and Without a Distal
Occlusion Embolic Protection Device
Gregory R. Giugliano, MD, SM, Richard E. Kuntz, MD, MSc, Jeffrey J. Popma, MD,
Donald E. Cutlip, MD, and Donald S. Baim, MD, on behalf of the Saphenous Vein
Graft Angioplasty Free of Emboli Randomized (SAFER) Trial Investigators
Distal balloon occlusion was approved as a means of
embolic protection during saphenous vein graft intervention based on its ability to decrease major adverse
clinical events (MACEs) by 42% in the 801-patient Saphenous Vein Graft Angioplasty Free of Emboli Randomized (SAFER) trial. However, the cost and technical complexity of this device have limited its widespread use
and prompted some to avoid its use in cases that appear
at “low risk” for complications. If predictors of MACEs
and their potential decrease by distal balloon occlusion
could be identified, this would have important clinical
implications in this challenging population. We therefore
used standard demographic and angiographic variables and 2 new angiographic markers (extent of graft
degeneration and estimated volume of plaque in the
target lesion) to construct multivariable logistic regres-
sion models of 30-day of MACEs in the SAFER trial.
Independent correlates of increased 30-day MACEs
were more extensive vein graft degeneration (p ⴝ
0.0001) and bulkier lesions (larger estimated plaque
volume, p ⴝ 0.0005). Use of a distal balloon occlusion
device was independently predictive of lower 30-day
rates of MACE (p ⴝ 0.01), with uniform benefit across
risk strata (no significant interaction between device use
and independent angiographic risk factors). Thus, the
risk of 30-day MACEs after percutaneous intervention in
aortocoronary saphenous vein grafts is increased in
more diffusely diseased grafts and in bulkier lesions, but
a significant benefit of the GuardWire was seen across
all levels of MACE risk rather than just those perceived to
be at highest risk. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:173–177)
lthough distal atheroemboli have been implicated
as the cause of major adverse clinical events
A
(MACEs) during stenting of saphenous vein graft
vessel diameters of 3 to 6 mm, ejection fraction
⬎25%, and serum creatinine levels ⬍2.5 mg/dl. The
primary end point was 30-day MACEs, defined as the
composite of death, myocardial infarction (creatine
kinase-MB level ⬎3 times the upper limit of normal),
emergency bypass surgery, or target vessel revascularization ⱕ30 days of the index procedure. Detailed
descriptions of the SAFER, including patient and lesion characteristics, participating hospitals, data gathering, analysis process, and end point definitions, have
been previously published.8
For the present analysis, standard clinical and angiographic variables were supplemented by 2 new ad
hoc angiographic variables derived from the core laboratory: SVG percent degeneration score, which was
defined in quartiles as the portion (0% to 25%, 26% to
50%, 51% to 75%, ⬎75%) of the treated graft with
evident (⬎20%) disease, and estimated lesion plaque
volume, which was defined as the volume of a cylinder whose diameter was equal to the reference vessel
diameter and whose length was the shoulder-to-shoulder lesion length, as measured by the core laboratory,
minus the volume of a cylinder of the same length and
a diameter equal to the measured minimal luminal
diameter (Appendix A). The former captures the state
of graft disease outside the target lesion, and the latter
captures the amount of plaque at risk for embolization
when the lumen of the target lesion is normalized by
stent placement.
(SVG) lesions,1– 8 only 1 study9 has reported correlates of MACEs after SVG percutaneous coronary
intervention that used distal embolic protection. We
therefore conducted this analysis of a selected set of
angiographic and demographic characteristics in the
Saphenous Vein Graft Angioplasty Free of Emboli
Randomized (SAFER) trial to identify the potential
determinants of MACEs at 30 days and their interaction with a randomized balloon occlusion device.
METHODS
The SAFER trial8 enrolled 801 patients between
June 1999 and August 2000 in a multicenter study in
which patients who underwent SVG graft intervention
were randomized to undergo stenting over a conventional guidewire or over the GuardWire distal protection device (Medtronic, Minneapolis, Minnesota). All
patients had coronary ischemia, culprit lesions of
⬎50% diameter stenosis by visual estimate, reference
From the Divisions of Cardiology and Clinical Biometrics, Brigham and
Women’s Hospital, Boston, Massachusetts; and the Division of
Cardiology, Beth Israel Deaconess Medical Center, Boston,
Massachusetts. Manuscript received August 13, 2004; revised manuscript received and accepted August 31, 2004.
Address for reprints: Donald S. Baim, MD, Brigham and Women’s
Hospital, One Brigham Circle, Boston, Massachusetts 02115. E-mail:
dbaim@partners.org.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.090
173
blocker, longer lesion, angiographic
evidence of thrombus, SVG angulaUnivariable Predictors of 30-day MACE
Odds Ratio
SE
p Value
tion ⬎45°, larger reference vessel diameter, postprocedural in-stent perAge (yrs)
1.00
0.017
0.99
cent diameter stenosis, and treatment
Male gender
1.02
0.439
0.96
History of diabetes mellitus
0.68
0.383
0.31
assignment (no distal protection).
History of myocardial infarction
1.28
0.359
0.49
The 2 new angiographic variables
History of hypertension
0.86
0.376
0.68
derived from the core laboratory,
History of hyperlipidemia
1.79
0.460
0.20
SVG degeneration score and estiSmoking ⱕ1 year
1.41
0.512
0.50
Conventional stenting with predilatation
3.32
0.382
0.002
mated plaque volume, also proved to
Vein graft age
1.03
0.033
0.33
be significant univariable associates
Use of glycoprotein IIb/IIIa inhibitors
1.67
0.230
0.03
of 30-day MACEs (Table 1).
Lesion length
1.06
0.015
0.0001
SVG percent degeneration was
Thrombus
1.89
0.343
0.06
categorized in quartiles based on the
Angulation ⬎45°
3.11
0.501
0.02
Reference vessel diameter
1.72
0.237
0.02
axial proportion (0% to 25%, 26% to
Postprocedural in-stent percent diameter stenosis
1.03
0.010
0.002
50%, 51% to 75%, 76% to 100%),
SVG degeneration score*
1.90
0.168
0.0001
measured by the core laboratory, of
†
Plaque volume
1.01
0.0006
0.0001
the graft that showed angiographic
*Four categories represent percent graft with angiographic evidence of atherosclerotic plaque (0% to
evidence of atherosclerotic disease.
25%, 26% to 50%, 51% to 75%, 76% to 100%).
The number of lesions in each group
†
2
2
Determined by the equation: ␲(lesion length) [(RVD/2) ⫺ (MLD/2) ], where ␲ equals 3.14, RVD is
and the corresponding incidence of
reference vessel diameter in millimeters, and MLD is minimal luminal diameter of the lesion in millimeters.
MACEa are presented in Table 2,
which associates higher degeneration
scores with a higher incidence of
Statistical analysis: Continuous variables were com- MACEs in the control group (p ⫽ 0.0002 for trend by
pared with Student’s t test when variables were nor- chi-square test) and GuardWire group (p ⫽ 0.0001)
mally distributed and with Wilcoxon’s rank-sum test and associates the use of a GuardWire with decreased
when variables were not normally distributed. Binary MACE at all levels of graft degeneration score.
variables were compared with chi-square test with
Estimated lesion plaque volume was calculated for
normal approximation or Fisher’s exact test when each target lesion, and the resulting values were diappropriate. A 2-tailed p value ⱕ0.05 was considered vided into equal quartiles (minimum to maximum
statistically significant. All data were analyzed based plaque volumes within each quartile: 8 to 66, 67 to
on the intention-to-treat principle. Logistic regression 108, 109 to 178, and 179 to 1,109 mm3, respectively).
was used to identify predictors of 30-day MACEs. Number and percentage of MACEs in each quartile of
Selected demographic and angiographic characteris- plaque volume are listed in Table 3. Lesions with
tics recorded in the SAFER trial in addition to graft larger estimated plaque volume were associated with a
degeneration score and lesion plaque volume were higher incidence of MACEs in the control group (p ⫽
considered in univariable analyses as potential predic- 0.0012 for trend by chi-square test) and GuardWire
tors. A candidate list of significant variables (p ⬍0.15) group (p ⫽ 0.0003). As presented in Table 3, use of
from the logistic univariable models was added to the GuardWire was associated with decreased MACEs
those variables of particular clinical interest into a at each level of estimated plaque volume.
stepwise final multivariable model in which a 2-sided
Multivariable determinants of MACEs: Multivariable
p value ⱕ0.05 was considered statistically significant. modeling identified the more conventional variable, leAn interaction was sought between randomized treat- sion length, as collinear with SVG degeneration score
ment assignment (distal occlusion embolic protection and not as strong a predictor; thus, lesion length and
device) and final multivariable main effects. All sta- reference vessel diameter were excluded from the multistical analyses were performed with SAS 6.12 (SAS tivariable model. Therefore, only greater SVG degenerInstitute, Cary, North Carolina).
ation score, larger plaque volume, and lack of GuardWire management were found to be significant
independent determinants of 30-day MACEs (Table 4).
RESULTS
SAFER trial: Between June 1999 and August 2000, Interaction terms between the 2 vein graft independent
801 patients who had 875 lesions were enrolled in the determinants and treatment assignment were then examSAFER trial. The trial showed a 42% relative decrease ined to explore whether the risk conferred by degenerain 30-day MACEs (16.5% for control patients vs 9.6% tion and plaque volume was modified by assignment to
for patients who received the GuardWire, p ⫽ 0.004), the GuardWire distal protection device. Those 2 interacindicating superior safety of stent implantation over tion terms were rejected, suggesting that the main effect
the GuardWire compared with a standard guidewire. of the GuardWire distal protection device is uniformly
The 2 arms of this trial were well matched for all additive across the other 2 risk prediction variables. That
is, MACEs increase with more diffuse SVG degenerabaseline clinical and angiographic variables.
Univariable correlates of MACEs: Factors linked to tion and larger estimated lesion plaque volume in the
an increased risk of MACEs included use of balloon actual analysis groups (Figures 1 and 2) and the model
predilatation, use of a platelet glycoprotein IIb/IIIa predictions (Figure 3), with significant relative decreased
TABLE 1 List of Candidate Variables
174 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
TABLE 2 Adverse Event Rates and Treatment Effect by Saphenous Vein Graft Degeneration Score
Rate of MACE
SVG Degeneration
n
Guidewire
GuardWire
0–25%
26–50%
51–75%
76–100%
435
208
78
45
9.5%
22.9%
32.5%
22.7%
4.7%
12.5%
21.1%
21.7%
OR (95% CI)
0.46
0.48
0.55
0.94
(0.21–1.0)
(0.23–1.0)
(0.20–1.54)
(0.23–3.85)
Relative Decrease
in MACE
p Value*
51%
45%
35%
4%
0.06
0.07
0.31
0.99
Relative Decrease
in MACE
p Value*
49%
77%
21%
32%
0.37
0.01
0.54
0.21
*For trend, p ⫽ 0.0002 and p ⫽ 0.0001.
CI ⫽ confidence interval; OR ⫽ odds ratio.
TABLE 3 Adverse Event Rates and Treatment Effect by Plaque Volume Quartile
30-Day Rate of MACE
Plaque Volume
Quartile
n
Guidewire
GuardWire
1
2
3
4
189
190
189
188
8.4%
13.0%
16.7%
25.3%
4.3%
3.1%
13.1%
17.2%
OR (95% CI)
0.48
0.21
0.76
0.61
(0.14–1.7)
(0.06–0.77)
(0.34–1.69)
(0.30–1.25)
*For trend, p ⫽ 0.0012 and p ⫽ 0.0003.
Abbreviations as in Table 2.
TABLE 4 Multivariable Regression Analysis of Predictors of 30-Day Rates of Main
Adverse Clinical Events
Predictors
Odds Ratio
Treatment assignment (1 ⫽ GuardWire, 0 ⫽
Guidewire)
SVG degeneration Score (1–4)
Plaque volume (continuous)
SE
p Value
0.55
0.23
0.01
1.56
1.002
0.12
0.0007
0.0001
0.0005
FIGURE 1. Percent SVG degeneration was categorized based on
the axial proportion determined by the core laboratory (0% to
25%, 26% to 50%, 51% to 75%, 76% to 100%) of each treated
graft that showed angiographic evidence of atherosclerotic disease. The decrease in relative risk of MACES using the GuardWire versus conventional guidewire is shown for each level of
degeneration. More diffusely diseased SVGs convey a higher
overall risk of MACES, with somewhat less relative benefit of
GuardWire protection.
rates of MACE by the GuardWire, are
seen across levels of degeneration or
plaque volume.
DISCUSSION
The SAFER trial8 showed that the
use of a distal occlusion/aspiration embolic protection device (MedtronicPercuSurge GuardWire and associated
Export catheter) during SVG intervention decreased incidences of 30-day
MACEs (by 42%, from 19% to 9%, p ⬍0.001) and
no-reflow (by 60%, from 8.3% to 3.3%, p ⬍0.001). This
finding implicates distal embolic debris as the proximate
cause of these 2 complications during SVG intervention.
Nonetheless, the clinical community has been slow to
adopt this device due in part to its high cost (⬎$1,000)
and the moderate technical complexity that requires careful coordination of 2 operators to minimize the obligate
4- to 6-minute ischemic occlusion time. One rationalization of that limited adoption might be the assertion that
certain techniques for performing the procedure (e.g., use
of direct stenting, use of platelet glycoprotein IIb/IIIa
receptor blockers) or ways of selecting patients at low
enough risk for embolic complications might obviate
distal protection.
The first variable we tested was percent graft
length that showed evident disease (angiographic
SVG degeneration score; Figure 1). In the control
group, MACEs increased when ⬎25% of the graft
was diseased (25%, 40 of 158) compared with when
⬍25% of the graft was diseased (9.5%, 21 of 220).
However, the relative decrease in MACEs using the
GuardWire was greatest (51%, from 9.5% to 4.7%) in
the group with the lowest SVG score. The observation
of least benefit in the group with the highest SVG
score (4%, from 22.7% to 21.7%) may suggest an
CORONARY ARTERY DISEASE/PREDICTORS OF MACE AFTER SVG PERCUTANEOUS CORONARY INTERVENTION
175
FIGURE 2. Relative risk decrease of MACES when using the
GuardWire versus conventional guidewire for each quartile of
plaque volume. The relative benefit of GuardWire protection is
greatest in the lower quartiles. Larger plaque volumes indicate a
higher overall risk of MACES, which is decreased significantly by
distal protection.
with a diameter equal to the measured minimal luminal diameter. This indirect measurement of plaque
burden proved to be a very potent predictor of complications, with monotonic increases in MACEs in the
conventional guidewire group (from 8.4% to 25.3%, p
⫽ 0.0012 for trend) from lowest to highest quartile of
estimated lesion plaque volume. Moreover, there was
a substantial decrease in MACEs when using the
GuardWire in even the 2 lowest quartiles of estimated
plaque volume (49%, from 8.4% to 4.3%, and 77%,
from 13.0% to 3.0%, in those quartiles, respectively,
p ⫽ 0.01; Figure 2). The somewhat lower treatment
effect in the 2 highest quartiles of estimated lesion
plaque volume (21% and 32%, respectively) may suggest overwhelming embolic load, inefficiency of the
export catheter with bulkier loads, or sufficient plaque
to protrude through the stent struts and embolize when
normal flow is restored after the distal protection
device has been removed.
This additional analysis of the 801-patient SAFER
trial thus shows that even the lowest risk grafts have
nontrivial (4% to 11%) event rates without distal
protection and that those risks are decreased across the
board using distal embolic protection. Over and above
these findings, the present study has
identified important new angiographic parameters that can be used
to compare expected baseline rates
of MACEs of patient cohorts across
various device trials, including distal
filters and proximal occlusion aspiration devices.
APPENDIX A
FIGURE 3. Predicted rates of MACES between control and treatment (GuardWire)
groups based on 2 predictor variables (plaque volume and percent SVG degeneration) in this analysis calculated with the equation presented in Appendix B. Use of
distal embolic protection provides treatment benefit across all levels of risk.
overwhelming embolic load or offsetting device-produced emboli when there is no relatively disease-free
landing zone for the GuardWire.
Beyond the state of degeneration of the graft as a
whole, we hypothesized that the amount of debris
available for embolization from the lesion site itself
might also be a predictor of potential embolic load and
consequent MACEs. Without intravascular ultrasound, no actual measurement of lesion volume could
be ascertained. However, we chose to calculate an
estimated plaque volume as the volume of a cylinder
whose diameter was equal to the vessel reference
diameter and whose length was equal to the lesion
length minus the volume of a cylinder of equal length
176 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
Angiographic variable definitions: All procedural angiograms were reviewed according to standard
morphologic criteria. Lesion length was defined as the
axial extent of the lesion that contained a shoulder-toshoulder luminal decrease by ⱖ20%. The degree of
graft degeneration (⬎20% luminal irregularity)
throughout the SVG was estimated as percent total
SVG length and reported in quartiles as 0% to 25%,
26% to 50%, 51% to 75%, and 75% to 100%.
Estimated lesion plaque volume was defined as the
volume of a cylinder whose diameter was equal to the
reference vessel diameter and whose length was the shoulder-to-shoulder lesion length as measured by the core
laboratory. Plaque volume was determined by this equation: ␲(lesion length)[(RVD/2)2 ⫺ (MLD/2)2], where ␲
equals 3.14, RVD is the reference vessel diameter in
millimeters, and MLD is the minimal luminal diameter of
the lesion in millimeters. The resultant patient lesion volumes were divided into 4 quartiles that contained equal
numbers of patients.
APPENDIX B
Model-predicted rates of MACEs: The probability of MACEs based on the
final multivariable model was calculated for each quartile of SVG degeneration
score and plaque volume. For SVG degeneration score, a number from 1 to 4 was
entered based on percent degeneration as previously defined. Plaque volume was
derived as a continuous variable and used as such for multivariable modeling. For
purposes of establishing predicted rates of MACE, plaque volume was categorized by quartiles, and the mean value within each quartile was used in the
following formula:
Probability of MACE
e(⫺2.44⫺0.59关treatment assignment兴⫹0.45关Degeneration Score兴⫹0.0024关plaque volume兴)
.
⫽
1 ⫹ e(⫺2.44⫺0.59关treatment assignment兴⫹0.45关Degeneration Score兴⫹0.0024关plaque volume兴)
JANUARY 15, 2005
1. Grube E, Gerckens U, Yeung AC, Rowold S, Kirchhof N, Sedgewick J, Yadav
JS, Stertzer S. Prevention of distal embolization during coronary angioplasty in
saphenous vein grafts and native vessels using porous filter protection. Circulation 2001;104:2436 –2441.
2. Webb JG, Carere RG, Virmani R, Baim D, Teirstein PS, Whitlow P, McQueen
C, Kolodgie FD, Buller E, Dodek A, Mancini GB, Oesterle S. Retrieval and
analysis of particulate debris after saphenous vein graft intervention. J Am Coll
Cardiol 1999;34:468 – 475.
3. Lefkovits J, Holmes DR, Califf RM, Safian RD, Pieper K, Keeler G, Topol EJ.
Predictors and sequelae of distal embolization during saphenous vein graft
intervention from the CAVEAT-II trial. Coronary Angioplasty Versus Excisional
Atherectomy Trial. Circulation 1995;92:734 –740.
4. Waksman R, Douglas JS Jr, Scott NA, Ghazzal ZM, Yee-Peterson J, King SB
III. Distal embolization is common after directional atherectomy in coronary
arteries and saphenous vein grafts. Am Heart J 1995;129:430 – 435.
5. Waksman R, Weintraub WS, Ghazzal Z, Scott NA, Shen Y, King SB III,
Douglas JS Jr. Short- and long-term outcome of narrowed saphenous vein bypass
graft: a comparison of Palmaz-Schatz stent, directional coronary atherectomy,
and balloon angioplasty. Am Heart J 1997;134:274 –281.
6. Hong MK, Mehran R, Dangas G, Mintz GS, Lansky AJ, Pichard AD, Kent
KM, Satler LF, Stone GW, Leon MB. Creatine kinase-MB enzyme elevation
following successful saphenous vein graft intervention is associated with late
mortality. Circulation 1999;100:2400 –2405.
7. Waksman R, Ghazzal ZM, Baim DS, Steenkiste AR, Yeh W, Detre KM, King
SB III. Myocardial infarction as a complication of new interventional devices.
Am J Cardiol 1996;78:751–756.
8. Baim DS, Wahr D, George B, Leon MB, Greenberg J, Cutlip DE, Kaya U,
Popma JJ, Ho KK, Kuntz RE. Randomized trial of a distal embolic protection
device during percutaneous intervention of saphenous vein aorto-coronary bypass
grafts. Circulation 2002;105:1285–1290.
9. Stone GW, Rogers C, Hermiller J, Feldman R, Hall P, Haber R, Masud A,
Cambier P, Caputo RP, Turco M, et al. Randomized comparison of distal
protection with a filter-based catheter and a balloon occlusion and aspiration
system during percutaneous intervention of diseased saphenous vein aorto-coronary bypass grafts. Circulation 2003;108:548 –553.
CORONARY ARTERY DISEASE/PREDICTORS OF MACE AFTER SVG PERCUTANEOUS CORONARY INTERVENTION
177
Artificial Neural Network Modeling of
Stress Single-Photon Emission Computed
Tomographic Imaging for Detecting
Extensive Coronary Artery Disease
J. Scott Allison,
MD,
Jaekyeong Heo,
MD,
and Ami E. Iskandrian,
MD
Single-photon emission computed tomographic imaging
is a useful, noninvasive method to detect coronary artery
disease (CAD). We tested the hypothesis that artificial
neural network modeling could predict CAD extent better than visual interpretation; 109 patients who underwent stress single-photon emission computed tomography and coronary angiography were selected. Twenty
patients who had a <5% probability of CAD were also
selected for calculation of normalcy rate. A model was
trained for each vessel. Stress images were decreased to
25 points by pixel averaging the polar map. The model
output was 1 for vessel stenosis >60% and 0 otherwise.
Model sensitivities were 92% (55 of 60) for left anterior
descending artery versus 62% (37 of 60) for visual
interpretation (p ⴝ 0.0002), 69% (20 of 29) for left
circumflex artery versus 55% for visual interpretation (p
ⴝ 0.30), and 94% (45 of 48) for right coronary artery
versus 78% for visual interpretation (p ⴝ 0.024). Model
specificities and normalcy rates were 78% and 85% for
the left anterior descending artery, 93% and 100% for
the left circumflex artery, and 85% and 90% for the right
circumflex artery, respectively. Single-vessel CAD was
predicted in 27 of 28 patients (96%) by modeling versus
23 of 28 patients (82%) by visual interpretation (p ⴝ 0.11).
Multivessel CAD was correctly predicted in 30 of 46 patients (65%) by modeling versus 16 of 46 patients (35%) by
visual interpretation (p ⴝ 0.004). Thus, artificial neural
network models can predict CAD from stress single-photon
emission computed tomographic images when using separate models for the 3 major epicardial vessels. Because of
their high sensitivity and specificity in detecting extensive
CAD, these models have great promise as an aid to correctly identify patients at high risk for CAD. 䊚2005 by
Excerpta Medica Inc.
(Am J Cardiol 2005;95:178 –181)
ecause of the powerful ability to model complex
input/output relations, nonlinear artificial neural netB
work models can be used to map single-photon emission
raphy separated by ⱕ3 months between January 2001
and April 2003 were selected. Patients who had previous coronary revascularization or any event between
studies were excluded. Based on ⱖ60% diameter stenosis by coronary angiography, 17 patients had 3-vessel CAD, 29 had 2-vessel CAD, 28 had 1-vessel CAD,
and 35 had no significant CAD. Patients who had left
main disease (n ⫽ 8) were considered to have 2-vessel
disease (LAD and left circumflex). Pertinent data on
the study group are presented in Table 1. Twenty
patients who had a ⬍5% probability of CAD and
normal stress SPECT studies by visual interpretation
were selected to calculate a normalcy rate. These 20
patients did not undergo coronary angiography.
Stress and imaging procedure: Patients underwent a
stress treadmill exercise test according to Bruce’s
protocol (n ⫽ 37) or received adenosine (n ⫽ 67) or
dobutamine (n ⫽ 5). Patients were instructed to not
use ␤ blockers and long-acting nitrates for 24 to 48
hours before the study. Methods of stress testing,
image acquisition and processing, and image interpretation have previously been described.6 All images
were acquired in a gated mode 60 minutes after injection of technetium 99m tetrofosmin with a dual-head
detector. Although attenuation correction was used
when available for visual interpretation, only data
from the uncorrected images were used for training
and validating the neural networks. Same-day or separate-day protocols during stress and at rest were used
as appropriate based on patients’ weight and body
computed tomographic (SPECT) perfusion images to
coronary narrowings detected by coronary angiography.
Previous studies have indicated that these models detect
coronary artery disease (CAD) at least as effectively as
human experts.1–5 However, these studies combined
right and left circumflex coronary arteries into a single
vascular territory, produced models from images during
stress and at rest with only a small number of inputs from
the stress data, or used models with a large number of
inputs and a small number of training cases. This study
tested the hypothesis that neural network models could
predict extensive CAD by using separate models for the
right, left circumflex, and left anterior descending (LAD)
arteries better than visual interpretation of SPECT images.
METHODS
Study patients: One hundred nine patients who had
stress SPECT perfusion studies and coronary angiog-
From the Division of Cardiovascular Diseases, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama.
Manuscript received June 24, 2004; revised manuscript received and
accepted September 3, 2004.
Address for reprints: Ami E. Iskandrian, MD, Division of Cardiovascular Diseases, University of Alabama at Birmingham, 318 LHRB,
1900 University Boulevard, Birmingham, Alabama 35294-0006.
E-mail: aiskand@uab.edu.
178
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.003
averaging the polar map with 25 equal-area sectors, as
shown in Figure 1. Before averaging, pixels were
scaled from 0 to 1, with 1 indicating the most perfuAge (yrs)
62 ⫾ 11
sion and 0 indicating the least. These 25 data points
Men
64 (59%)
were used as inputs to the model. Because this study did
Systemic hypertension
83 (76%)
not attempt to identify ischemia, perfusion data at rest
Diabetes mellitus
38 (35%)
Hyperlipidemia (LDL ⬎100 or on therapy)
40 (37%)
were not included in the model inputs. The model output
Previous myocardial Infarction
13 (12%)
was 1 for vessel stenosis ⬎60% and 0 otherwise.
Ejection fraction (%)
52 ⫾ 14
Model creation: Separate models were created and
Chest pain during stress
18 (17%)
trained for each of the 3 vessels. The model architecPositive electrocardiogram during stress
30 (28%)
ture (Figure 2) used the logistic function, 1.0/[1.0 ⫹
LDL ⫽ low-density lipoprotein.
e(⫺x)], as the transfer function, f. The same inputs
were used for all models as previously described, with the output
based on the existence of a significant lesion in the vessel modeled.
The model’s input layer contained
25 neurons that corresponded to the
25 inputs and 1 output layer neuron.
The number of hidden layer neurons
was optimized for each model, with
the LAD model containing 8, the left
circumflex model containing 4, and
the right model containing 5.
EasyNN Plus 3.0 (Neural Planner
Software, Cheadle, United Kingdom) was used for model creation,
training, and validation. Each model
was trained with back-propagation.
Data from the 109 patients were split
randomly into 66% training (72 of
FIGURE 1. Stress polar map divided into 25 equal-area sectors.
109) and 33% validation (37 of 109)
datasets for each model. More patients were assigned to the training
data set than to the validation data
set because of the relatively small
number of patients.7
Statistical analysis: Data are presentedby model (vessel) and by extent
of disease (single vessel or multivessel). Model results were compared
with visual interpretation. Model predictions by vessel and extent of disease were compared with visual interFIGURE 2. Multilayer perception neural network. f ⴝ transfer function; w ⴝ synaptic
pretation by Fisher’s exact test. A p
weight.
value ⬍0.05 was considered statistically significant.
TABLE 1 Pertinent Demographic Data on Study Population (n
⫽ 109)
habitus. Perfusion defects in the anterior wall or septum were considered to represent LAD disease, those
in the inferior wall were considered to represent right
coronary artery disease, and those in the lateral wall
were considered to represent left circumflex disease.
Polar maps were generated with AutoQUANT 4.2.1
(ADAC, Sunnyvale, California).
Coronary angiography was performed according to
standard techniques. Significant stenosis was defined as
ⱖ60% diameter stenosis in the LAD, left circumflex, or
right coronary arteries. Angiograms and the SPECT images were read without knowledge of the other results.
Model inputs and output: Stress SPECT perfusion
images were decreased to 25 data points by pixel
RESULTS
There were 60 diseased LADs, 48 diseased right,
and 29 diseased left circumflex coronary arteries. Table 2 lists the sensitivity of the model for detecting
CAD compared with visual interpretation. These results are presented graphically in Figures 3 and 4.
In patients who had no significant CAD, all 3
models had correct predictions in 21 of 35 patients,
with a specificity of 60% (visual interpretation 29%, p
⫽ 0.0095). The low specificity most likely reflected
post-test referral bias, and most patients had CAD but
⬍60% stenosis, which could be physiologically significant. Results in the training and validation groups
were similar (data not shown).
CORONARY ARTERY DISEASE/ARTIFICIAL NEURAL NETWORK MODELING
179
The most common method of
SPECT interpretation is visual, which
often is supplemented by semiquantiModel
Visual
p Value
tative algorithms. These algorithms include the use of a summed stress score,
Location of disease
LAD
92% (55/60)
62% (37/60)
0.0002
polar maps, or circumferential proLeft circumflex
69% (20/29)
55% (16/29)
0.2964
files.8 –13 The use of gated imaging and
Right
94% (45/48)
78% (37/48)
0.0237
attenuation
correction have improved
No. of coronary artery narrowings
diagnostic accuracy.14 –20 It is for these
1
96% (27/28)
82% (23/28)
0.1080
2
59% (17/29)
31% (9/29)
0.0402
reasons that SPECT imaging has been
3
76% (13/17)
41% (7/17)
0.0466
firmly established in detection of CAD
2/3
65% (30/46)
35% (16/46)
0.0040
and risk assessment. There are 2 issues
that need further refinement, the ability
to correctly identify patients at high
risk (those with extensive CAD) and to improve the
interpretation of less experienced readers.
Multiple methods of automated interpretation of
SPECT imaging have previously been studied. Most
of these studies have used the polar map as input.
Perfusion quantification applied to assigned vessel
regions has been used to localize fixed and reversible
defects to a vessel.8 –13 Expert systems have been
developed that use a knowledge base of rules to interpret images.21 Case-based systems match a study
patient to the best fit in a library of patients.22 These
methods have the advantage that some justification
can be made for the results produced. Garcia et al21
FIGURE 3. Sensitivity of the model (gray bars) versus that of vifound that interpretations of an expert system agree
sual interpretation (black bars) for detecting CAD.
well with expert visual interpretation and are as accurate when coronary angiography is used as the gold
standard. Similarly, Khorsand et al22 showed that a
case-based system has a diagnostic accuracy similar to
visual interpretation and the use of normal limits.
Neural networks are true “black box” models that
contain complex interconnections of neurons with no
interpretable intermediate data. However, because of
their ability to perform complex pattern-recognition
tasks, neural networks have the potential to outperform
the previously described methods. Multiple previous
studies have examined neural network models of myocardial perfusion images. Fujita et al1 showed that a
neural network with 256 inputs could be trained to interpret SPECT polar map images better than a radiology
FIGURE 4. Sensitivity of the model (gray bars) versus that of visual interpretation (black bars) for detecting extent of CAD.
resident but worse than an expert radiologist. However,
the study used a very large number of inputs, 256, with
Twenty patients who had a ⬍5% likelihood of CAD a small number of test cases, 71. The study also did not
were used to determine normalcy rates. The LAD model contain any cases in which the angiographic results did
correctly predicted 17 of 20 patients (85%), the left not agree with the visual interpretation of the SPECT
circumflex model correctly predicted 20 of 20 patients images. Porenta et al2 reported that a neural network with
(100%), and the right model correctly predicted 18 of 20 45 inputs from thallium-201 dipyridamole stress-redispatients (90%). The combined normalcy rate was 75% tribution planar images predicted significant CAD with a
(15 of 20) for all 3 models.
sensitivity of 51% and a specificity of 90% compared
with angiography. The study examined 159 patients, 81
of whom underwent angiography, and considered 2 vasDISCUSSION
The present results show that fully automated neural cular territories (LAD and combined right/left circumflex
network models can be used to detect any CAD and arteries). The neural network was not significantly better
extensive CAD with a high sensitivity and specificity than expert visual interpretation compared with angiogthat are as good as or better than visual interpretation raphy. A neural network trained to predict expert visual
alone. Sensitivities were 96% for single-vessel disease interpretation performed better than the neural network
and 65% for correctly identifying multivessel disease. trained to predict angiographic lesions. Hamilton et al3
These values were higher than those by visual interpre- examined 410 men and found that a neural network
tation (82% and 35%, respectively).
could be trained to detect hypoperfused segments in a
TABLE 2 Model Sensitivity Versus Visual Interpretation in Detecting Coronary
Artery Disease
180 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
polar map divided into 8 or 24 segments. Hypoperfused
segments were defined as ⬍2.5 SD of the expected mean
count from a normal database. Coronary angiography
was not used in this study. Lindahl et al4 reported that a
neural network could be trained to predict angiographic
LAD lesions significantly better than visual interpretation (77.2% vs 73% sensitivity, p ⫽ 0.038). The study
examined 135 patients and considered 2 vascular territories (LAD and combined right/left circumflex arteries).
The neural network trained for the right/left circumflex
territory was not significantly better than visual interpretation. Inputs to the 2 neural networks were 30 Fourier
components, which were obtained by applying a 2-dimensional Fourier transform to the polar maps during
stress and at rest, which has a smaller root-mean-square
error than pixel averaging. However, neural networks
were not trained with both methods. Lindahl et al,5 in a
separate study, found that physicians who classify polar
maps benefit from an artificial neural network as measured by an increased area under the receiver-operating
characteristic curve.5 The areas under the receiver-operating characteristic curve increased from 0.65 to 0.70 (p
⫽ 0.18) and from 0.79 to 0.82 (p ⫽ 0.006) for the 2
vascular territories in the same 135 patients assessed in
the previous study. Based on the results of this and other
studies, neural network models appear to be a promising
method to improve diagnostic accuracy of SPECT imaging.
This study has a few limitations that must be considered. First, coronary angiography was used as the gold
standard for CAD, which may not be valid, especially in
patients who have intermediate severity of stenoses.23–25
Further, no independent method to validate the physiologic significance of coronary stenosis such as flow reserve ratio, pressure gradients, or resistance was used.
These methods have shown better correlation with visual
interpretation of SPECT images than with coronary angiography.24 –27 We elected to use 60% diameter stenosis
as a cutoff but even that is open to criticism. Second, all
data obtained were retrospective with no prospective
testing. Third, the sample size of 109 patients was relatively small for creation of a neural network, but the
cases were diverse. Diverse cases are more desirable
than number of patients because a neural network trained
on data from many patients who do not have CAD will
not perform well in patients who do have CAD. Further
work needed in this area includes modeling left main
disease and prospective testing. A model to improve the
detection of significant left main disease would be valuable. Because of results from these 3 models, a left main
model with good performance should be possible with
enough training data. Prospective testing that measures
improvement in visual interpretations with model advice
is needed.
1. Fujita H, Katafuchi T, Uehara T, Nishimura T. Application of artificial neural
network to computer-aided diagnosis of coronary artery disease in myocardial
SPECT bull’s eye images. J Nucl Med 1992;33:272–276.
2. Porenta G, Dorffner G, Kundrat S, Petta P, Duit-Schedlmayer J, Sochor H.
Automated interpretation of planar thallium-201-dipyridamole stress-redistribution scintigrams using artificial neural networks. J Nucl Med 1994;35:2041–2047.
3. Hamilton D, Riley PJ, Miola UJ, Amro AA. A feed forward neural network for
classification of bull’s-eye myocardial perfusion images. Eur J Nucl Med 1995;
22:108 –115.
4. Lindahl D, Palmer J, Ohlsson M, Peterson C, Lundin A, Edenbrandt L.
Automated interpretation of myocardial SPECT perfusion images using artificial
neural networks. J Nucl Med 1997;38:1870 –1875.
5. Lindahl D, Lanke J, Lundin A, Palmer J, Edenbrandt L. Improved classifications of myocardial bull’s eye scintigrams with computer-based decision support
system. J Nucl Med 1999;40:96 –101.
6. Iskandrian AE, Verani MS. Nuclear Cardiac Imaging: Principles and Applications, 3rd Ed. New York: Oxford University Press, 2003:137–162.
7. Patterson DW. Artificial Neural Networks: Theory and Applications. Singapore:
Prentice-Hall, 1996:213.
8. Garcia EV, Van Train K, Maddahi J, Prigent F, Friedman J, Areeda J, Waxman
A, Berman DS. Quantification of rotational thallium-201 myocardial tomography.
J Nucl Med 1985;26:17–26.
9. DePasquale EE, Nody AC, DePuey EG, Garcia EV, Pilcher G, Bredlau C,
Roubin G, Gober A, Gruentzig A, D’Amato P, Berger HJ. Quantitative rotational
thallium-201 tomography for identifying and localizing coronary artery disease.
Circulation 1988;77:316 –327.
10. Maddahi J, Van Train K, Prigent F, Garcia EV, Friedman J, Ostrzega E, Berman
D. Quantitative single photon emission computed thallium-201 tomography for
detection and localization of coronary artery disease: optimization and prospective
validation of a new technique. J Am Coll Cardiol 1989;14:1689 –1699.
11. Klein JL, Garcia EV, DePuey EG, Campbell J, Taylor AT, Pettigrew RI,
D’Amato P, Folks R, Alazraki N. Reversibility bull’s-eye: a new polar bull’s eye
map to quantify reversibility of stress-induced SPECT thallium-201 myocardial
perfusion defects. J Nucl Med 1990;31:1240 –1246.
12. Garcia EV, DePuey EG, Sonnemaker RE, Neely HR, DePasquale EE,
Robbins WL, Moore WH, Heo J, Iskandrian AS, Campbell J. Quantification of
the reversibility of stress-induced thallium-201 myocardial perfusion defects: a
multicenter trial using bull’s-eye polar maps and standard normal limits. J Nucl
Med 1990;31:1761–1765.
13. Van Train KF, Areeda J, Garcia EV, Cooke CD, Maddahi J, Kiat H, Germano
G, Silagan G, Folks R, Berman DS. Quantitative same-day rest-stress technetium99m-sestamibi SPECT: definition and validation of stress normal limits and
criteria for abnormality. J Nucl Med 1993;34:1494 –1502.
14. Chua T, Kiat H, Germano G, Maurer G, Van Train K, Friedman J, Berman
D. Gated technetium 99m sestamibi for simultaneous assessment of stress myocardial perfusion, post-exercise regional ventricular function and myocardial
viability. J Am Coll Cardiol 1994;23:1107–1114.
15. DePuey EG, Rozanski A. Using gated technetium-99m-sestamibi SPECT to
characterize fixed myocardial defects as infarct or artifact. J Nucl Med 1995;36:
952–955.
16. Smanio PE, Watson DD, Segalla DL, Vinson EL, Smith WH, Beller GA.
Value of gating of technetium-99m sestamibi single-photon emission computed
tomographic imaging. J Am Coll Cardiol 1997;30:1687–1692.
17. Taillefer R, DePuey EG, Udelson JE, Beller GA, Latour Y, Reeves F.
Comparative diagnostic accuracy of Tl-201 and Tc-99m sestamibi SPECT imaging (perfusion and ECG-gated SPECT) in detecting coronary artery disease in
women. J Am Coll Cardiol 1997;29:69 –77.
18. Sharir T, Germano G, Kavanagh PB, Lai S, Cohen I, Lewin HC, Friedman
JD, Zellweger MJ, Berman DS. Incremental prognostic value of post-stress left
ventricular ejection fraction and volume by gated myocardial perfusion single
photon emission computed tomography. Circulation 1999;100:1035–1042.
19. DePuey EG, Garcia EV. Optimal specificity of thallium-201 SPECT through
recognition of imaging artifacts. J Nucl Med 1989;30:441– 449.
20. Links J, DePuey EG, Taillefer R, Becker LC. Attenuation correction and
gating synergistically improve the diagnostic accuracy of myocardial perfusion
SPECT. J Nucl Cardiol 2002;9:183–187.
21. Garcia EV, Cooke CD, Folks RD, Santana CA, Krawczynska EG, De Braal
L, Ezquerra NF. Diagnostic performance of an expert system for the interpretation of myocardial perfusion SPECT studies. J Nucl Med 2001;42:1185–1191.
22. Khorsand A, Haddad M, Graf S, Moertl D, Sochor H, Porenta G. Automated
assessment of dipyridamole 201Tl myocardial SPECT perfusion scintigraphy by
case-based reasoning. J Nucl Med 2001;42:189 –193.
23. White CW, Wright CB, Doty DB, Hiratza LF, Eastham CL, Harrison DG,
Marcus ML. Does visual interpretation of the coronary arteriogram predict the
physiologic importance of a coronary stenosis? N Engl J Med 1984;310:819 – 824.
24. Miller DD, Donohue TJ, Younis LT, Bach RG, Aguirre FV, Wittry MD,
Goodgold HM, Chaitman BR, Kern MJ. Correlation of pharmacologic 99mTcsestamibi myocardial perfusion imaging with poststenotic coronary flow reserve in
patients with angiographically intermediate coronary artery stenoses. Circulation
1994;89:2150 –2160.
25. Pijls NH, De Bruyne B, Peels K, Van Der Voort PH, Bonnier HJ, Bartunek J,
Koolen JJ. Measurement of myocardial fractional flow reserve to assess the functional severity of coronary artery stenosis. N Engl J Med 1996;334:1703–1708.
26. Donohue TJ, Kern MJ, Aguirre FV, Bach RG, Wolford T, Bell CA, Segal J.
Assessing the hemodynamic significance of coronary artery stenoses: analysis of
translesional pressure-flow velocity relationships in patients. J Am Coll Cardiol
1993;22:449 – 458.
27. Heller LI, Cates C, Popma J, Deckelbaum LI, Joye JD, Dahlberg ST, Villegas
BJ, Arnold A, Kipperman R, Grinstead WC, et al. Intracoronary Doppler assessment of moderate coronary artery disease: comparison with 201Tl imaging and
coronary angiography. FACTS Study Group. Circulation 1997;96:484 – 490.
CORONARY ARTERY DISEASE/ARTIFICIAL NEURAL NETWORK MODELING
181
Incremental Prognostic Power of
Single-Photon Emission Computed
Tomographic Myocardial Perfusion
Imaging in Patients With Known or
Suspected Coronary Artery Disease
Salvador Borges-Neto, MD, Linda K. Shaw, MS, Robert H. Tuttle, MSPH,
John H. Alexander, MD, William T. Smith IV, MD, Marianna Chambless, BS,
R. Edward Coleman, MD, Robert A. Harrington, MD, and Robert M. Califf, MD
Noninvasive stress testing provides prognostic information in patients who have suspected coronary artery
disease, but limited data are available on the incremental value of myocardial perfusion testing in high-risk
patients. We studied 3,275 patients who underwent
cardiac catheterization and single-photon emission computed tomographic (SPECT) perfusion imaging. Median
follow-up was 3.1 years for death, cardiovascular
death, and a composite of cardiovascular death or nonfatal myocardial infarction. Using Cox’s proportional
hazards regression models, we examined the relation of
SPECT summed stress score (SSS) to each outcome. A
1-unit change in SSS was associated with increased risks
of 4%, 7%, and 5% for death, cardiovascular death, and
death or nonfatal myocardial infarction, respectively (all
p <0.0001). To examine the prognostic utility of SPECT,
after baseline adjustments, SSS and angiographic results provided incremental prognostic information for
each outcome. Thus, SPECT SSS provides information
beyond clinical and angiographic data in patients who
have known or suspected coronary artery disease. This
information may be useful for stratifying patients into
multiple risk categories for future cardiovascular events
and potentially guiding therapy. 䊚2005 by Excerpta
Medica Inc.
(Am J Cardiol 2005;95:182–188)
e undertook this study (1) to evaluate the value
of single-photon emission computed tomoW
graphic (SPECT) myocardial perfusion imaging for
Center (Durham, North Carolina) specifically for percutaneous coronary intervention often have incomplete angiographic data in the Duke Databank for
Cardiovascular Disease. Ten percent of otherwise eligible patients had incomplete angiograms. The study
population consisted of 3,275 patients who had complete angiographic and SPECT myocardial perfusion
imaging studies. This research was performed with the
approval of the institutional review board of the Duke
University Medical Center.
Clinical information: Clinical characteristics were
prospectively collected for each patient at the time of
cardiac catheterization and stored in the Duke Databank for Cardiovascular Disease.1,2 Clinical variables
included age; history of congestive heart failure, peripheral vascular disease, cerebrovascular disease, hypertension, diabetes, or myocardial infarction; New
York Heart Association congestive heart failure classification; presence of carotid bruits or ventricular
gallop; chest pain (pain type, course, and frequency);
gender; race; and a modified Charlson’s co-morbidity
index.3 Histories of congestive heart failure and myocardial infarction were removed from Charlson’s index and examined independently because of their predictive value in high-risk patients who have cardiac
problems.
Stress testing: Patients who were capable of exercising underwent exercise treadmill stress testing according to a standard Bruce’s protocol, unless another
protocol was specifically requested by the attending
predicting adverse events in a high-risk patient population for coronary artery disease (CAD) and (2) to
determine the incremental benefit of SPECT data versus those of clinical and angiographic data. With the
increasing number of options available for medical
and interventional management of CAD, risk stratification of patients who have CAD may help to guide
and monitor responses to therapy.
METHODS
Study population: Between September 1993 and
July 2000, 17,048 patients underwent SPECT procedures during stress (exercise or pharmacologic) or at
rest at our laboratory. Of these patients, 4,829 underwent coronary angiography ⱕ180 days before or after
the nuclear procedure. Patients were excluded from
the study if they underwent subsequent revascularization ⱕ60 days after SPECT perfusion imaging and if
the angiographic results were incomplete. Patients
who are referred to the Duke University Medical
From the Departments of Medicine (Cardiology) and Radiology (Nuclear
Medicine) and the Duke Clinical Research Institute, Duke University
Medical Center, Durham, North Carolina. Manuscript received March
11, 2004; revised manuscript received and accepted September 8,
2004.
Address for reprints: Salvador Borges-Neto, MD, Duke University
Medical Center, Box 3949, Durham, North Carolina 27710. E-mail:
borge001@mc.duke.edu.
182
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.004
FIGURE 1. Twelve-segment system used to generate the perfusion
score for nuclear imaging at rest and during stress: (1) high anterior, (2) low anterior, (3) high anterolateral, (4) low anterolateral,
(5) anteroseptal, (6) inferoseptal, (7) high posterolateral, (8) low
posterolateral, (9) inferior, (10) posterobasal, (11) inferoapical, and
(12) anteroapical.
physician.4 Patients who were unable to exercise underwent pharmacologic stress testing.
SPECT myocardial perfusion imaging protocol: The
protocol for performing SPECT myocardial perfusion
imaging studies has been previously described.5,6 In
summary, SPECT data were obtained with multihead
detectors using a step-and-shoot protocol. Images at
rest were obtained for 30 seconds/projection and those
during stress were obtained for 20 seconds/projection.
A dual-isotope protocol was used for most patients,
with thallium-201 being the agent of choice for images at rest and technetium-99m perfusion agents being used for images during stress. Patients who
weighed ⬎280 lb (⬎127 kg) had technetium-99m
agents used for portions of the examination at rest and
during stress.
Image interpretation and candidate nuclear variables: Studies were independently reviewed by physi-
cian specialists in nuclear medicine or nuclear cardiology. A 12-segment reporting system (Figure 1) was
used to quantify perfusion in various segments. The
relative perfusion grade to each segment was quantified by 4 gradations, with 0 representing no defect, 1
representing a mild defect, 2 representing a moderate
defect, and 3 representing a severe defect. A cumulative summed stress score (SSS) was obtained by adding the scores of the 12 segments. Therefore, SSS
would equal 0 in a normal study, and the maximum
score would be 36 (severe perfusion defect in all 12
segments). The score variable has been shown to be
highly predictive of cardiovascular outcomes in a 20segment model.7 The summed rest score, which is the
perfusion score at rest, and the summed difference
score, the score difference between stress and rest,
were also constructed by applying the same scoring
system.
Coronary angiography: Coronary angiography was
performed in multiple left and right anterior oblique
projections. Angiograms were interpreted by consensus of ⱖ2 experienced angiographers who applied the
following ordinal scale: 0, ⬍25%, 25%, 50%, 75%,
95%, or 100% occlusion. Extent of CAD was determined by the traditional 1-, 2-, or 3-vessel disease
characterization. Significant disease was defined as
ⱖ75% occlusion of a major epicardial coronary
artery.
The number of major epicardial vessels with significant stenosis (ⱖ75% narrowing of luminal diameter) was used to evaluate the prognostic value of
angiographic disease severity with SPECT.
Follow-up: During follow-up, information was collected on death, cardiovascular death, nonfatal myocardial infarction, and date of last known status. These
data were prospectively obtained by mailed questionnaires or telephone interviews at 6 months, 1 year, and
annually thereafter. Information from clinic visits and
rehospitalizations was also evaluated to ascertain
follow-up and end point determination. Considering
all time intervals and all patients, follow-up was 93%
complete. An independent clinical events committee
reviewed and classified all deaths and nonfatal myocardial infarctions without knowledge of the patient’s
clinical, cardiac catheterization, or SPECT myocardial
perfusion results. Criteria used to diagnose nonfatal
myocardial infarction and cardiovascular death have
been described and validated.8,9
Statistical analysis: Clinical characteristics of subjects were described in terms of percentage for categorical variables and by the median (25th and 75th
percentiles) for continuous variables. Significant differences at the 0.05 level across risk strata were determined with Pearson’s chi-square tests for categorical variables and the Kruskal-Wallis test for continuous
variables. Cumulative survival rates as a function of time
after SPECT myocardial perfusion imaging were calculated using the Kaplan-Meier method.10 These rates are
reported in tabular form at specific time points stratified
by SSS score category and by number of diseased
vessels.
Cox’s proportional hazards regression modeling
techniques were used to assess unadjusted and adjusted relations between SSS and patient outcomes.11,12 After examining the results of a flexible
Cox’s model-fitting approach involving cubic polynomial spline functions,13–17 the linearity of the unadjusted relation between SSS and each outcome was
assessed. Further application of Cox’s regression techniques resulted in multivariable clinical models that
were developed to determine the most important predictors of the outcomes based on patients’ histories
and physical examinations. Candidate variables were
selected on the basis of clinical importance and from
previous models.1,18 –20 Multivariable models for each
end point that consisted of variables that were statistically significant (p ⬍0.05) were derived from a stepwise selection process of the candidate variables: age,
gender, race, history of hypertension, vascular disease,
Charlson’s co-morbidity index, history and severity of
CORONARY ARTERY DISEASE/PROGNOSTIC POWER OF SPECT
183
TABLE 1 Clinical Characteristics of Study Population
Clinical Characteristics
Men
White
Age (yrs)
Systemic hypertension
Previous myocardial infarction
Previous congestive heart failure
Diabetes mellitus
Peripheral vascular disease
Previous revascularization
SRS
SDS
Angiographic results
No. of diseased vessels
0
1
2
3
SSS 0 – 4
(n ⫽ 1,669)
SSS 5–14
(n ⫽ 1,297)
SSS ⱖ15
(n ⫽ 309)
p Value
54.2%
77.7%
61 (52, 70)
65.6%
26.8%
28.1%
26.1%
14.9%
42.5%
0 (0, 0)
0 (0, 2)
71.3%
77.2%
61 (52, 69)
64.6%
58.9%
32.1%
31.2%
17.7%
58.4%
5 (2, 8)
2 (0, 5)
78.6%
78.3%
62 (54, 71)
65.1%
78.3%
50.0%
37.9%
18.8%
66.0%
15 (11, 18)
2 (0, 6)
0.001
0.910
0.208
0.851
0.001
0.001
0.001
0.054
0.001
0.001
0.001
42.9%
26.2%
14.7%
16.2%
16.7%
23.3%
24.8%
35.2%
7.4%
17.2%
21.0%
54.4%
0.001
Values are percentages or medians (25th, 75th percentiles).
SDS ⫽ summed difference score; SRS ⫽ summed rest score.
congestive heart failure, ventricular gallop, diabetes,
previous myocardial infarction, and previous revascularization with percutaneous coronary intervention or
coronary artery bypass grafting. After determining the
most robust model for each end point, angiographic
results (defined by number of diseased vessels) and
perfusion SSS were added to the model in a stepwise
manner to assess their incremental benefit. An interaction term consisting of SSS by angiographic results
was tested in each model for each end point to determine whether the relation between outcome and SSS
varied significantly over the extent of CAD. In addition, the measurements summed rest score and
summed difference score were evaluated separately in
a similar incremental fashion, with each score replacing SSS in the modeling scenario. Thus, we used
prognostic risk-adjusted models that contained significant clinical variables to provide a base for
evaluating the incremental prognostic value of angiography and perfusion imaging versus that of
clinical data.
RESULTS
Clinical characteristics: Clinical characteristics, perfusion results, and angiographic results for the 3,275
patients stratified by levels of SSS are presented in
Table 1. Higher-risk SSS levels were associated with
male gender, previous revascularization procedures,
history of myocardial infarction, congestive heart failure, diabetes, and increased severity of coronary disease on angiography. Differences in age, race, and
anginal symptoms were not noted across perfusion
levels.
Follow-up, end points, and outcome events: Of 3,275
subjects, 571 died and 231 developed nonfatal myocardial infarctions. Of the deaths, 377 (66%) were
classified as cardiovascular. The composite end point
of cardiovascular death or myocardial infarction in-
184 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
cluded 558 patients, 50 of whom had a nonfatal myocardial infarction and later died. Median follow-up
was 3.1 years (25th and 75th percentiles 1.6 and 5.1),
and maximum follow-up was 8.1 years.
Unadjusted end point relations: The unadjusted relation between perfusion and angiography with study
end points was statistically significant in all patients.
Chi-square values associated with SSS for death, cardiovascular death, and the composite of cardiovascular death or nonfatal myocardial infarction were 36.6,
57.6, and 51.7, respectively. Each unit change in SSS
was associated with an absolute 4% increase in risk
for death (95% confidence interval [CI] 3 to 6), a 7%
increase in risk for cardiovascular death (95% CI 5 to
8), and a 5% increase in risk for death or nonfatal
myocardial infarction (95% CI 4 to 7, p ⬍0.001 for
each). Unadjusted chi-square values associated with
angiography for these end points were 49.5, 65.7, and
74.8, respectively. Each increase in the number of
vessels with significant stenosis was associated with a
31% increased risk for death (95% CI 21 to 41), a 47%
increased risk for cardiovascular death (95% CI 33 to
61), and a 39% increased risk for death or nonfatal
myocardial infarction (95% CI 29 to 50, p ⬍0.001 for
each). Figure 2 shows the relation between SSS and
cardiovascular death that was produced with a Cox’s
model-fitting technique that involved cubic spline
functions. A continuous linear plot of the spline between SSS and the logarithm of the hazard ratio for
cardiovascular death exists, thus satisfying the linearity assumption of Cox’s proportional hazards model.
Similar linear trends were observed for SSS and the
other 2 end points.
To illustrate risk stratification by SSS, we grouped
SSS values into the following 3 categories of 0 to 4, 5
to 14, and ⱖ15. Survival rate by SSS category is
shown in Figure 3. Similar trends were observed for
the all-cause mortality rate and the composite of nonJANUARY 15, 2005
fatal myocardial infarction and cardiovascular mortality (Figure 4). Unadjusted 5-year event rates for each
risk group stratified by angiographic
results are presented in Table 2. Unadjusted 5-year event rates for patients who had an SSS of 0 stratified
by clinical characteristics are presented in the Table 3.
Multivariable Analyses: CLINICAL
Significant risk factors
from the clinical multivariable
models for all 3 outcomes included
age, New York Heart Association
congestive heart failure severity
class, Charlson’s co-morbidity index, diabetes, vascular disease, history of myocardial infarction, and
previous percutaneous coronary intervention. Each variable was assoFIGURE 2. Polynomial spline plot of a linear relation between cardiovascular morciated with an increased risk, extality rate and results of SPECT imaging (SSS). LL ⴝ lower limit of confidence
cept for previous percutaneous
band; UL ⴝ upper limit of confidence band.
coronary intervention, which was
associated with decreased risk. In
addition to this list, the all-cause
mortality end point model included
ventricular gallop. The covariate
list for the cardiovascular mortality end point model
also included race, with minority race being associated with a lower cardiovascular survival rate.
The cardiovascular mortality or nonfatal myocardial infarction end point model also included hypertension and minority race as significant
covariates.
ANGIOGRAPHIC AND SPECT MODELING RESULTS: Clinical information was a powerful predictor of outcomes; however, coronary angiography and SSS provided additional and incremental prognostic information.
Figure 5 shows chi-square values for the 3 models after
angiographic data and SSS were added to the clinical
models. The 8 clinical variables provided 91% of the
prognostic information for estimating the risk of death,
83% for cardiovascular death, and 85% for cardiovascular death or myocardial infarction. However, SSS and
angiographic data provided further prognostic information beyond these clinical variables and beyond each
other (p ⬍0.001 for all incremental chi-square values;
Figure 5). For each outcome (cardiovascular death or
cardiovascular death/myocardial infarction), summed
rest score (chi-square 33.4, p ⫽ 0.0001 and chi-square
17.5, p ⫽ 0.0001, respectively) and summed difference score (chi-square 4.1, p ⫽ 0.0418 and chi-square
5.4, p ⫽ 0.0197, respectively) contributed no more
than a similar amount of additional prognostic information to the clinical variables compared with SSS
results.
None of the interaction terms, SSS by angiographic
results, tested in the multivariable models were significant (all interaction terms p ⬎0.2). Thus, the relation between SSS and outcome remained constant
FIGURE 3. Kaplan-Meier cardiovascular event-free survival
across the extent of CAD (defined by the number of
curves demonstrating risk represented by SSS values 0 to 4, 5 to
diseased vessels) for each end point.
14, and >15.
VARIABLES:
CORONARY ARTERY DISEASE/PROGNOSTIC POWER OF SPECT
185
graphic results or did not use clinical
models that adjusted for noncardiac
(Charlson’s co-morbidity index)3 and
cardiac co-morbidities. In our series,
in which all patients underwent cardiac catheterization and SPECT imaging, we observed 5.1% and 6.3%
1-year rates for death and combined
cardiovascular death or myocardial
infarction, respectively.
Despite a high prevalence of previous myocardial infarction in patients
who had an SSS ⱖ15, in the overall
patient population, the summed difference score and summed rest score also
provided independent prognostic information. When we observed a higher
prevalence of patients who had previFIGURE 4. Event rates at follow-up intervals for all end points for SSS values 0 to 4, 5
ous myocardial infarction, the summed
to 14, and >15. CV ⴝ cardiovascular; MI ⴝ myocardial infarction.
rest score and, hence, SSS were higher,
indicating worsening outcomes.
Although we excluded patients
DISCUSSION
This study involved a high-risk cohort of patients who underwent revascularization ⱕ60 days, we exwho had a high rate of revascularization and under- amined and determined that these patients had higher
went cardiac catheterization ⱕ6 months of their index summed difference scores compared with the overall
SPECT study. We have collected a comprehensive list study population (p ⫽ 0.0001). Further, the summed
of clinical, angiographic, and nuclear imaging vari- rest score was not different (p ⫽ 0.3976). Including
ables and have developed a robust clinical model to these patients did not change the interpretation of
understand their prognostic value. We found a strong SPECT results.
No interactions between SSS and number of disassociation between SSS and all-cause mortality rate
and between SSS and cardiovascular mortality rate. eased vessels were found; however, even after adjustClinical variables provided ⬎80% of the prognostic ment for clinical and angiographic data, SSS was
information available from the combination of clinical prognostically important. Thus, these data support the
data, SPECT imaging results, and angiographic re- use of SPECT imaging to provide important informasults. After adjustment for these clinical and angio- tion to physicians, even when angiographic results are
graphic variables, SPECT imaging provided addi- known.
Our research used a 12-segment system model to
tional important prognostic information.
Many diagnostic tests and procedures, including describe the location of cardiac perfusion defects becardiac catheterization, are available to risk stratify cause our database supported only 12 segments at the
patients who have CAD into high- and low-risk sub- outset of the data collection period. Although use of
groups to guide further invasive or noninvasive treat- the recommended segmental model27 would provide a
ment. The Duke Databank for Cardiovascular Disease greater ability to define regions of defect within the
has been used to validate the prognostic importance of myocardium and possibly offer a higher resolution of
exercise treadmill testing and of clinical variables and the SSS, the use of a 12-segment model does not
to assess the prognostic utility of exercise ejection detract from the validity of our results.
Despite multiple previous studies that have demfraction calculated by radionuclide angiography. More
recently, it was used to demonstrate the incremental onstrated the usefulness of the poststress ejection fracprognostic value of ejection fraction by first-pass tion to predict outcome,28 we did not incorporate a
gated SPECT ejection fraction into this study. At the
RNA during pharmacologic stress testing.
Although previous studies have evaluated the time of data collection for our database, gated SPECT
relative importance of thallium 201 imaging in pa- imaging was not a standard component of the diagtients who undergo cardiac catheterization21,22 and nostic study in our laboratory. Based on previous
have made efforts to quantify the relative contribu- results, it is likely that a gated SPECT ejection fraction of various radionuclide, clinical, and catheter- tion would provide further incremental prognostic
ization variables,23 they have not addressed the value beyond the clinical variables and perhaps beprognostic contribution when the extent of diseased yond the perfusion score, particularly for the prediccoronary anatomy was known.24 –26 The use of tion of cardiovascular mortality rate.
Despite these limitations but because of the strong
SPECT as a prognostic aid has been validated in
numerous studies,7,24 –26 but these populations were association between SSS and clinical outcomes, the
generally at low risk for disease or had low ob- potential exists to use SPECT perfusion imaging as an
served event rates, or their respective methodolo- intermediate biomarker to draw inferences about those
gies did not assess the effect of coronary angio- therapies that may ultimately produce clinical benefit.
186 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
TABLE 2 Unadjusted Five-year Event Rates Categorized by Summed Stress Score Risk Strata and Number of Diseased Vessels
SSS Group
No. of
Diseased Vessels
No. of
Patients/Group
Death (%)
CV Death (%)
CV Death/MI (%)
0
1
2
3
0
1
2
3
0
1
2
3
717
437
245
270
217
302
322
456
23*
53
65
168
14.8
21.1
22.5
28.9
14.3
21.8
24.6
33.9
—
21.3
28.9
42.9
8.2
10.2
14.9
22.2
10.2
13.3
18.1
26.2
—
14.0
27.8
35.3
14.0
18.1
26.3
29.8
14.6
17.6
27.3
36.1
—
17.3
47.7
42.0
0–4
5–14
ⱖ15
*Insufficient number of patients to assess event rates.
CV ⫽ cardiovascular; MI ⫽ myocardial infarction.
TABLE 3 Unadjusted Five-year Event Rates at Zero Summed Stress Score
Clinical Characteristic
No. of diseased vessels
Stress type
Age group
Charlson’s co-morbidity categories
Levels
No. of
Patients/Group
Death Rate
(%)
CV Death Rate
(%)
CV Death/MI
Rate (%)
0
1
Multiple
Exercise
Pharmacologic
Young (⬍70 yrs)
Old (ⱖ70 yrs)
⬎1
0 or 1
364
247
238
344
505
613
236
204
645
17.7
20.0
21.9
10.1
36.0
16.7
28.8
27.3
17.0
9.5
9.4
12.7
5.8
19.2
14.8
23.3
14.7
9.0
14.4
16.6
20.9
10.9
27.7
7.9
18.1
22.9
15.0
Abbreviations as in Table 2.
approach, if successful, would further support the usefulness of SSS to
cardiology and clinical decision
making.
1. Pryor DB, Shaw L, McCants CB, Lee KL, Mark DB,
Harrell FE Jr, Muhlbaier LH, Califf RM. Value of the
history and physical in identifying patients at increased
risk for coronary artery disease. Ann Intern Med 1993;
118:81–90.
2. Borges-Neto S, Shaw LJ, Kesler K, Sell T, Peterson
ED, Coleman RE, Jones RH. Usefulness of serial radionuclide angiography in predicting cardiac death after
coronary artery bypass grafting and comparison with
clinical and cardiac catheterization data. Am J Cardiol
1997;79:851– 855.
3. Charlson ME, Pompei P, Ales KL, MacKenzie CR.
A new method of classifying prognostic comorbidity in
longitudinal studies: development and validation.
J Chron Dis 1987;40:373–383.
4. Gibbons RJ, Balady GJ, Beasley JW, Bricker JT,
Duvernoy WF, Froelicher VF, Mark DB, Marwick TH,
FIGURE 5. Model chi-square values demonstrating incremental prognostic information
McCallister BD, Thompson PD, et al. ACC/AHA
for clinical variables (black bars), clinical ⴙ angiographic variables (bars with horiguidelines for exercise testing: executive summary. A
zontal lines), clinical variables ⴙ SSS (white bars), and clinical variables ⴙ angioreport of the American College of Cardiology/Amerigraphic variables ⴙ SSS (bars with vertical lines). Abbreviations as in Figure 4.
can Heart Association Task Force on Practice Guidelines (Committee on Exercise Testing). Circulation
1997;96:345–354.
Novel anti-ischemic regimens or new revasculariza- 5. Mast ST, Shaw LK, Ravizzini GC, Chambless M, Joski P, Coleman RE,
S. Incremental prognostic value of RNA ejection fraction measuretion techniques could potentially be screened initially Borges-Neto
ments during pharmacologic stress testing: a comparison with clinical and perfor an effect on SSS. Interventions that decrease SSS fusion variables. J Nucl Med 2001;42:871– 877.
would be promising to test in adequately powered 6. Borges-Neto S. Perfusion and function assessment by nuclear cardiology
techniques. Curr Opin Cardiol 1997;12:581–586.
clinical trials with long-term follow-up to determine 7.
Berman DS, Kang X, Van Train KF, Lewin HC, Cohen I, Areeda J, Freidman
whether they also improve clinical outcomes. This JD, Germano G, Shaw LJ, Hachamovitch R. Comparative prognostic value of
CORONARY ARTERY DISEASE/PROGNOSTIC POWER OF SPECT
187
automatic quantitative analysis versus semiquantitative visual analysis of exercise
myocardial perfusion single-photon emission computed tomography. J Am Coll
Cardiol 1998;32:1987–1995.
8. Harris PJ, Harrell FE Jr, Lee KL, Rosati RA, Nonfatal myocardial infarction
in medically treated patients with coronary artery disease. Am J Cardiol 1980;
46:937–942.
9. Shaw LJ, Hachamovitch R, Heller G, Marwick TH, Travin MI, Iskandrian AE,
Kesler K, Lauer MS, Hendel R, Borges-Neto S, et al. Noninvasive strategies for
the estimation of cardiac risk in stable chest pain patients. Am J Cardiol 2000;
86:1–7.
10. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 481.
11. Cox DR. Regression models and life tables (with discussion). J R Stat Soc B
1972;34:187–220.
12. Breslow NE. Covariance analysis of censored survival data. Biometrics
1974;30:89 –99.
13. Harrell FE Jr, Pollock BG, Lee KL. Graphical methods for the analysis of
survival data. In: Proceedings of the 12th Annual SAS Users Group International
Conference. Cary, NC: SAS Institute, 1987:1107–1115.
14. Smith PL. Splines as a useful and convenient statistical tool. Am Stat
1979;33:57– 62.
15. Stone CJ, Koo C. Additive splines in statistics. In: Proceedings of the
Statistical Computing Section. Alexandria, VA: American Statistical Association,
1985:45– 48.
16. Devlin TF, Weeks BJ. Spline functions for logistic regression modeling. In:
Proceedings of the 11th Annual SAS Users Group International Conference.
Cary, NC: SAS Institute, 1986:646 – 651.
17. Harrell FE Jr, Lee KL, Pollock BG. Regression models in clinical studies:
determining relationships between predictors and response. J Natl Cancer Inst
1988;80:1198 –1202.
18. Rosati RA, McNeer JF, Starmer CF, Mittler BS, Morris JJ Jr, Wallace AG.
A new information system for medical practice. Arch Intern Med 1985;135:
1017–1024.
19. Califf RM, Mark DB, Harrell FE Jr, Hlatky MA, Lee KL, Rosati RA, Pryor
DB. Importance of clinical measures of ischemia in the prognosis of patients with
documented coronary artery disease. J Am Coll Cardiol 1988;11:20 –26.
188 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
20. Harris PJ, Lee KL, Harrell FE Jr, Behar VS, Rosati RA. Outcome in
medically treated coronary artery disease: ischemic events: nonfatal infarction
and death. Circulation 1980;62:718 –726.
21. Kaul S, Finkelstein DM, Homma S, Leavitt M, Okada RD, Boucher CA.
Superiority of quantitative exercise thallium-201 variables in determining longterm prognosis in ambulatory patients with chest pain: a comparison with cardiac
catheterization. J Am Coll Cardiol 1988;12:25–34.
22. Kaul S, Lilly DR, Gascho JA, Watson DD, Gibson RS, Oliner CA, Ryan JM.
Prognostic utility of the exercise thallium-201 tests in ambulatory patients with
chest pain: comparison with cardiac catheterization. J Am Coll Cardiol 1988;
12:25–34.
23. Lee KL, Pryor DB, Pieper KS, Harrell FE Jr, Califf RM, Mark DB, Hlatky
MA, Coleman RE, Cobb FR, Jones RH. Prognostic value of radionuclide angiography in medically treated patients with coronary artery disease. Circulation
1990;82:1705–1717.
24. Hachamovitch R, Berman DS, Shaw LJ, Kiat H, Cohen I, Cabico JA,
Friedman J, Diamond GA. Incremental prognostic value of myocardial perfusion
single photon emission computed tomography for the prediction of cardiac death.
Differential stratification for risk of cardiac death and myocardial infarction.
Circulation 1998;97:535–543.
25. Hachamovitch R, Berman DS, Kiat H, Cohen I, Cabico JA, Friedman J,
Diamond GA. Exercise myocardial perfusion SPECT in patients without known
coronary artery disease: incremental prognostic value and impact on subsequent
patient management. Circulation 1996;93:905–914.
26. Iskandrian S, Iskandrian AE. Risk assessment using single-photon emission
computed tomographic technetium-99m sestamibi imaging. J Am Coll Cardiol
1998;32:57– 62.
27. Berman D, Port SC, Garcia E, Sinusas A, Wackers F. Imaging guidelines for
nuclear cardiology procedures. J Nucl Cardiol 1996;6:G49 –G84.
28. Sharir T, Germano G, Kang X, Lewin HC, Miranda R, Cohen I, Agafitei RD,
Friedman JD, Berman DS. Prediction of myocardial infarction versus cardiac
death by gated myocardial perfusion SPECT: risk stratification by the amount of
stress-induced ischemia and the poststress ejection fraction. J Nucl Med 2001;
42:831– 837.
JANUARY 15, 2005
Comparative Effects of Rosuvastatin and
Gemfibrozil on Glucose, Insulin, and
Lipid Metabolism in Insulin-Resistant,
Nondiabetic Patients With
Combined Dyslipidemia
Cindy Lamendola, MSN, Fahim Abbasi, MD, James W. Chu, MD,
Howard Hutchinson, MD, Valerie Cain, MS, Elizabeth Leary, PhD,
Tracey McLaughlin, MD, Evan Stein, MD, PhD, and Gerald Reaven, MD
To evaluate the pharmacologic intervention most likely to
decrease cardiovascular disease risk in insulin-resistant
patients with combined dyslipidemia, 39 patients with this
abnormality were assessed before and after 3 months of
treatment with gemfibrozil (1,200 mg/day) or rosuvastatin
(40 mg/day) with regard to: (1) steady-state plasma glucose concentration at the end of a 180-minute infusion of
octreotide, insulin, and glucose; (2) fasting lipid, lipoprotein, and apolipoprotein concentrations; and (3) daylong
glucose, insulin, triglyceride, and remnant lipoprotein cholesterol concentrations in response to breakfast and lunch.
The 2 groups were similar at baseline in age, gender, body
mass index and in measurements of carbohydrate and
lipoprotein metabolism. Neither gemfibrozil nor rosuvastatin enhanced insulin sensitivity or lowered daylong glucose and insulin concentrations in insulin-resistant patients
with combined dyslipidemia, but both drugs significantly
decreased fasting triglyceride concentrations. However,
only rosuvastatin treatment significantly (p <0.05 to
<0.001) reduced fasting low-density lipoprotein cholesterol, apolipoprotein B-100, apolipoprotein C-III, apolipoprotein C-III:B particles, the apolipoprotein B-100:apolipoprotein A-I ratio, and increased apolipoprotein A-I (p
<0.05). The degree of improvement in fasting and postprandial remnant lipoprotein cholesterol concentrations
was significantly greater (p <0.05) in rosuvastatin-treated
patients, and this difference in the relative effectiveness of
the drugs was also true of the decrease in non– highdensity lipoprotein cholesterol concentrations. 䊚2005 by
Excerpta Medica Inc.
(Am J Cardiol 2005;95:189 –193)
pidemiologic studies have demonstrated that cardiovascular disease (CVD) risk is greater in paE
tients with combined dyslipidemia, defined by eleva-
in insulin-resistant, nondiabetic patients with combined dyslipidemia.
tions in both plasma triglyceride and low-density
lipoprotein (LDL) cholesterol concentrations, than in
those with either abnormality alone.1,2 In addition to
their abnormal lipoprotein metabolism, patients with
combined dyslipidemia are also insulin resistant and
hyperinsulinemic.3 Because the clinical utility of statin versus fibric acid drug treatment has not been
evaluated in patients with this clinical syndrome, we
compared the relative effects of gemfibrozil with rosuvastatin treatment on multiple aspects of fasting and
postprandial carbohydrate and lipoprotein metabolism
From the Department of Medicine, Stanford University School of Medicine, Stanford, California; AstraZeneca, Wilmington, Delaware; Pacific Biometrics Inc., Seattle, Washington; and Medical Research
Laboratories International, Highland Heights, Kentucky. This study was
supported by Research Grant RR-00070 from the National Institutes of
Health, Bethesda, Maryland, and a grant from AstraZeneca,
Wilmington, Delaware. Drs. Reaven and Stein received funding from
AstraZeneca for this study. Manuscript received June 8, 2004; revised
manuscript received and accepted September 8, 2004
Address for reprints: Gerald M. Reaven, MD, Falk Cardiovascular Research Center, Stanford University School of Medicine,
300 Pasteur Drive, Stanford, California 94305. E-mail: greaven@
cvmed.stanford.edu.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
METHODS
The Human Subjects Committee approved this
study, and all subjects provided written, informed
consent at the time of their initial visit to the General
Clinical Research Center. The experimental group
consisted of 39 apparently healthy patients who volunteered for this study in response to advertisements
in local newspapers. They were in good general
health, with a body mass index ⬍33.0 kg/m2, and had
elevations (⬎200 mg/dl) of both total plasma cholesterol and triglyceride concentrations while not taking
any lipid-lowering drugs for ⱖ6 weeks. In addition,
they had no known disease other than dyslipidemia,
were nondiabetic, with normal blood count, urinalysis,
and liver, kidney, and thyroid function tests. Volunteers meeting these inclusion criteria were instructed
on a weight-maintaining, low saturated fat diet and
were seen at intervals over the next 6 weeks. Those
who continued to have a total cholesterol concentration ⬎200 mg/dl and triglyceride concentrations between 200 and 600 mg/dl (with values differing by
⬍30%) at the end of the 6-week dietary run-in were
then randomized to receive treatment with either gemfibrozil or rosuvastatin. The age, gender distribution,
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.005
189
ide concentrations were measured as
previously described, and remnant lipoproteins were isolated by an imp Value
munoseparation method based on the
use of monoclonal antibodies to hu0.45
0.62
man apolipoproteins B-100 and A-I
0.28
to remove most of the apolipoprotein
B-100 – and apolipoprotein A-I– containing lipoproteins. The unbound
fraction is designated as remnant lipoproteins and quantified by determining the cholesterol concentration with a highly sensitive enzymatic
assay.7–9
On another morning after an overnight fast, insulin-mediated glucose disposal was quantified by a
modification10 of the insulin suppression test.11,12
Subjects were infused for 180 minutes with octreotide
acetate (0.27 ␮g/m2/min), insulin (32 mU/m2/min),
and glucose (267 mg/m2/min), blood drawn at 10minute intervals from 150 to 180 minutes, plasma
glucose and insulin concentrations measured, and the
mean of these values used as the steady-state plasma
insulin and glucose concentrations for each patient.
The steady-state plasma glucose concentration provides a direct measure of the ability of insulin to
mediate disposal of an infused glucose load; the
higher the steady-state plasma glucose concentration,
the more insulin-resistant the patient.
After completion of these baseline measurements,
treatment was initiated with gemfibrozil (600 mg
twice daily) or rosuvastatin (40 mg/day). Subjects
were seen at 2-week intervals to evaluate their wellbeing, drug and dietary compliance, and liver function; creatine kinase was measured at monthly intervals. All baseline measurements were repeated after 3
months of drug treatment.
Results are expressed as mean ⫾ SEM unless
otherwise stated. Statistical significance of the absolute changes in the fasting concentrations of the experimental variables resulting from each treatment
were compared by Student’s paired t test, with comparisons of the differences between the effects of each
treatment by Student’s unpaired t test. Absolute
changes in postprandial values were evaluated by
2-way analysis of variance. In all cases, p values
⬍0.05 were considered statistically significant.
TABLE 1 Baseline Clinical Characteristics of the Two Experimental Groups
Variable
Gemfibrozil
(n ⫽ 19)
Rosuvastatin
(n ⫽ 20)
Age (yrs)
Men/women
Body mass index (kg/m2)
51 ⫾ 10
11/8
29.5 ⫾ 2.9
54 ⫾ 9
10/10
28.1 ⫾ 4.1
Values are expressed as mean ⫾ SD.
and body mass index of the 2 groups are compared in
Table 1, and it is apparent that these variables were
similarly distributed.
Before starting drug treatment, blood samples were
obtained after an overnight fast, and plasma separated
for determination of fasting lipid, lipoprotein, and
apolipoprotein concentrations in the following manner. Total cholesterol and triglyceride concentrations
were determined by enzymatic methods on a Hitachi
747 analyzer (Boehringer Mannheim Diagnostics, Indianapolis, Indiana) as previously described.4 Highdensity lipoprotein (HDL) was isolated using heparin-2 mol/L manganese chloride,5 and the cholesterol
and triglyceride content of the isolated HDL measured. Very-low-density lipoprotein (VLDL) and LDL
cholesterol and triglyceride concentrations were determined after the ultracentrifugal separation of LDL and
HDL from VLDL at a density ⬎1.006.6 Cholesterol
and triglyceride concentrations were measured at a
density ⬎1.006 (bottom) fraction, and the previously
measured HDL cholesterol and HDL triglyceride values subtracted to obtain LDL cholesterol and LDL
triglyceride levels. VLDL cholesterol and VLDL triglyceride levels were calculated by subtracting the
values obtained on the bottom fraction from those
measured in whole plasma. Apolipoproteins A-I, B,
and E were analyzed by rate immunonephelometry
(Dade-Behring, Marburg, Germany). Apolipoprotein
C-III was measured by electroimmunoassay, and the
apolipoprotein C-III content of nonapolipoprotein Bcontaining lipoproteins (apolipoprotein C-III:non B)
similarly measured after precipitation of all apolipoprotein B-containing particles with antisera to apolipoprotein B. The amount of apolipoprotein B associated C-III (apolipoprotien C-III:B) was obtained by
subtraction.
Plasma glucose and insulin concentrations were
measured every hour in response to 2 meals, given at
8 A.M. and at noon. Both meals contained (as percentages of total calories) 15% protein, 43% carbohydrate,
and 42% fat (⬍10% saturated fat), with breakfast
comprising 20% and lunch comprising 40% of estimated daily caloric requirements. Insulin was measured in a stepwise sandwich enzyme-linked immunosorbent assay procedure on an ES 300 (Boehringer
Mannheim Diagnostics). Glucose was measured using
the hexokinase method on the Hitachi 747 (Boehringer Mannheim Diagnostics).
During the daylong meal tolerance test, blood was
also obtained at 2-hour intervals for determination of
fasting and postprandial plasma triglyceride and remnant lipoprotein cholesterol concentrations. Triglycer190 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
RESULTS
Fasting plasma lipid and lipoprotein concentrations,
before and after treatment, are listed in Table 2; importantly, at baseline both groups had the adverse CVD risk
profile of patients with combined dyslipidemia (i.e., elevations in total cholesterol and triglyceride concentrations) associated with low HDL cholesterol concentrations. Total cholesterol concentrations decreased in
response to both drugs (p ⬍0.001), but the reduction was
greater in response to rosuvastatin (⫺104 to ⫺17 mg/dl).
Not surprisingly, LDL cholesterol concentrations decreased significantly (p ⬍0.001) after administration of
rosuvastatin (⫺76 mg/dl), but did not change in response
to gemfibrozil (⫹5 mg/dl). VLDL and non-HDL cholesterol concentrations decreased significantly (p ⬍0.001)
JANUARY 15, 2005
TABLE 2 Fasting Lipid and Lipoprotein Concentrations Before and After Treatment
Gemfibrozil
Variable
Total cholesterol
VLDL cholesterol
LDL cholesterol
HDL cholesterol
Non–HDL cholesterol
Triglycerides
VLDL triglycerides
LDL triglycerides
HDL triglycerides
RLP cholesterol
Total cholesterol/HDL
cholesterol
Non–HDL cholesterol/
HDL cholesterol
Before
223
52
126
41
181
284
199
26
43
15
5.6
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
4
4
5
2
4
17
15
4
3
2
0.2
4.6 ⫾ 0.2
Rosuvastatin
After
p Value
Before vs After
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
5
4
5
3
4
23
20
4
4
1
0.2
⬍0.001
⬍0.001
NS
⬍0.001
⬍0.001
⬍0.001
⬍0.001
NS
⬍0.01
⬍0.005
⬍0.001
3.7 ⫾ 0.2
⬍0.001
206
31
131
46
160
166
106
26
34
10
4.7
Before
242
66
138
40
202
324
235
33
45
23
6.6
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
5
7
7
3
6
19
27
6
2
4
0.4
5.6 ⫾ 0.4
After
p Value
Before vs After
p Value*
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
6
4
4
3
6
18
19
2
2
1
0.2
⬍0.001
⬍0.001
⬍0.001
NS
⬍0.001
⬍0.001
⬍0.05
⬍0.05
⬍0.05
⬍0.002
⬍0.001
⬍0.001
NS
⬍0.001
NS
⬍0.001
NS
NS
⬍0.05
NS
⬍0.05
⬍0.001
2.4 ⫾ 0.2
⬍0.001
⬍0.001
138
35
62
42
96
211
163
14
39
9
3.4
Values are expressed as mean ⫾ SEM.
*Changes in gemfibrozil treatment compared with changes in rosuvastatin treatment.
RLP ⫽ remnant lipoprotein.
TABLE 3 Fasting Apolipoprotein Concentrations Before and After Treatment in Patients With Combined Dyslipidemia
Gemfibrozil
Variable
Apolipoprotein A-I
Apolipoprotein B-100
Apolipoprotein C-III
Apolipoprotein C-III:B
Apolipoprotein C-III:non-B
Apolipoprotein E
Apolipoprotein B-100/
apolipoprotein A-I
Before
141
137
4.5
2.5
2.1
4.8
1.0
⫾
⫾
⫾
⫾
⫾
⫾
⫾
6
4
0.3
0.2
0.2
0.3
0.1
After
144
129
3.8
2.1
1.8
4.2
0.9
⫾
⫾
⫾
⫾
⫾
⫾
⫾
8
4
0.4
0.2
0.2
0.3
0.0
Rosuvastatin
p Value
Before vs After
NS
NS
NS
NS
NS
⬍0.05
NS
Before
132
147
5.2
3.1
2.1
6.0
1.2
⫾
⫾
⫾
⫾
⫾
⫾
⫾
7
6
0.3
0.3
0.2
0.6
0.1
After
p Value
Before vs After
p Value*
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⬍0.05
⬍0.001
⬍0.01
⬍0.05
NS
⬍0.001
⬍0.001
NS
⬍0.001
NS
NS
NS
⬍0.001
⬍0.001
146
80
4.3
2.5
2.0
3.4
0.6
8
4
0.4
0.3
0.1
0.3
0.0
Values are expressed as mean ⫾ SEM.
*p values ⫽ changes in gemfibrozil treatment compared with changes in rosuvastatin treatment.
in response to either treatment (⫺21 and ⫺21 mg/dl,
respectively, for gemfibrozil vs ⫺31 and ⫺106 mg/dl for
rosuvastatin), but the decrease in non-HDL cholesterol
concentration was significantly greater in rosuvastatintreated patients (p ⬍0.001). The only variable that improved in response to gemfibrozil, but not rosuvastatin,
was an increase in HDL cholesterol concentrations (5 vs
2 mg/dl), but there was no significant difference when
the change associated with the 2 treatments was directly
compared.
In contrast to the disparate effects of the 2 drugs on
LDL cholesterol concentrations, both drugs lowered
total plasma triglyceride concentrations (p ⬍0.001),
and to a comparable degree (⫺118 vs ⫺113 mg/dl). In
addition, VLDL triglyceride and HDL triglyceride
concentrations were significantly lower after treatment
with either drugs, and the magnitude of the decrease
was not significantly different as a function of the drug
used. However, a decrease in LDL triglycerides
was only seen in rosuvastatin-treated patients (⫺19
mg/dl), and these patients also had a significantly
greater decrease in fasting remnant lipoprotein cholesterol concentrations (⫺14 mg/dl for rosuvastatin vs
⫺5 mg/dl for gemfibrozil). Finally, although the ratios
of total cholesterol/HDL cholesterol and non-HDL
cholesterol/HDL cholesterol were significantly lower
after treatment with either drug, they decreased to a
significantly greater degree in response to rosuvastatin
(⫺3.2 and ⫺3.2 for rosuvastatin vs ⫺0.9 and ⫺0.9 for
gemfibrozil, respectively).
Treatment-associated changes in apolipoprotein
concentrations are listed in Table 3. The only
change after gemfibrozil administration was a significant decrease in plasma apolipoprotein E concentrations (⫺0.6 mg/dl). Apolipoprotein E concentrations also decreased in rosuvastatin-treated
patients (⫺2.6 mg/dl), and the magnitude of the
decline was significantly greater (p ⬍0.001). In
addition, rosuvastatin treatment was associated with
an increase in apolipoprotein A-I concentrations (p
⬍0.05) of 14 mg/dl, and decreases of ⫺67 mg/dl in
the concentrations of apolipoprotein B-100 (p
⬍0.001), ⫺0.9 mg/dl in apolipoprotein C-III (p
⬍0.01), and ⫺0.7 mg/dl in apolipoprotein C-III:B
(p ⬍0.05), as well as ⫺0.6 in the plasma concen-
PREVENTIVE CARDIOLOGY/ROSUVASTATIN AND GEMFIBROZIL IN COMBINED DYSLIPIDEMIA
191
FIGURE 1. Comparison of daylong plasma triglycerides and remnant lipoprotein cholesterol (RLP) concentration in response to test
meals before and after treatment with gemfibrozil or rosuvastatin. The test meals contained (as a percentage of total calories) 15%
protein, 42% fat (<10% saturated fat), and 43% carbohydrates. Subjects ate breakfast at 8 A.M. (20% of daily caloric requirements)
and lunch at noon (40% of daily caloric requirements) with blood samples obtained after fasting and then every 2 hours for 8 hours.
tration ratio of apolipoprotein B-100/apolipoprotein
A-I (p ⬍0.001).
Steady-state plasma glucose concentrations did not
change in response to treatment with either gemfibrozil (192 ⫾ 15 to 201 ⫾ 14 mg/dl) or rosuvastatin
(192 ⫾ 13 to 207 ⫾ 16 mg/dl). Because steady-state
plasma insulin concentrations were also similar before
and after treatment with either drug (⬃80 mU/L),
insulin-mediated glucose disposal did not change with
treatment. Consequently, it was not surprising that
daylong plasma glucose and insulin levels were comparable before and after in both treatment groups. It is
worth emphasizing that the steady-state plasma glucose concentrations were quite high in both groups,
similar to previous data indicating that insulin resistance characterizes patients with combined dyslipidemia.3 Daylong concentrations of plasma triglycerides,
and remnant lipoprotein cholesterol, before and after
treatment with the 2 drugs are shown in Figure 1. It is
apparent from these data that treatment with either
drug led to significant (p ⬍0.001) decreases in daylong plasma triglyceride and remnant lipoprotein cholesterol concentrations. Although no difference was
observed between them with regard to effects on daylong triglyceride concentrations, the reductions in
remnant lipoprotein cholesterol were significantly
192 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
greater (p ⬍0.01) than with gemfibrozil after rosuvastatin treatment.
DISCUSSION
The results presented indicate that rosuvastatin
treatment of individuals with combined dyslipidemia not only decreases LDL cholesterol concentration to a value that could no longer be considered to
contribute to CVD risk but also lowers triglyceride,
VLDL triglyceride, and VLDL cholesterol concentrations to approximately the same degree as that in
gemfibrozil-treated patients. The finding that all of
the lipoprotein abnormalities that characterize subjects with combined dyslipidemia improved significantly in rosuvastatin-treated patients raises the
possibility that effective treatment of these patients
frequently may not require combination therapy
with a statin and a fibric acid.
Although treatment with either drug was associated
with significant decreases in plasma non-HDL cholesterol concentrations, as well in the plasma concentration ratios of total cholesterol/HDL cholesterol, and
non-HDL cholesterol/HDL cholesterol, in all 3 cases
the magnitude of the change was significantly greater
in rosuvastatin-treated patients. In addition, rosuvasJANUARY 15, 2005
tatin-treated patients had several changes in plasma
apolipoprotein concentrations that would suggest decreased CVD risk, including a statistically significant
increase in apolipoprotein A-I and decreases in apolipoprotein B-100, apolipoprotein C-III, apolipoprotein
C-III:B, apolipoprotein E, and the ratio of apolipoprotein B-100/apolipoprotein A-I. In contrast, the only
statistically significant changes in plasma apoprotein
concentrations seen in gemfibrozil-treated patients
were decreases in apolipoprotein E levels; even in this
case, the decrease in rosuvastatin-treated patients was
significantly greater.
Finally, the observation that the decrease in the
postprandial accumulation of remnant lipoproteins
was significantly greater after treatment with rosuvastatin is worthy of comment for 2 reasons. Clinically,
the atherogenic potential of increased concentrations
of triglyceride-rich remnant lipoproteins, initially emphasized by Zilversmit,13 has been repeatedly confirmed.14 –16 Physiologically, the daylong decreases in
postprandial triglyceride concentrations were similar
after both drugs, and the decrements in remnant lipoprotein cholesterol concentrations in gemfibroziltreated patents were similar to changes in postprandial
triglyceride levels. In contrast, the decrease in daylong
remnant lipoprotein cholesterol concentrations in rosuvastatin-treated subjects were accentuated compared with the decreases in postprandial triglycerides,
resulting in much lower postprandial remnant lipoprotein levels than levels in gemfibrozil-treated patients.
These data (1) are similar to our findings in patients
with combined dyslipidemia who also had type 2
diabetes,17 (2) demonstrate that the ability of satins to
upregulate the hepatic LDL receptor will also increase
the removal rate from plasma of remnant lipoproteins,
and (3) offer another example of how statin treatment
can decrease CVD risk above and beyond the ability
to lower LDL cholesterol levels.
1. Manninen V, Tenkanen L, Koskinen P, Huttunen JK, Manttari M, Heinonen
OP, Frick MH. Joint effects of serum triglyceride and LDL cholesterol and HDL
cholesterol concentrations on coronary heart disease risk in the Helsinki Heart
Study. Implications for treatment. Circulation 1992;85:37– 45.
2. Assmann G, Schulte H, Funke H, Von Eckardstein A. The emergence of
triglycerides as a significant independent risk factor in coronary artery disease.
Eur Heart J 1998;19(suppl):M8 –M14.
3. Sheu WH-H, Shieh SM, Fuh MM, Shen DD, Jeng CY, Chen Y-D, Reaven GM.
Insulin resistance, glucose intolerance, and hyperinsulinemia. Hypertriglyceridemia versus hypercholesterolemia. Arterioiscler Thromb Vasc Biol 1993;13:367–
370.
4. Steiner P, Freidel J, Bremner W, Stein E. Standardization of micromethods for
plasma cholesterol, triglyceride and HDL-cholesterol with the lipid clinics’
methodology. J Clin Chem 1981;19:850.
5. Warnick G, Albers J. A comprehensive evaluation of the heparin manganese
precipitation procedure for estimating high-density lipoprotein cholesterol. J
Lipid Res 1978;19:65–76.
6. Lipid Research Clinics Program. Manual of laboratory operations: Lipid and
lipoprotein analysis. Washington DC: US Department of Health Education and
Welfare. Publication NIH 75-628, 1982.
7. Campos E, Nakajima K, Tanaka A, Havel RJ. Properties of an apolipoprotein.
E-enriched fraction of triglyceride-rich lipoproteins isolated from human blood
plasma with a monoclonal antibody to apolipoprotein B-100. J Lipid Res 1992;
33:369 –380.
8. Nakajima K, Saito T, Tamura A, Suzuki M, Nakano T, Adachi M, Tanaka A,
Tada N, Nakamura H, Campos E, Havel R. Cholesterol in remnant-like lipoproteins in human serum using monoclonal anti apo B-100 and anti apo A-I
immunoaffinity mixed gels. Clin Chim Acta 1993;223:53–71.
9. Leary ET, Wang T, Baker DJ, Cilla DD, Zhong J, Warnick GR, Nakajima K,
Havel RJ. Evaluation of an immunoseparation method for quantitative measurement of remnant-like particle-cholesterol in serum and plasma. Clin Chem 1998;
44:2490 –2498.
10. Pei D, Jones CN, Bhargava R, Chen YD, Reaven GM. Evaluation of
octreotide to assess insulin-mediated glucose disposal by the insulin suppression
test. Diabetolgia 1994;37:843– 845.
11. Shen S-W, Reaven GM, Farquhar J. Comparison of impedance to insulinmediated glucose uptake in normal subjects and in subjects with latent diabetes.
J Clin Invest 1980;49:2151–2160.
12. Greenfield MS, Doberne L, Kraemer FB, Tobey TA, Reaven GM. Assessment of insulin resistance with the insulin suppression test and the euglycemic
clamp. Diabetes 1981;30:387–92.
13. Zilversmit DB. Atherogenesis: a postprandial phenomenon. Circulation
1979;60:473– 485.
14. Groot PH, van Stiphout WA, Krauss XH, Jansen H, van Tol A, van
Ramshorst E, Chin-On S, Hofman A, Cresswell SR, Havekes L. Postprandial
lipoprotein metabolism in normolipidemic men with and without coronary artery
disease. Arterioiscler Thromb Vasc Biol 1991;11:653– 662.
15. Patsch JR, Miesenbock G, Hopferwieser T, Muhlberger V, Knapp E, Dunn
JK, Gotto AM Jr, Patsch W. Relation of triglyceride metabolism and coronary
artery disease. Studies in the postprandial state. Arterioiscler Thromb Vasc Biol
1992;12:1336 –1345.
16. Karpe F, Bard JM, Steiner G, Carlson LA, Fruchart JC, Hamsten A. HDLs
and alimentary lipemia. Studies in men with previous myocardial infarction at a
young age. Arterioiscler Thromb Vasc Biol 1993;13:11–22.
17. McLaughlin T, Abbasi F, Lamendola C, Leary E, Reaven GM. Comparison
in patients with type 2 diabetes of fibric acid versus hepatic hydroxymethyl
glutaryl-coenzyme A reductase inhibitor treatment of combined dyslipidemia.
Metabolism 2002;51:1355–1359.
PREVENTIVE CARDIOLOGY/ROSUVASTATIN AND GEMFIBROZIL IN COMBINED DYSLIPIDEMIA
193
Relation Between Atherogenic
Dyslipidemia and the Adult Treatment
Program-III Definition of Metabolic
Syndrome (Genetic Epidemiology of
Metabolic Syndrome Project)
Diego F. Wyszynski, PhD, Dawn M. Waterworth, PhD, Philip J. Barter, MD,
Jonathan Cohen, PhD, Y. Antero Kesäniemi, MD, Robert W. Mahley, MD, PhD,
Ruth McPherson, MD, Gérard Waeber, MD, Thomas P. Bersot, MD,
Sanjay S. Sharma, BSc, Vikki Nolan, MPH, Lefkos T. Middleton, MD, PhD,
Scott S. Sundseth, PhD, Lindsay A. Farrer, MD, Vincent Mooser, MD, and
Scott M. Grundy, MD, PhD
Genetic Epidemiology of Metabolic Syndrome is a multinational, family-based study to explore the genetic
basis of the metabolic syndrome. Atherogenic dyslipidemia (defined as low plasma high-density lipoprotein
cholesterol with elevated triglycerides (<25th and
>75th percentile for age, gender, and country, respectively) identified affected subjects for the metabolic syndrome. This report examines the frequency at which
atherogenic dyslipidemia predicts the metabolic syndrome of the National Cholesterol Education Program
Adult Treatment Panel III (ATP-III). One thousand four
hundred thirty-six (854 men/582 women) affected patients by our criteria were compared with 1,672 (737
men/935 women) unaffected persons. Affected patients
had more hypertension, obesity, and hyperglycemia,
and they met a higher number of ATP-III criteria (3.2 ⴞ
1.1 SD vs 1.3 ⴞ 1.1 SD, p <0.001). Overall, 76% of
affected persons also qualified for the ATP-III definition
(Cohen’s ␬ 0.61, 95% confidence interval 0.59 to 0.64),
similar to a separate group of 464 sporadic, unrelated
cases (75%). Concordance increased from 41% to 82%
and 88% for ages <35, 36 to 55, and >55 years,
respectively. Affected status was also independently associated with waist circumference (p <0.001) and fasting glucose (p <0.001) but not systolic blood pressure (p
ⴝ 0.43). Thus, the lipid-based criteria used to define
affection status in this study substantially parallels the
ATP-III definition of metabolic syndrome in subjects aged
>35 years. In subjects aged <35 years, atherogenic
dyslipidemia frequently occurs in the absence of other
metabolic syndrome risk factors. 䊚2005 by Excerpta
Medica Inc.
(Am J Cardiol 2005;95:194 –198)
he present report is derived from the Genetic Epidemiology of Metabolic Syndrome project. This
T
is a large, multinational, family-based study exploring
affection status: elevated triglycerides and concomitant low high-density lipoprotein (HDL) cholesterol.
These criteria were used because they are (1) primary
features of atherogenic dyslipidemia, (2) associated
with insulin resistance, (3) detectable early in the
development of metabolic syndrome, (4) individually
highly heritable, and (5) easy to measure. The essential question addressed was how strongly familial
atherogenic dyslipidemia is associated with other
components of the metabolic syndrome as defined by
the ATP-III.1 The analysis thus was designed to determine whether most instances of atherogenic dyslipidemia occur together with the metabolic syndrome or
as an isolated dyslipidemia.
the genetic basis of the metabolic syndrome. Two
simple lipid-based criteria were used to define the
From the Department of Medicine (Genetics Program), Boston University School of Medicine, Boston, Massachusetts; Medical Genetics,
Genetic Research, GlaxoSmithKline, Collegeville, Pennsylvania; Heart
Research Institute, New South Wales, and Hanson Institute, University
of Adelaide, Adelaide, Australia; Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, Texas; Department
of Internal Medicine and Biocenter Oulu, University of Oulu, Oulu,
Finland; Gladstone Institute of Cardiovascular Disease and the University of California, San Francisco, California, and American Hospital,
Istanbul, Turkey; University of Ottawa Heart Institute, Ottawa,
Canada; and Department of Medicine, Lausanne University Hospital,
Lausanne, Switzerland. This research project was funded by GlaxoSmithKline, Inc., Collegeville, Pennsylvania. Manuscript received June
22, 2004; revised manuscript received and accepted August 31,
2004.
Address for reprints: Scott M. Grundy, MD, PhD, Center for
Human Nutrition, University of Texas Southwestern Medical Center at
Dallas, 5323 Harry Hines Boulevard, Y3.206, Dallas, Texas 753909052. E-mail: Scott.Grundy@UTSouthwestern.edu.
194
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
METHODS
Accrual centers in Australia, Canada, Finland,
Switzerland, Turkey, and the United States participated in this study. Analysis was performed in the
epidemiology center in Boston University. Caucasian
participants (aged 18 to 70 years) were considered
affected if they had both plasma triglycerides ⱖ75th
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.091
TABLE 1 Plasma Triglycerides and High-density Lipoprotein (HDL) Cholesterol Values Corresponding to Triglycerides ⱖ75th
Percentile and HDL in ⱕ25th Percentile for Specific Ages by Sex and Site*
Australia
TG
HDL
Canada
TG
Finland
HDL
TG
Switzerland
Turkey
United States
HDL
TG
HDL
TG
HDL
TG
HDL
46
41
41
40
40
40
124
124
133
151
151
151
41
41
42
43
43
39
109
155
191
189
168
120
35
35
35
35
35
35
138
159
198
194
184
174
39
40
36
37
37
37
52
51
50
49
49
49
106
106
106
115
124
124
54
53
52
54
51
51
83
110
136
165
146
145
40
40
40
40
40
40
116
129
145
183
195
171
45
44
44
45
46
46
Male Subjects
Age (yrs)
20
30
40
50
60
70
133
151
195
204
204
195
43
39
39
39
39
39
130
154
193
202
201
198
39
39
39
39
39
39
98
148
178
198
190
190
Female Subjects
Age (yrs)
20
30
40
50
60
70
124
115
124
168
186
195
46
46
46
46
46
46
121
118
120
170
182
197
46
46
46
46
46
46
94
110
112
133
161
161
*The youngest participants in both Switzerland and Australia were 18 years of age.
Despite both Australia and the United States using NHANES data to calculate percentiles, the cutpoints were originally calculated using different age groupings,
resulting in slightly different cutpoints by country.
Values are expressed as milligrams per deciliter.
percentile and serum HDL cholesterol ⱕ25th percentile, both adjusted for age and gender. Those meeting
the criteria will be referred to throughout as affected
by current criteria. Lipid values were either currently
obtained (80%) or obtained up to 3 years before entry
(20%). Population values vary in different countries;
in particular, HDL levels in Turkey are lower.2 Thus,
cut points were defined by national databases. The
following sources were used for different sites: Australia (National Health and Nutrition Examination
Survey [NHANES III] survey website [http://www.
cdc.gov/nchs/about/major/nhanes/nh3data.htm]); Canada (for subjects ⱖ18 years, Canadian Heart Health
Surveys 1986 to 1992 data,3 and for younger subjects,
Lipid Research Clinics prevalence study),4 Finland,
cardiovascular risk factor data5; Switzerland, Geneva
database on the epidemiology of cardiovascular risk
factors in public servants, Turkey [Turkish Heart
Study],6 and United States (NHANES III survey as
mentioned). Age and gender cut points for the 75th
percentiles for triglyceride and HDL cholesterol levels
for each study site are listed in Table 1.
Subjects were excluded if they (1) had a fasting
blood glucose levels ⱖ126 mg/dl (probands only), (2)
had a body mass index ⱖ35 kg/m2, (3) were positive
for human immunodeficiency virus, (4) were recipients of an organ transplant, (5) had familial hypercholesterolemia, or (6) were heavy alcohol users (⬎8
U/day). The minimal unit for collection of a family
was an affected sib-pair, but larger nuclear or extended families were collected if possible. In a small
number of instances, some patients originally believed
to be affected, on closer examination turned out not to
be affected, leading to a small proportion of families
with only 1 affected. An additional 464 singly-ascer-
tained patients were recruited from the same sites with
the same lipid phenotype as that used for the family
collection. The institutional review board at each participating site approved the protocol and informed
consent forms.
A standardized questionnaire was administered to
every family member. In addition to demographic
data, it captured co-morbid conditions, medications,
tobacco, and alcohol. Height, weight, and waist circumference, as well as 3 blood pressure measurements, were obtained for each subject. The average of
the second and third systolic and diastolic blood pressure readings was used in the analyses. Blood was
collected after a 12-hour fast. Laboratory tests for the
lipids and glucose levels were performed according to
standard procedures.
Statistical analyses were performed using SAS version 8 (SAS Institute, Cary, North Carolina) and Stata
version 8. Descriptive statistics are expressed as number (percent) or mean ⫾ SD, except for triglyceride
levels and blood pressure measurements, which are
expressed as medians. Statistical significance was determined using univariate logistic regression for continuous variables or a chi-square test for categorical
variables. For the linear regression analysis, PROC
GLM in SAS (SAS Institute) was used; waist circumference, fasting glucose, and systolic blood pressure
were treated as continuous variables and currently
affected status was a categorical variable.
RESULTS
In all, 3,273 patients from 504 families were recruited in the study. The 6 accrual centers contributed
33 to 136 families, and the size of the families ranged
from 2 to 179. Of 3,273 participants from the family
PREVENTIVE CARDIOLOGY/ATHEROGENIC DYSLIPIDEMIA AND METABOLIC SYNDROME
195
TABLE 2 Characteristics of Participants Based on the Genetic Epidemiology of Metabolic Syndrome Criteria*
Men
Variable
Age (yrs)
Ever smoked
Current smoker
No. of cigarettes/d
No. of alcoholic drinks/wk
Waist circumference (cm)
High waist circumference ⬎102 cm (M), 88 cm (F)
Waist-to-hip ratio
Body mass index (kg/m2)
Body mass index ⱖ30 kg/m2
Diagnosed diabetes mellitus
Diagnosed systemic hypertension
Taking blood pressure medication
Blood pressure (median, mm Hg)
Elevated blood pressure†
Taking lipid-altering medication
Total cholesterol (mg/dl)
HDL cholesterol (mg/dl)
HDL cholesterol ⬍40 mg/dl (M), ⬍50 mg/dl (F)
LDL cholesterol (mg/dL)
Triglycerides (median, mg/dl)
Triglycerides (ⱖ150 mg/dl)
Fasting glucose (mg/dl)
Glucose ⱖ110 mg/dl‡
Women
Affected
(n ⫽ 854)
Unaffected
(n ⫽ 737)
Affected
(n ⫽ 582)
Unaffected
(n ⫽ 935)
45.9 ⫾ 14.0
541 (63%)
229 (26%)
21.2 ⫾ 14.3
5.5 ⫾ 6.7
100.2 ⫾ 12.0
333 (41%)
0.94 ⫾ 0.06
28.2 ⫾ 3.9
291 (34%)
75 (9%)
275 (32%)
208 (24%)
133/84
583 (69%)
282 (33%)
211.9 ⫾ 56.1
32.4 ⫾ 5.4
829 (97%)
118.2 ⫾ 40.5
265.5
796 (93%)
99.6 ⫾ 23.5
143 (17%)
40.4 ⫾ 18.4§
392 (53%)§
210 (29%)§
19.3 ⫾ 13.1
5.3 ⫾ 6.4
90.0 ⫾ 14.4§
121 (17%)§
0.89 ⫾ 0.08§
25.1 ⫾ 4.4§
96 (13%)§
25 (3%)§
110 (15%)§
74 (10%)
128/80§
378 (53%)§
64 (9%)§
188.2 ⫾ 46.0§
46.0 ⫾ 11.4§
221 (30%)§
118.1 ⫾ 39.9
109.0§
190 (26%)§
91.9 ⫾ 17.9§
52 (7%)§
45.3 ⫾ 16.4
244 (42%)
105 (18%)
15.5 ⫾ 11.3
3.6 ⫾ 3.8
91.5 ⫾ 14.0
332 (60%)
0.85 ⫾ 0.08
28.1 ⫾ 4.7
211 (36%)
60 (11%)
198 (34%)
165 (28%)
131/83
380 (67%)
140 (24%)
210.0 ⫾ 49.6
37.3 ⫾ 6.5
573 (98%)
121.8 ⫾ 42.8
215.0
476 (82%)
99.0 ⫾ 31.3
84 (15%)
42.1 ⫾ 18.0§
326 (35%)储
151 (16%)
12.5 ⫾ 9.4储
4.1 ⫾ 5.5
81.7 ⫾ 12.9§
270 (30%)§
0.82 ⫾ 0.09§
24.7 ⫾ 4.5§
128 (14%)§
19 (2%)§
148 (16%)§
108 (12%)§
129/79§
422 (46%)§
60 (7%)§
192.6 ⫾ 43.9§
57.0 ⫾ 14.0§
290 (31%)§
114.8 ⫾ 38.1储
88.7§
129 (14%)§
88.3 ⫾ 16.8§
32 (3%)§
*The Genetic Epidemiology of Metabolic Syndrome plasma lipid criteria: triglycerides ⱖ75th percentile, HDL cholesterol ⱕ25th percentile (age-adjusted, country
and gender-specific).
†
Systolic ⱖ130 mm Hg and/or diastolic ⱖ85 mm Hg.
‡
Does not include treated diabetics
§
p ⱕ0.05; 储p ⱕ0.01; §p ⱕ0.001 versus affected subjects.
Values are expressed as number (%) or mean ⫾ SD.
ATP-III criteria to define metabolic syndrome are depicted in italics.
TABLE 3 Characteristics of Singly-ascertained Cases
Variable
Men
Age (yrs)
Ever smoked
Current smoker
No. of cigarettes/d
No. of alcoholic drinks/wk
Waist circumference (cm)
High waist circumference ⬎102 cm
(M), ⬎88 cm (W)
Waist-to-hip ratio
Body mass index (kg/m2)
Body mass index ⱖ30 kg/m2
Diagnosed hypertension
Blood pressure (median, mm Hg)
Elevated blood pressure
Taking lipid medication
Total cholesterol (mg/dl)
HDL cholesterol (mg/dl)
HDL cholesterol ⬍40 mg/dl (M),
⬍50 mg/dl (W)
LDL cholesterol (mg/dl)
Triglycerides (median, mg/dl)
Triglycerides ⱖ150 mg/dl
Fasting glucose (mg/dl)
Glucose ⱖ110 mg/dl
Cases (n ⫽ 464)
340 (73%)
50.6 ⫾ 8.6
295 (64%)
109 (24%)
21.0 ⫾ 13.3
NA
97.6 ⫾ 15.3
230 (50%)
0.92 ⫾ 0.07
28.9 ⫾ 3.3
179 (39%)
160 (34%)
132/81
279 (60%)
206 (45%)
221.9 ⫾ 45.3
35.7 ⫾ 6.2
429 (92%)
132.1 ⫾ 45.3
310.7
443 (95%)
96.1 ⫾ 11.2
43 (11%)
Values are presented as number percentage or mean ⫾ SD.
LDL ⫽ low-density lipoprotein.
196 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
dataset, 1,436 (44%) were considered as affected using the current definition, 1,672 (51%) were unaffected, and the remaining 165 patients (5%) were
considered unknown due to absence of reliable measurements of either HDL cholesterol or triglycerides
and thus were excluded from analysis. The clinical
characteristics of the 1,436 (854 men/582 women)
affected and the 1,672 (737 men/935 women) unaffected family members, as well as the proportion of
these patients who met each patient ATP-III criterion,
are listed in Table 2. Age, waist circumference, waistto-hip ratio, body mass index, the proportion of obese
patients, blood pressure, and the proportion of hypertensive subjects, blood glucose levels, and proportion
of persons with blood glucose levels ⬎110 mg/dl were
all significantly higher among affected versus unaffected men and women (p ⬍0.001). These differences
were not due to differences in alcohol consumption or
current cigarette smoking. In addition, a separate set
of 464 unrelated patients was ascertained based on the
same criteria as the familial subjects, except that they
were not required to have an affected relative. The
characteristics of this singly-ascertained group are
listed in Table 3.
To explore the concordance between the current
and the ATP-III definitions for metabolic syndrome,
JANUARY 15, 2005
FIGURE 1. Frequency distribution (percent) of genetic epidemiology of metabolic syndrome-affected (closed bars) and -unaffected (open bars) patients according to the number of ATP-III
criteria and age. (ATP-III criteria to define metabolic syndrome
are listed in Table 1.) The diagnosis of metabolic syndrome is
made if >3 of these criteria are met (corresponding to right side
of dotted line).
we first partitioned 3,030 subjects from the family
dataset who had sufficient information for ATP-III
classification according to these 2 definitions. Among
those affected by current criteria, 76% met the ATPIII definition for metabolic syndrome, whereas 85% of
those unaffected by current criteria were also negative
for the ATP-III definition, with a Cohen’s ␬ of 0.61
(95% confidence interval [CI] 0.59 to 0.64), indicating
overall substantial agreement. To establish how much
of the discordance between the 2 definitions was due
to our different criteria for dyslipidemia, and how
much was due to coexistence of dyslipidemia with
other metabolic syndrome factors, we repeated this
analysis after replacing the absolute thresholds for
triglyceride and HDL cholesterol in ATP-III with the
current thresholds to create a pseudodefinition and
again looked for concordance with the current definition. Among those positive for both definitions, 81%
also met the ATP-III definition for metabolic syndrome, whereas 91% of those unaffected by current
criteria were also unaffected by ATP-III, with a Cohen’s ␬ of 0.72 (95% CI 0.69 to 0.74). Thus, overall
concordance was increased, and only a small proportion of subjects (4.7%) affected by current criteria did
not achieve ATP-III status due to differences in the
lipid thresholds.
We next stratified the participants in this study
according to the number of ATP-III criteria they met
and by age categories (Figure 1). In this figure, solid
bars located left of the dotted line correspond to “false
positives,” whereas solid bars on the right correspond
to “true positives,” if the ATP-III definition is used as
the “gold standard” to define metabolic syndrome.
This analysis revealed that in the age group ⱕ35 years
old, ⬎50% of the participants (58%) who were affected by current criteria were not affected by ATPIII. In contrast, the rate of false positives decreased to
14% in those aged 36 to 55 years and to 12% in older
subjects. The overall agreement between the 2 criteria
as described by the Cohen’s ␬ showed a similar pattern, with only moderate agreement in those ⬍35
years (␬ 0.42, 95% CI 0.36 to 0.48), but substantial
agreement in both those 35 to 55 years (␬ 0.66, 95%
CI 0.60 to 0.71) and in those ⬎55 years (␬ 0.58, 95%
CI 0.51 to 0.65).
Because currently defined status was fairly well
correlated with ATP-III status overall, it was interesting to explore which nondyslipidemic components of
ATP-III (waist, fasting glucose, and systolic blood
pressure) were most closely associated with the current definition of dyslipidemia. A linear regression
analysis performed on waist circumference, fasting
glucose, and systolic blood pressure is presented in
Table 4. The parameter estimates correspond to the
mean increase in the trait under evaluation in affected
subjects by current criteria compared with unaffected
subjects. Univariate analysis indicates unadjusted association, whereas model 1 adjusts for standard confounders (age, gender, center, smoking, and low-density lipoprotein cholesterol), and model 2 further
adjusts for metabolic syndrome-associated factors.
Affection status by current data was closely associated
with each of the 3 traits in the univariate analysis, and
the association remained significant after adjustment
for confounders, although with a reduced magnitude
of effect for both waist circumference and fasting
glucose levels. In contrast, after adjustment for factors
related to metabolic syndrome, systolic blood pressure
was no longer associated with status by current criteria, suggesting that the observed association with current data was a secondary effect of ⱖ1 of the syndromic factors. Further analysis showed that waist
was the specific variable that ablated this association
(data not shown).
To assess if the concordance between the 2 definitions was influenced by the familial nature of dyslipidemia in the present study, the same comparison
was performed in the additional set of 464 patients
affected by current criteria who were singly ascertained. Overall, 75% of these sporadic cases also were
affected for ATP-III metabolic syndrome, a figure
similar to that observed in the family sample. Thus,
the level of concordance between the 2 criteria does
not appear to have been substantially affected by the
familial nature of affected family members.
PREVENTIVE CARDIOLOGY/ATHEROGENIC DYSLIPIDEMIA AND METABOLIC SYNDROME
197
TABLE 4 Association of the Genetic Epidemiology of Metabolic Syndrome Criteria With Systolic Blood Pressure, Waist
Circumference, and Fasting Glucose
Systolic Blood Pressure (mm Hg)
Univariate
Model 1*
Model 2†
Parameter
Estimate
95% CI
7.1
3.8
0.6
5.6–8.6
2.4–5.1
⫺0.9–2.2
p Value
⬍0.0001
⬍0.0001
0.43
Waist Circumference (cm)
Parameter
Estimate
95% CI
11.3
8.2
7.0
10.3–12.3
7.3–9.1
6.0–7.9
p Value
⬍0.0001
⬍0.0001
⬍0.0001
Fasting Glucose (mg/dl)
Parameter
Estimate
95% CI
9.5
7.0
3.6
7.9–11.1
5.4–8.6
1.8–5.3
p Value
⬍0.0001
⬍0.0001
⬍0.0001
*Adjusted for age, gender, center, smoking, and LDL cholesterol.
†
Systolic blood pressure further adjusted for waist, fasting glucose, and hypertension medications; waist circumference further adjusted for systolic blood pressure,
fasting glucose, and hypertension medications; fasting glucose further adjusted for waist, systolic blood pressure, and hypertension medications.
DISCUSSION
In our family collection, 86% of Caucasian patients
aged ⬎35 years affected with atherogenic dyslipidemia met the ATP-III definition for metabolic syndrome. A similar high association was present in sporadic cases of atherogenic dyslipidemia. Thus, most
middle-aged subjects who have atherogenic dyslipidemia seemingly will have it as a component of the
metabolic syndrome. This supports our contention that
atherogenic dyslipidemia, whether familial or sporadic, is a robust identifier of predisposition for the
syndrome. The strongest association was with abdominal obesity. The weaker concordance between atherogenic dyslipidemia and ATP-III metabolic syndrome
before age 35 years may signify a particularly strong
predisposition to an isolated lipid disorder, but in
many cases, it likely foreshadows the development of
the metabolic syndrome later in life.
The relation between dyslipidemia and systolic
blood pressure was less striking, suggesting a weaker
causal linkage. This corroborates well with published
findings, where a factor analyses investigating metabolic syndrome shows that blood pressure often falls
out as a distinct factor,7 although it often correlates
with obesity.8 Elevated plasma glucose was also not
strongly associated, but here it is known that compensatory hyperinsulinemia can prevent the development
of overt hyperglycemia for many years in persons with
the metabolic syndrome. Finally, it is known that the
metabolic syndrome manifests differently in different
populations, and the essential findings obtained in the
current Caucasian population will not necessarily be
applicable to other populations, notably blacks, Hispanics, and Asians.
Acknowledgment: We thank all the families whose
participation made this project possible. We gratefully
acknowledge the contributions of the study personnel
at each of the collaborating sites: John Farrell,
198 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
Nicholas Nikolopoulos and Maureen Sutton (Boston);
Judy Walshe, Monica Prentice, Anne Whitehouse,
Julie Butters, and Tori Nicholls (Australia); Heather
Doelle, Lynn Lewis, and Anna Toma (Canada); Kari
Kervinen, Seppo Poykko, Liisa Mannermaa, and Sari
Paavola (Finland); Claire Hurrel, Diane Morin, Alice
Mermod, Myriam Genoud, and Roger Darioli
(Switzerland); Guy Pepin, Sibel Tanir, Erhan Palaoglu, Kerem Ozer, Linda Mahley, and Aysen Agacdiken (Turkey); and Deborah A. Widmer, Rhonda
Harris, and Selena Dixon (United States). We also
would like to thank Allen Roses, Shyama Brewster,
and Julia Perry for their support, and Dennis Sprecher,
Annie McNeill, and Sara Ephross for their useful
discussions.
1. National Cholesterol Education Program (NCEP) Expert Panel on detection,
evaluation, and treatment of high blood cholesterol in adults (Adult Treatment
III). Third report of the National Cholesterol Education Program (NCEP) Expert
Panel on detection, evaluation, and treatment of high blood cholesterol in adults
(Adult Treatment Panel III) final report. Circulation 2002;106:3143–3421.
2. Bersot TP, Vega GL, Grundy SM, Palaoglu KE, Atagunduz P, Ozbayrakci S,
Gokdemir O, Mahley RW. Elevated hepatic lipase activity and low levels of high
density lipoprotein in a normotriglyceridemic, nonobese Turkish population.
J Lipid Res 1999;40:432– 438.
3. Langille DB, Joffres MR, MacPherson KM, Andreou P, Kirkland SA, MacLean DR. Prevalence of risk factors for cardiovascular disease in Canadians 55 to
74 years of age: results from the Canadian Heart Health Surveys, 1986 –1992.
Can Med Assoc J 1999;161(suppl 8):S3–S9.
4. The Lipid Research Clinics Program Epidemiology Committee. Plasma lipid
distributions in selected North American populations: the Lipid Research Clinics
Program Prevalence Study. Circulation 1979;60:427– 439.
5. Vartiainen E, Jousilahti P, Alfthan G, Sundvall J, Pietinen P, Puska P.
Cardiovascular risk factor changes in Finland, 1972–1997. Int J Epidemiol
2000;29:49 –56.
6. Mahley RW, Palaoglu KE, Atak Z, Dawson-Pepin J, Langlois AM, Cheung V,
Onat H, Fulks P, Mahley LL, Vakar F. Turkish Heart Study: lipids, lipoproteins,
and apolipoproteins. J Lipid Res 1995;36:839 – 859.
7. Hanley AJ, Karter AJ, Festa A, D’Agostino R Jr, Wagenknecht LE, Savage P,
Tracy RP, Saad MF, Haffner S. Factor analysis of metabolic syndrome using
directly measured insulin sensitivity: The Insulin Resistance Atherosclerosis
Study. Diabetes 2002;51:2642–2647.
8. Meigs JB, D’Agostino RB Sr, Wilson PW, Cupples LA, Nathan DM, Singer
DE. Risk variable clustering in the insulin resistance syndrome. The Framingham
Offspring Study. Diabetes 1997;46:1594 –1600.
JANUARY 15, 2005
New Predictors of Outcome in Idiopathic
Pulmonary Arterial Hypertension
Steven M. Kawut, MD, Evelyn M. Horn, MD, Ketevan K. Berekashvili, MD,
Robert P. Garofano, EdD, Rochelle L. Goldsmith, EdD, Allison C. Widlitz, PA, MS,
Erika B. Rosenzweig, MD, Diane Kerstein, MD, and Robyn J. Barst, MD
Idiopathic pulmonary arterial hypertension (PAH) is a
rare disease with a poor prognosis. New therapies
have improved the outcome of this condition; accordingly, the factors that determine outcome may have
changed. We aimed to identify determinants of survival in a cohort of consecutive patients with PAH:
which was idiopathic, familial, or associated with
anorexigen use. We performed a retrospective cohort
study of 84 consecutive patients with PAH who underwent initial evaluation at our center from January
1994 to June 2002. The primary outcome was death
or lung transplantation. Survival at 1, 2, and 3 years
was 87%, 75%, and 61%, respectively. Multivariate
analysis showed that being of African-American or
Asian descent was associated with an increased risk
of death. Warfarin use was associated with a reduced
risk of death. Higher serum albumin and cardiac index and acute vasoreactivity were independently associated with improved survival. These data suggest
that the determinants of outcome have changed. Race
is identified as a new risk factor, which may be attributable to biologic or socioeconomic differences.
Cardiac function and acute reactivity of the pulmonary vascular bed remain strong independent predictors of outcome. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:199 –203)
he National Institutes of Health (NIH) Patient Registry for the Characterization of Primary Pulmonary
T
Hypertension found that right-sided heart function (as-
PAH who underwent initial evaluation from January
1994 to June 2002 with follow-up through June 2003
at our center. The study was approved by the institutional review board.
Study subjects: To assemble the cohort, we queried
the Data Warehouse at our institution for all patients
with an International Classification of Diseases-Ninth
Revision code for primary or secondary pulmonary
hypertension who were assessed by clinicians at our
center. Second, we hand searched all outpatient
records from our center. Last, we reviewed a comprehensive list of all patients from our center who underwent cardiac catheterization since 1994.
The following criteria were required for inclusion
in the study: (1) PAH that was idiopathic, familial, or
associated with anorexigen use5; (2) age ⬎16 years;
and (3) initial evaluation performed at our center
between January 1994 and July 2002. The exclusion
criteria were (1) previous cardiac catheterization with
acute vasodilator study and initiation of PAH therapy
and (2) other forms of PAH. Usual clinical and laboratory criteria were used to rule out other diseases
associated with PAH (e.g., systemic sclerosis and
systemic lupus erythematosus).
Measurements: Patients routinely underwent laboratory testing, pulmonary function testing, radionuclide angiography,6 echocardiography, cardiopulmonary exercise testing, and cardiac catheterization at
initial evaluation. During cardiac catheterization,
short-acting vasodilators, such as inhaled nitric oxide
or intravenous epoprostenol, were administered to test
acute vasoreactivity. Acute vasoreactivity was defined
as a reduction in mean pulmonary artery pressure of
⬎20% without a decrease in cardiac index. Exercise
testing, cardiac catheterization, or vasodilator testing
were not performed if deemed unsafe (e.g., severe
symptoms at rest).
sessed by right atrial pressure and cardiac index) and
severity of pulmonary arterial hypertension (PAH) (assessed by pulmonary artery pressure) were associated
with survival.1 Since the NIH Registry, clinical trials
have demonstrated the effectiveness of several therapies
in idiopathic PAH: continuous intravenous epoprostenol,
continuous subcutaneous treprostinil, and oral bosentan
have been approved for treatment of PAH.2– 4 In light of
these new therapies, it is possible that the prognosis of
PAH has been altered. Accordingly, the determinants of
survival may have also changed. If so, the NIH Registry
prediction rule may no longer provide accurate prognostic information. Our first aim was to identify the determinants of survival in idiopathic PAH in the recent
therapeutic era. Our second aim was to reevaluate the
accuracy and generalizability of the NIH prediction rule
for patients with idiopathic PAH diagnosed during the
last decade.
METHODS
Study design: We performed a retrospective cohort
study of all consecutive adult patients with idiopathic
From the Departments of Internal Medicine and Pediatrics, College of
Physicians and Surgeons, and Department of Epidemiology, Joseph L.
Mailman School of Public Health, Columbia University, New York,
New York. Dr. Kawut was supported by Grant HL67771 from the
National Institutes of Health, Bethesda, Maryland, and the Florence
and Herbert Irving Clinical Research Career Award, New York, New
York. Manuscript received July 21, 2004; revised manuscript received
and accepted September 8, 2004.
Address for reprints: Steven Kawut, MD, Division of Pulmonary,
Allergy, and Critical Care Medicine, PH 8E, Room 101, 622 West
168th Street, New York, New York 10032. E-mail: sk2097@
columbia.edu.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.006
199
TABLE 1 Baseline Laboratory (n ⫽ 84), Pulmonary Function
Tests (n ⫽ 84), and Cardiopulmonary Exercise Tests (n ⫽
72) Data
Characteristic
Blood urea nitrogen (mg/dl)
Creatinine (mg/dl)
White blood cell count (10 · 9/l)
Hemoglobin (g/dl)
Hematocrit (%)
Mean corpuscular volume (fl)
Platelets (10 · 9/l)
Total bilirubin (mg/dl)*
Direct bilirubin (mg/dl)*
Alkaline phosphatase (U/L)*
Alanine aminotransferase (U/L)*
Aspartate aminotransferase (U/L)*
Albumin (g/dl)*
Prothrombin time (s)†
Positive antinuclear antibody‡
Forced vital capacity (% predicted)
Forced vital capacity (L)
Forced expiratory volume in 1 s
(% predicted)
Forced expiratory volume in 1 s (L)
Total lung capacity (% predicted)§
Total lung capacity (L)§
Diffusion capacity for carbon monoxide
adjusted for hemoglobin (ml/min/
mm Hg)‡
Diffusion capacity for carbon monoxide
adjusted for hemoglobin (%
predicted)‡
Peak heart rate (beats/min)
Peak systolic blood pressure (mm Hg)储
Peak diastolic blood pressure (mm Hg)储
Peak oxygen consumption (ml/kg/min)
Peak oxygen pulse (ml/beat)
Peak respiratory exchange ratio
Peak ventilatory equivalent for carbon
dioxide
Peak minute ventilation (L/min)
TABLE 2 Hemodynamic Parameters at Baseline and With
Acute Vasodilator Testing
Value
17 (14–20)
0.9 (0.8–1.1)
7.8 (6.8–9.6)
14.8 (14–16)
43.9 (41.6–47.6)
88.6 (86.5–92.4)
212 (178–262)
0.9 (0.7–1.3)
0.2 (0.1–0.3)
84 (62–111)
23 (17–32)
24.5 (20–32)
4.5 (4.2–4.9)
13.4 (12.7–14.3)
16 (20%)
90 ⫾ 20
3.1 ⫾ 1.0
86 ⫾ 18
2.4
93
4.8
16
⫾
⫾
⫾
⫾
0.7
16
1.1
6
63 ⫾ 22
137
133
77
11.4
6.1
1.1
56
⫾
⫾
⫾
⫾
⫾
⫾
⫾
21
26
13
4
2.5
0.1
15
47 ⫾ 17
Values are expressed as mean ⫾ SD, median (interquartile range) or number
(%).
*n ⫽ 82, †n ⫽ 81, ‡n ⫽ 80, §n ⫽ 70, 储n ⫽ 71.
The choice of therapy adhered to usual algorithms.7
Patients with acute vasoreactivity were treated with calcium channel blockers. Nonreactive patients were treated
with bosentan (when available) for New York Heart
Association (NYHA) class III and IV symptoms, treprostinil (when available) for NYHA III or IV symptoms,
and intravenous epoprostenol for NYHA III symptoms
(before the availability of bosentan and treprostinil) and
NYHA IV symptoms. Although certain therapies were
available at different times during the study period, we
adjusted for these differences in our analysis.
The primary combined end point was death or lung
transplantation. We assessed outcomes through chart
review and query of the Social Security Death Index.
A patient was censored as alive at the last medical
contact recorded in the computerized clinical database
through June 2003.
Statistical analysis: Continuous variables were summarized by mean ⫾ SD or median (interquartile
range). Categorical variables were summarized by frequencies with 95% confidence intervals (CIs). Sur200 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
Variable
Heart rate (beats/min)
Mean pulmonary artery
pressure (mm Hg)
Mean right atrial pressure
(mm Hg)
Cardiac index (L/min/m2)
Pulmonary vascular resistance
index (U · m2)
Mixed venous oxygen
saturation (%)
Baseline
(n ⫽ 80)*
With Acute
Vasodilator
(n ⫽ 74)
82 (74–93)
55 (48–61)
81 (67–92)
52 (42–58)
8 (5–12)
8 (5–12)†
1.8 (1.5–2.4)
27 (16–33)
2.1 (1.7–2.5)
21 (13–29)
60 (50–66)
64 (57–69)
Values are expressed as median (interquartile range).
*Baseline hemodynamics were similar for patients with acute vasodilator
studies.
†
n ⫽ 60.
vival analyses were performed using Cox proportional-hazards models.8 We constructed univariate models
for each variable. Therapies were modeled using timevarying covariates, which allow patient-time to be
accurately classified according to treatment status. For
example, we included patient-time while receiving
various therapies (e.g., warfarin, calcium channel
blockers, epoprostenol, bosentan, and/or treprostinil),
while also including patient-time before initiation and
after discontinuation of these therapies. This statistical
method allowed us to control for the availability and
use of different therapies in patients at different times
over the course of the study.
Multivariate models were constructed using covariates with a p value ⬍0.20 on univariate analysis or
that were believed to be clinically important. Likelihood ratio testing was used to identify the most parsimonious model. The variables in the final multivariate model met the assumption of proportional
hazards.9
We also assessed the accuracy of the NIH Registry
prediction rule.10 “Discrimination” describes the ability
of the rule to distinguish patients who die from patients
who live and is expressed by the area under the receiveroperating characteristic curve (AUC). The AUC ranges
from 0.5 (for chance) to 1.0 (for perfect prediction).
“Calibration” is the ability of the rule to accurately predict the probability of death for a patient. To assess
calibration, we divided the cohort into quintiles based on
the predicted probability of death or transplant and calculated the Hosmer-Lemeshow goodness-of-fit statistic.
A p value ⬎0.05 indicated adequate prediction.
Univariate analyses were performed with available
data. We performed simple imputation for missing
data points in the multivariable analysis. We performed sensitivity analyses using mean values and the
“worst” value in the data set in place of the missing
values for the multivariate analysis. A p value ⬍0.05
was considered statistically significant. Stata, version
7.0 (College Station, Texas), was used for all analyses.
JANUARY 15, 2005
RESULTS
Our cohort was composed of 84 newly diagnosed
patients with PAH (mean age 42 ⫾ 14 years).
Sixty-eight (81%) were women. The cohort included 69 whites (82%, of whom 9 were Hispanic
and 60 were non-Hispanic) and 15 non-whites
(18%, 9 Asians and 6 African-Americans). Sixtysix patients (78%) had idiopathic PAH, 14 (17%)
had familial PAH, and 4 (5%) had PAH related to
anorexigen use. The median time between the performance of echocardiography and/or cardiac catheterization consistent with a diagnosis of pulmonary
hypertension and evaluation at our center was 181
days (interquartile range 91 to 374). The median
duration of symptoms before our evaluation was
584 days (interquartile range 251 to 1,392; n ⫽ 81).
Seventy-nine patients (94%) were treated with warfarin and 72 (86%) with digoxin at baseline or after
evaluation. Thirty-three patients (39%) received
oral calcium channel blockers and 61 (73%) received spironolactone. Thirty-eight patients (45%)
were treated with intravenous epoprostenol, 31
within 60 days of the baseline catheterization.
Twelve patients (14%) were treated with subcutaneous treprostinil and 23 (27%) with oral bosentan.
Eleven patients (13%) received combination therapy with bosentan and either intravenous epoprostenol or subcutaneous treprostinil.
Laboratory, pulmonary function, cardiopulmonary
exercise test, and hemodynamic data are listed in
Tables 1 and 2. Four patients did not undergo rightsided cardiac catheterization due to severity of illness
and safety concerns; 10 patients did not undergo vasodilator testing. Echocardiography showed a pericardial effusion in 28 of 83 patients (34%). The peak
tricuspid regurgitant gradient was 83 ⫾ 24 mm Hg
(n ⫽ 74). Right ventricular ejection fraction on radionuclide angiography was 30 ⫾ 9% (n ⫽ 65, normal
ⱖ45%).11
Survival: The median follow-up was 764 days (interquartile range 505 to 1,573). Twenty-four patients
died and 1 patient underwent lung transplantation.
Transplant-free survival at 1 year was 87% (95% CI
77% to 93%), at 3 years 75% (95% CI 63% to 84%),
and at 5 years 61% (95% CI 45% to 73%) (Figure 1).
Univariate survival analyses: We found an increased
risk of death for non-whites (Asians and AfricanAmericans) in our cohort (Table 3). Hepatic and renal
dysfunction were associated with an increased risk of
death. The presence of antinuclear antibodies (titer
ⱖ1:40) was not associated with outcome (hazard ratio
[HR] 1.6, 95% CI 0.63 to 4.1, p ⫽ 0.32, n ⫽ 80). The
time of patient evaluation was not associated with
survival (data not shown).
Lower systolic blood pressure and higher ventilatory equivalent for carbon dioxide at peak exercise
were associated with worse outcome (p ⬍0.05); however, other parameters from cardiopulmonary exercise
testing were not (data not shown). A pericardial effusion on echocardiography more than doubled the risk
of death (HR 2.6, 95% CI 1.2 to 5.9, p ⫽ 0.017),
whereas the peak tricuspid regurgitant gradient was
FIGURE 1. Kaplan-Meier estimate of transplant-free survival in
patients with idiopathic pulmonary arterial hypertension.
not associated with outcome (HR 1.0, 95% CI 0.98 to
1.02, p ⫽ 0.83). A 5% increase in right ventricular
ejection fraction on radionuclide angiography was associated with a reduced risk of death (HR 0.62, 95%
CI 0.43 to 0.89, p ⫽ 0.009).
Multivariate survival analysis: Asian or AfricanAmerican race, low serum albumin, lack of warfarin
use, low cardiac index, and absence of acute vasoreactivity were independently associated with an increased risk of death (Table 4). Replacement of cardiac index with right ventricular ejection fraction in
the model showed similar results. Repeat analysis of
the cohort without potentially influential and/or outlying subjects did not change the conclusions, supporting the robustness of the final multivariate model.
The effect estimates from the sensitivity analyses for
missing data were not substantially different from
those using simple imputation (data not shown).
Assessment of the NIH prediction rule: The NIH
prediction rule was poorly calibrated to predict the
probability of 1-, 2-, and 3-year survival in our cohort
(p ⬍0.005). Reclassifying the 1 patient who underwent lung transplantation as alive did not change the
results. Although better than chance alone, the NIH
prediction rule was not sufficiently discriminating between patients who died and those who were alive at
each time period (1-year AUC 0.76 [95% CI 0.59 to
0.93], 2-year AUC 0.70 [95% CI 0.55 to 0.86], and
3-year AUC 0.72 [95% CI 0.57 to 0.88]).
DISCUSSION
This is the largest study of outcomes in patients
with idiopathic PAH treated with currently approved
PAH therapies. We have shown that cardiac index,
acute vasoreactivity, race, and serum albumin independently predict transplant-free survival in patients
with idiopathic PAH. Although warfarin use was associated with increased survival, the use of other PAH
medications was not. These results differ from those
of recently published studies.12–15 Our data also demMISCELLANEOUS/NEW PREDICTORS OF OUTCOME IN PAH
201
relation to outcome. Although rightsided cardiac function continues to
Variable
HR
95% CI
p Value
be an important factor in determining
survival, the currently approved theraAge
0.99
0.96–1.0
0.44
peutic agents may have “uncoupled”
Gender: men
1.0
0.40–2.6
0.98
Body mass index
0.96
0.90–1.0
0.29
the former association between seRace
verity of PAH and outcome.
Asian or African-American
3.2
1.3–7.6
0.008
The purpose of acute vasoreactivWhite
1.0
—
—
ity testing in PAH is to predict the
Diagnosis
Idiopathic
1.0
—
—
response to chronic oral calcium
Familial
1.6
0.58–4.2
0.37
channel blockade.20,21 Recently, inRelated to anorexigens
2.3
0.53–10
0.27
vestigators
have questioned the vaWarfarin
0.33
0.12–0.90
0.03
lidity of using only a 20% decrease
Digoxin
0.57
0.23–1.4
0.24
in mean pulmonary artery pressure in
Spironolactone
2.1
0.87–5.1
0.10
Calcium channel blockers
0.34
0.08–1.4
0.14
defining acute vasoreactivity, beEpoprostenol
1.4
0.62–3.1
0.42
cause near normalization of pulmoBosentan
0.64
0.15–2.7
0.54
nary artery pressures is a better preTreprostinil
0.64
0.15–2.7
0.54
dictor of the long-term response to
Combination therapy
0.75
0.10–5.6
0.78
Blood urea nitrogen
1.05
1.0–1.1
0.04
chronic calcium channel blockade.22
Creatinine
6.1
1.7–22
0.006
However, our data suggest that paHematocrit
1.0
0.95–1.1
0.41
tients with an acute response to vaPlatelet count
1.0
0.99–1.0
0.28
sodilator testing have an improved
Prothrombin time
1.1
1.0–1.1
0.025
survival compared with patients who
Total bilirubin
1.9
1.2–2.8
0.003
Direct bilirubin
15.3
3.7–63
0.001
are nonresponders, irrespective of
Albumin
0.46
0.19–1.11
0.09
treatment. This suggests that acute
Alanine aminotransaminase
1.01
1.0–1.02
0.13
vasoreactivity has implications reAspartate transaminase
1.02
1.0–1.03
0.055
garding survival beyond the theraAlkaline phosphatase
1.01
1.0–1.02
0.006
Forced vital capacity % predicted
0.97
0.95–0.99
0.015
peutic response to calcium channel
Forced expiratory volume in 1-s %
0.97
0.95–0.99
0.035
blockers.
predicted
Although we did not find signifiTotal lung capacity % predicted
0.96
0.93–0.99
0.025
cant associations between the variDiffusion capacity for carbon dioxide
1.0
0.99–1.03
0.54
ous PAH therapies and survival,
adjusted for hemoglobin %
predicted
these results should be interpreted
Heart rate
1.06
1.02–1.1
0.005
with caution. Confounding by disMean pulmonary artery pressure
1.02
0.98–1.05
0.29
ease severity may bias the associaMean right atrial pressure
1.05
0.99–1.1
0.09
tions between therapies and outCardiac index
0.36
0.17–0.76
0.008
Pulmonary vascular resistance index
1.03
1.01–1.06
0.005
comes. For example, sicker patients
Mixed venous oxygen saturation
0.94
0.90–0.98
0.003
may receive intravenous epoprosteAcute vasoreactivity
0.11
0.01–0.81
0.03
nol earlier than healthier patients.
Therefore, epoprostenol may appear
to be associated with worse outcomes in a cohort study, despite definitive evidence
TABLE 4 Multivariate Analysis of Transplant-free Survival
that epoprostenol improves survival.4 Accordingly, a
Variable
HR
95% CI
p Value
cohort study cannot estimate the clinical effects of
therapies. However, we can conclude that adjustment
Asian or African-American
4.3
1.7–11
0.002
for differing therapeutic regimens did not affect the
race
Warfarin use
0.35
0.12–0.99
0.05
main determinants of survival in the final multivariate
Albumin
0.37
0.16–0.84
0.019
model.
Cardiac index
0.41
0.19–0.90
0.026
We found an association between warfarin use and
Acute vasoreactivity
0.13
0.02–0.96
0.046
improved survival, consistent with other nonrandomized studies.20,23–25 Anticoagulation with warfarin
may prevent in situ thrombosis, a frequent finding in
onstrate that right ventricular ejection fraction on ra- the pulmonary vasculature in PAH, and venous thromdionuclide angiography is associated with survival in boembolism. However, anticoagulation is withheld in
PAH. This could be a useful clinical parameter for patients with previous bleeding episodes and/or severe
evaluating disease severity and a potential surrogate thrombocytopenia, so that confounding by disease severity may explain these findings. A randomized conend point for PAH clinical trials.16 –19
The degree of PAH and right-sided cardiac dys- trolled trial of anticoagulation in PAH may therefore
function have been considered the most significant be warranted.
Traditionally, liver function test abnormalities in
factors in the disease course.1 We found that cardiac
index and right ventricular ejection fraction predicted PAH have been believed to merely reflect increased right
survival, whereas pulmonary artery pressure had no atrial pressure. However, we identified low albumin as a
TABLE 3 Univariate Analyses of Transplant-free Survival
202 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
risk factor for mortality, independent of right atrial pressure or cardiac index. Although a low albumin may
result from malnutrition or chronic losses, these findings
highlight the need for further investigation of hepatic
dysfunction in patients with PAH.
To our knowledge, our study is the first to suggest a difference in survival in PAH based on
race.1,12,14,26,27 Although 1 study found an increased
age-specific mortality for African-American women
with PAH, the diagnosis was based on International
Classification of Diseases-Ninth Revision coding
without clinical data, and actual survival was not
assessed.28 Minority patients may be identified later
in their disease course; however, we did not find
associations between race and duration of symptoms, referral time, or hemodynamics. Our findings
are unfortunately consistent with studies of health
disparities in other cardiovascular diseases and
should spur outreach to “at-risk” patients and the
physicians who serve them.29
Despite similar demographics and possibly lower
mean cardiac index in our cohort compared with the
NIH Registry, the NIH prediction rule underestimated
the probability of survival in each quintile of risk in
our cohort (data not shown), likely attributable to
advances in treatment. Because adequate calibration
of a predictive index is most important for counseling
patients regarding their prognosis,10 the NIH rule may
no longer be sufficient for that task.
These findings require confirmation in other centers. Although there were missing data for some variables, sensitivity analyses based on different imputation methods did not change our conclusions. Finally,
the small number of events limited the precision of
certain analyses; however, we did not find significant
model “overfitting.”
In conclusion, cardiac function and acute reactivity
of the pulmonary vascular bed at diagnosis remain
strong predictors of outcome in PAH regardless of
which therapy is initiated. Asian or African-American
race and low albumin are new independent risk factors. Right ventricular ejection fraction from radionuclide angiography predicts long-term outcome in
PAH. Future studies should derive and validate new
prediction tools for patients with PAH diagnosed and
treated in the 21st century.
1. D’Alonzo GE, Barst RJ, Ayres SM, Bergofsky EH, Brundage BH, Detre KM,
Fishman AP, Goldring RM, Groves BM, Kernis JT, et al. Survival in patients
with primary pulmonary hypertension. Results from a national prospective registry. Ann Intern Med 1991;115:343–349.
2. Rubin LJ, Badesch DB, Barst RJ, Galie N, Black CM, Keogh A, Pulido T,
Frost A, Roux S, Leconte I, Landzberg M, Simonneau G. Bosentan therapy for
pulmonary arterial hypertension. N Engl J Med 2002;346:896 –903.
3. Simonneau G, Barst RJ, Galie N, Naeije R, Rich S, Bourge RC, Keogh A,
Oudiz R, Frost A, Blackburn SD, Crow JW, Rubin LJ. Continuous subcutaneous
infusion of treprostinil, a prostacyclin analogue, in patients with pulmonary
arterial hypertension: a double-blind, randomized, placebo-controlled trial. Am J
Respir Crit Care Med 2002;165:800 – 804.
4. Barst RJ, Rubin LJ, Long WA, McGoon MD, Rich S, Badesch DB, Groves
BM, Tapson VF, Bourge RC, et al. A comparison of continuous intravenous
epoprostenol with conventional therapy for primary pulmonary hypertension.
N Engl J Med 1996;334:296 –302.
5. Simonneau G, Galie N, Rubin LJ, Langleben D, Seeger W, Domenighetti G,
Gibbs S, Lebrec D, Speich R, Beghetti M, Rich S, Fishman A. Clinical classification of pulmonary hypertension. J Am Coll Cardiol 2004;43:5S–12S.
6. Nichols K, Saouaf R, Ababneh AA, Barst RJ, Rosenbaum MS, Groch MW,
Shoyeb AH, Bergmann SR. Validation of SPECT equilibrium radionuclide
angiographic right ventricular parameters by cardiac magnetic resonance imaging. J Nucl Cardiol 2002;9:153–160.
7. Galie N, Seeger W, Naeije R, Simonneau G, Rubin LJ. Comparative analysis
of clinical trials and evidence-based treatment algorithm in pulmonary arterial
hypertension. J Am Coll Cardiol 2004;(suppl)43:81S– 88S.
8. Cox DR. Regression models and life tables (with discussion). J Royal Stat Soc
1972;34:187–220.
9. Grambsch PM, Therneau TM. Proportional hazards tests and diagnostics based
on weighted residuals. Biometrika 1994;81:515–526.
10. Justice AC, Covinsky KE, Berlin JA. Assessing the generalizability of
prognostic information. Ann Intern Med 1999;130:515–524.
11. Berger HJ, Matthay RA, Loke J, Marshall RC, Gottschalk A, Zaret BL.
Assessment of cardiac performance with quantitative radionuclide angiocardiography: right ventricular ejection fraction with reference to findings in chronic
obstructive pulmonary disease. Am J Cardiol 1978;41:897–905.
12. McLaughlin VV, Shillington A, Rich S. Survival in primary pulmonary hypertension: the impact of epoprostenol therapy. Circulation 2002;106:1477–1482.
13. Sitbon O, Humbert M, Nunes H, Parent F, Garcia G, Herve P, Rainisio M,
Simonneau G. Long-term intravenous epoprostenol infusion in primary pulmonary hypertension: prognostic factors and survival. J Am Coll Cardiol 2002;40:
780 –788.
14. Kuhn KP, Byrne DW, Arbogast PG, Doyle TP, Loyd JE, Robbins IM.
Outcome in 91 consecutive patients with pulmonary arterial hypertension receiving epoprostenol. Am J Respir Crit Care Med 2003;167:580 –586.
15. Wensel R, Opitz CF, Anker SD, Winkler J, Hoffken G, Kleber FX, Sharma
R, Hummel M, Hetzer R, Ewert R. Assessment of survival in patients with
primary pulmonary hypertension: importance of cardiopulmonary exercise testing. Circulation 2002;106:319 –324.
16. Kawut SM, Palevsky HI. Surrogate endpoints in pulmonary arterial hypertension. Am Heart J 2004;148:559 –565.
17. Kawut SM, Palevsky HI. New answers raise new questions in pulmonary
arterial hypertension. Eur Respir J 2004;23:799 – 801.
18. Peacock A, Naeije R, Galie N, Reeves JT. End points in pulmonary arterial
hypertension: the way forward. Eur Respir J 2004;23:947–953.
19. Hoeper MM, Oudiz RJ, Peacock A, Tapson VF, Haworth SG, Frost AE,
Torbicki A. End points and clinical trial designs in pulmonary arterial hypertension: clinical and regulatory perspectives. J Am Coll Cardiol 2004;43(suppl):
48S–55S.
20. Rich S, Kaufmann E, Levy PS. The effect of high doses of calcium-channel
blockers on survival in primary pulmonary hypertension. N Engl J Med 1992;
327:76 – 81.
21. Weir EK, Rubin LJ, Ayres SM, Bergofsky EH, Brundage BH, Detre KM,
Elliott CG, Fishman AP, Goldring RM, Groves BM, et al. The acute administration of vasodilators in primary pulmonary hypertension. Experience from the
National Institutes of Health Registry on Primary Pulmonary Hypertension. Am
Rev Respir Dis 1989;140:1623–1630.
22. Sitbon O, Humbert M, Ioos V, Jais X, Parent F, Garcia G, Herve P,
Simonneau G. Who benefits from long-term calcium-channel blocker therapy in
primary pulmonary hypertension? (abstr). Am J Respir Crit Care Med 2004;167:
A440.
23. Fuster V, Steele PM, Edwards WD, Gersh BJ, McGoon MD, Frye RL.
Primary pulmonary hypertension: natural history and the importance of thrombosis. Circulation 1984;70:580 –587.
24. Cohen M, Fuster V, Edwards WD. Anticoagulation in the treatment of
pulmonary hypertension. In: Fishman AP, ed. The Pulmonary Circulation: Normal and Abnormal. Philadelphia, PA: University of Pennsylvania Press, 1990:
501–510.
25. Frank H, Mlczoch J, Huber K, Schuster E, Gurtner HP, Kneussl M. The effect
of anticoagulant therapy in primary and anorectic drug-induced pulmonary hypertension. Chest 1997;112:714 –721.
26. Kawut SM, Taichman DB, Archer-Chicko CL, Palevsky HI, Kimmel SE.
Hemodynamics and survival in patients with pulmonary arterial hypertension
related to systemic sclerosis. Chest 2003;123:344 –350.
27. Bossone E, Paciocco G, Iarussi D, Agretto A, Iacono A, Gillespie BW,
Rubenfire M. The prognostic role of the ECG in primary pulmonary hypertension. Chest 2002;121:513–518.
28. Lilienfeld DE, Rubin LJ. Mortality from primary pulmonary hypertension in
the United States, 1979 –1996. Chest 2000;117:796 – 800.
29. Cooper R, Cutler J, Desvigne-Nickens P, Fortmann SP, Friedman L, Havlik
R, Hogelin G, Marler J, McGovern P, Morosco G, et al. Trends and disparities in
coronary heart disease, stroke, and other cardiovascular diseases in the United
States: findings of the national conference on cardiovascular disease prevention.
Circulation 2000;102:3137–3147.
MISCELLANEOUS/NEW PREDICTORS OF OUTCOME IN PAH
203
ANTHONY NICHOLAS DeMARIA, MD:
A Conversation With the Editor*
ony DeMaria (Figure 1) was born in Elizabeth,
New Jersey on January 12, 1943, and grew up in
T
Bayonne, New Jersey. After public and parochial
schools, he went to the College of the Holy Cross,
graduating in 1964, and then to the New Jersey College of Medicine, graduating in 1968. His internship
in internal medicine was at the St. Vincent Hospital in
Worcester, Massachusetts, and his junior and senior
medical residency was at the US Public Health Service Hospital in Staten Island, New York. His fellowship in cardiovascular medicine was at the University
of California at Davis. Following completion of his
fellowship in 1973, he remained on the faculty until
June 1981, when he moved to Lexington, Kentucky,
to be Professor of Medicine and Chief of the cardiovascular division of the University of Kentucky College of Medicine. In May 1992 he returned to California, this time to San Diego, as Professor of
Medicine and Chief of the cardiovascular division of
the University of California at San Diego. In October
2001, he was made the Judith and Jack White Chair in
Cardiology. Dr. DeMaria has been, for nearly 3 decades, an international authority in cardiology, particularly in the area of echocardiography. His investigations have led to the publication of ⬎450 articles,
almost all in peer-reviewed medical journals. He is
also the author or editor of 3 books. For his investigative work and teaching abilities, he has received
many honors, including several Internal Medicine Distinguished Faculty Teaching Awards, Teaching Scholar
of the American Heart Association, Distinguished Alumnus Award of the University of Medicine & Dentistry
of New Jersey, and honorary degrees from the Kagawa Medical University and the University of Bordeaux, among others. He was president of the American College of Cardiology in 1988 and president of
the American Society of Echocardiography from 1985
to 1987. He became editor in chief of the Journal of
the American College of Cardiology in January 2002.
He has lectured widely, both in the USA and abroad.
He and his lovely wife Lori are the proud parents of 3
offspring, Christine and Anthony who are now married, and Jonathan. Both Tony and Lori are wonderful
people and fun to be around.
William C. Roberts, MD† (hereafter WCR): Dr. DeMaria, I appreciate your willingness to talk to me and
therefore to the readers of The American Journal of
Cardiology. We’re onboard the “Silver Whisper” ship
traveling around the Italian peninsula. Could we start
by my asking you to discuss your parents and siblings,
early upbringing, and some early memories?
*This series of interviews is underwritten by an unrestricted grant from
Bristol-Myers Squibb.
†
Executive Director, Baylor Cardiovascular Institute, Baylor University
Medical Center, Dallas, Texas, 75246.
204
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
FIGURE 1. Anthony Nicholas DeMaria, MD, during the interveiw
(photo by WCR).
Anthony Nicholas DeMaria, MD‡ (hereafter AND): I
was born in Elizabeth, New Jersey, but I actually grew
up in a town not far away, called Bayonne, New
Jersey, which like Jersey City and Hoboken, is in the
New York metropolitan area. My dad owned a fruit
and vegetable market. He went to the farmer’s market
very early 5 days a week for produce. He worked
extraordinarily long hours, opening the store at 6 A.M.
after coming back from the farmer’s market, and he
closed it at 10 P.M. 7 days a week. He made his own
sausage. During Christmas time, we sold Christmas
trees. It was in the selling of Christmas trees that I
learned that the price for anything was variable and
subject to negotiation. We sold flowers at Easter and
for Mother’s Day. I worked in the store with my dad.
My mom was a homemaker (Figure 2). She took care
of my 3 younger sisters and me. Bayonne was, and is,
a largely an ethnic community. If I knew somebody’s
address, I could pretty much tell what their nationality
was and what church they went to. I went through the
local schools and then went off to college in Worcester, Massachusetts.
‡
Director, Cardiovascular Center, Judith and Jack White Chair in
Cardiology Professor of Medicine, University of California, San
Diego, California, 92103-8411.
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.10.004
FIGURE 3. AND’s father’s grocery store.
FIGURE 2. AND and his mother and sister.
WCR: How big was Bayonne?
AND: Bayonne had a population of 60,000 packed
into a peninsula that was 3 miles long and at its very
widest, about 1 mile wide. On one end it was connected to Staten Island, New York, via the Bayonne
Bridge, and on the other end, it continued to Jersey
City. We could almost see the Statue of Liberty from
my house.
WCR: Did you get to Manhattan or Staten Island
very much?
AND: Not very much. The Verrazano Narrows
Bridge, which connects Staten Island to Brooklyn,
hadn’t been built yet. It was alleged (and I’m sure it’s
true) that there were people who had been born on
Staten Island that never in their lives had left the
island, even for 1 day. Going to New York City was
an adventure. The traffic was worse, of course, than in
Bayonne. The people were a little different. We
mostly stayed on the Jersey side.
WCR: You are the oldest of 4 siblings. You were
born in 1943. What’s the difference in your siblings’
ages?
AND: Each of us was born roughly 2 years apart.
We spanned 8 to 10 years between the 4 of us.
WCR: What was your daddy’s store like?
AND: It was very small and fascinating. I have a
picture of it in my office (Figure 3). When my father
died and we went through his things, we found a
crumpled up photograph of the store. I had to smooth
it out to have it made into a nice big print. The whole
store couldn’t have been 1,000 square feet. It was
mostly fruits and vegetables with canned goods on 1
side and a big freezer in the corner at the back. It
contained a little delicatessen, where we sliced lunch
meats. His major business was selling to people who
wanted very fresh fruits and vegetables or alternatively, to people who wanted delivery. The latter is
what I did most of the time. Customers would call in
orders (a pound of this and a sack of sugar) and we’d
put together the order and then I’d deliver it to their
houses. It was pretty convenient.
WCR: On a bicycle?
AND: Remember those bicycles with the small
wheels in front? The rear wheel was the usual size and
the front wheel was very small with a huge basket
above it. I could usually fit 3 or 4 orders in the basket
and deliver them. It was a marvelous education in life.
My dad died prematurely at 59 years of age from an
aortic dissection. My mom is still alive and my sisters
Barbara, Carol, and Adele are doing well.
WCR: When was your father born?
AND: He was born in 1916 and died in 1976.
WCR: He worked up to the day he died?
AND: My father had 2 jobs at the time he died. The
store had burned down. The building that the store
was in was very old. We were home on a Sunday and
somebody came by and told us the store was on fire.
The building had kerosene heaters. This was firstgeneration America. It was not the suburbs. Because
my father had records in the store of who owed him
money, he ran into the building. I ran to the building,
a judgment that compounded the stupidity of the moment, but I convinced him to get out; shortly thereafter
the building went up in a blaze.
WCR: Neither of you got burned?
AND: Neither one of us got hurt. We got away in
time.
WCR: You were how old then?
AND: I was 23. My dad then went to work for the
city of Bayonne. He was in charge of all their parking
meters and helped with zoning issues, etc. He’d do
that all day, and then in the evening he went to some
truck terminals in Jersey City and loaded trucks until
about 2, and then come home and go to bed.(Jersey
City was a port.) He died of aortic dissection while
loading a truck.
WCR: Did your father smoke?
INTERVIEW/ANTHONY NICHOLAS DEMARIA
205
AND: Yes, cigars, continuously. I think he was
born with a cigar in his mouth. He’d send me down to
the store to get his cigars. I’d get Dutch Master’s
panatelas. That’s all he smoked.
WCR: How far was the store from your home?
AND: A block or maybe a block and a half.
WCR: What was your home like?
AND: My home was pretty good. We were, by the
standards of our neighborhood, well off. We had our
own house. It was a small split-level house. My father
had it built. There was a lot in the area that was empty.
He and my uncle purchased the lot and then contracted
out the building of 2 houses next to one another.
WCR: Did you have a room of your own?
AND: I did. It was in the attic. The house had 2
bedrooms and an attic. My parents slept in one and my
3 sisters had the master bedroom. That was the only
one all 3 could fit in. My parents slept in a smaller
bedroom. My dad, who was very handy, put a floor in
the attic and plastered the ceiling. You could only
stand up in the middle of the room because the roof
slanted. But it was mine! I was the only one living in
it! It was a different time, Bill. My folks thought it was
important for a son to go to college. That was
preached to me from day 1. They repeated over and
over again that I needed to go to college so that I
didn’t wind up running a fruit and vegetable market.
But my sisters didn’t seem to matter so much. If they
went to college that was nice; it was icing on the cake.
For the first born son, it was absolutely necessary.
WCR: Were your parents first-generation Americans?
AND: Yes. Both my parents were born in the USA.
Their parents had just come over. I never got to know
my grandparents because I couldn’t speak their language.
WCR: Where were they from?
AND: On my father’s side, from Italy (the Naples
area), and on my mother’s side, Poland. They couldn’t
speak to one another.
WCR: How did they meet?
AND: They met at one of the beaches in Jersey. My
father was black-haired with a darker complexion. My
mother was blonde with very fair skin. I guess opposites attract. My mother was from Elizabeth, New
Jersey. That is why I was born there. My dad was from
Bayonne. I was only born in Elizabeth. We always
lived in Bayonne.
WCR: How far is Bayonne from Elizabeth?
AND: Two stops on the Jersey Turnpike, maybe 10
miles.
WCR: Where does your mother live now?
AND: She lives in the same house, in Bayonne,
New Jersey. My uncle still lives next door. Two other
uncles, my father’s brothers, lived across the street
until they died. The whole family on my dad’s side
lived in that area.
WCR: You had a lot of cousins and family?
AND: Tons of cousins. There were lots and lots of
DeMaria kids in that 3- to 4-block area.
WCR: It sounds like your father wasn’t home very
much because he was always at the store or working
206 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
a second job. But you could always go down to the
store and see him. Did you work every day after
school after you reached a certain age? How did that
work out?
AND: No. When I got into high school my parents
made it clear to me that my job in life was to get to
college. They didn’t want me to do anything that
would interfere with that. I only worked in the store
summers and on weekends. During the week, my job
was to study and get good grades. My mother did not
graduate from high school. My dad did graduate from
high school and actually could have gone to college.
He had that opportunity, but he had to go into the
family business. At that time my paternal grandfather
was a peddler. He had 4 sons. He had horses and
wagons and as a kid, I often rode the horse and wagon.
They’d go up and down the streets of Bayonne with
fruits and vegetables. By those standards, my father
was a big success. He got his own store! He didn’t
have to peddle anymore. Frankly, if I didn’t want to
work weekends, I didn’t have to. If I ever went to him
and told him that I had to study, he’d say, “That’s all
right. You do whatever you want. Go study. You don’t
have to do work here.”
In contrast, my kids grew up competitive swimmers and I was the starter of the swim team. Every
weekend Lori and I packed up the car and went to a
swim meet. I tried to see every little league and soccer
game my kids were in. I would make the time. My dad
was just too busy. He had to be in the store.
WCR: What was your father like? Were you and
your father close?
AND: My dad was very smart. He was kind of an
overwhelming presence. In our neighborhood, he
managed to get his own house. Very few had their
own house. Families lived in rooms. He was very
facile politically. A lot of Italians in our neighborhood
came to the store to seek his advice on various matters. If they had a problem with the city, he knew who
to call to get a pothole or whatever fixed. He was
active politically. If neighbors needed things done, he
knew how to get them done. He spoke Italian fluently,
as well as English. He was sought after as a political
ally for the elections because a lot of people voted the
way he recommended that they should. In our neighborhood, he was a big “chimichanga.” I was always
quite awed by the respect he was afforded, and by the
fact that we lived in our own house and I could go to
a private Catholic school.
WCR: You had to pay tuition?
AND: Yes. As a kid I thought my dad was about the
biggest success in the world. I wasn’t sure that I could
ever quite achieve what he had achieved. If he had had
the opportunities that I had, only God knows what he
would have accomplished. He was the real McCoy!
WCR: It sounds like he was very resourceful and
that he was a magnet that attracted people to him.
AND: Yes. He was the eldest brother and that’s
why he couldn’t go to college. He had to go into the
business because the other brothers were a bit
younger. He was the glue that held that family together. He was very resourceful and was very handy.
JANUARY 15, 2005
He managed by hook or by crook. I’ll never know in
a million years where he learned it. I couldn’t contract
out a house, but my father knew how to hire plumbers
and stuff like that. What he didn’t know, he knew
friends who would provide him with the information
he needed. He knew everybody. He was very active in
the community.
WCR: Did you take family vacations?
AND: Not really. When I was nearly out of high
school, my parents rented a bungalow in Seaside for a
couple of weeks during the summer, but I worked in
the store.
WCR: What did you do in the summertime other
than work at your father’s store?
AND: Nothing. I played the summer sports—Little
League, Babe Ruth league.
WCR: What was dinner like at home? It sounds like
your father rarely made it.
AND: He didn’t make it very often. During the
summers, he would close the store at noon on Sundays
and we’d pack up the car with a picnic lunch and go
out to one of the lakes. In those days people swam off
Staten Island at a place called South Beach where it
was still swimmable. I remember well those excursions to South Beach or to Lake Hopatcong.
WCR: It sounds like your father was the dominant
figure around the house as well as at work with his
male colleagues.
AND: Yes. My father ran the house and everything
and my mother ran my father!
WCR: What was your mother like?
AND: She was and is a very supportive person. She
was a homemaker. She had great aspirations for all her
kids. She demanded that we achieve at whatever we
did. My sisters took ballet lessons and she’d be with
them, making sure that they could dance on their toes.
She was a meticulous housekeeper. She had her own
circle of friends. Behind the scenes she was the person
who took care of the family because my father was
always at the store or somewhere. My mother got us
ready for school and carted us off to wherever we had
to get carted off to.
WCR: You had a car?
AND: Yes, our family had a 1948 Mercury, James
Dean’s car.
WCR: What about the safety in the neighborhood?
Were you safe in the neighborhood where you lived?
AND: Yes. All of Bayonne was ethnic. It was all
shuffled together. There was no real slum. The rich
area, so to speak, wasn’t all that rich and it wasn’t all
that big. There wasn’t that much that separated the
“haves” from the “have-nots.” It was all mixed up.
That’s true of where I lived. Our house was on a nice
street. I could go 2 blocks away where my father’s
store was and there were the low-income “housing
projects,” consisting of very big box apartment buildings. Bayonne had 2 sets of projects that went on for
several city blocks. One of them started on 49th and
Broadway and went up to 50th. My dad’s store was
between 48th and 49th on Broadway. We lived on
48th off Avenue C, which was 1 block up. The
projects were the toughest part of town. I can’t say that
I ever worried very much about safety. Our house was
never burglarized.
WCR: Your father’s store was never robbed?
AND: Yes, it was a couple times. There wasn’t
much to rob, however. The robbers could steal salami
or something. He didn’t have a large amount of money
there. Everybody knew everybody. There was no way
to commit a crime and not have everybody in the
whole damn town knowing it.
WCR: You spoke English in your home?
AND: My parents were very adamant about that.
My name is mispronounced. The correct pronunciation is not “de Marry A;” it’s “day Maria”. When my
grandfather came over, the Americans called him
“DeMaria” and the concept was that you had to assimilate into English as quickly as possible. My
mother and father had to speak English to each other
because, although my mother was fluent in Polish and
my father was fluent in Italian, neither knew the other’s language, so they had to communicate in English.
That’s what we did.
WCR: What was your mother’s name before marriage?
AND: It was Tess Kupis. That’s not what the actual
name was. It was shortened. It was a long Polish
name. When her parents came through Ellis Island, the
name had 17 or so vowels and it was shortened.
Everyone called her Tess.
WCR: Do you remember dinner at night in your
home when you were growing up? Your daddy was at
the store. What was it like? Did you, your mother, and
3 sisters talk about what happened in school or what?
AND: It wasn’t like my father was never there. He
occasionally came home for supper. We didn’t have a
family dinner exactly. It just depended on who was
doing what. If one of the kids had a recital or had to
go off to Cub Scouts or Girl Scouts or got home late
from track practice, then we ate when we got there.
My mother, who I hope will never read this, was not
a great cook. She has many virtues in this world, but
being a great cook was not one of them.
WCR: But you could always go to the store and get
something to eat if you wanted to.
AND: Yes. We were all skinny as can be. When I
was in high school, I became aware of women and
things like that. I realized then that I was the “90pound weakling.” I went off to college at 6-feet tall
and 122 pounds. I was just skin and bones.
WCR: What do you weigh now?
AND: More than that.
WCR: Are you still 6-feet tall?
AND: No, I shrunk a little bit. I’m 5 feet 11 inches.
My intervertebral disks are desiccating. So are yours.
WCR: I’m sure they are. Was your family strongly
religious? Did you go to mass every Sunday?
AND: My mother always did. There was a children’s mass and we had to go to that one. Those who
went to Catholic school sat in one section at mass and
those who went to public school in another section. It
was fairly regimented. We always went. My dad always ascribed to religion, but he often missed mass
until after the store burned down and then he attended
INTERVIEW/ANTHONY NICHOLAS DEMARIA
207
mass regularly and indeed became an usher, helping
attendees find seats and later passing the collection
baskets. He was very active. They were both pretty
religious by today’s standard.
WCR: Did you have prayers before you ate meals at
home?
AND: No.
WCR: Were there books around the house? Did
your mother read much?
AND: She read some but not a lot. It was not a very
literary environment.
WCR: Were there teachers in grammar school, junior
high, or high school who had an impact on you?
AND: Yes. I attended public grammar school. (My
father felt very strongly about that.) Bayonne, however, was about 75% Catholic. Every parish had its
own school. The concept was that your best chance of
getting to college was to go to one of the Catholic
schools. My dad, however, insisted that I go to public
grammar school because he thought it was very important to learn how to live with people with different
cultural backgrounds than our own. After grammar
school, my parents switched the children to Catholic
private schools, believing they provided a better
chance of getting into college than the public schools.
It wasn’t that kids who went to Bayonne High School
didn’t get into college, but fewer of them went on to
college, whereas most kids who went to the Catholic
schools were targeting college. Then the question was,
“How do you get into a Catholic high school?” We
lived in a town of 75% Catholics and there was only
1 men’s and 1 women’s Catholic high school, and
they couldn’t take all the Catholic high schoolers.
There were other Catholic high schools in Jersey City,
but most were even more competitive. You had to take
an entrance exam. I don’t know what the acceptance
rates of those Catholic high schools were in those
days, but it was very competitive to get in.
In the eighth grade I was assigned an elderly
teacher named Mrs. Murphy. Early on she asked,
“How many of you children in this class are going to
try to go to Catholic high school and take the entrance
exams?” There were about 8 of us in my class of 25
students. Our parents had decided that we were going
to try to go to a Catholic high school. She said, “I’m
going to come in 1 hour early 3 days a week and we
are going to start preparing for the exams.” She did.
The tests to see if you could qualify were given in the
spring. Until April or May she came early 3 days each
week to go through vocabulary with us. I don’t want
to sound like I was in any way disadvantaged, because
I wasn’t. My parents had a standard vocabulary. They
didn’t have advanced language skills. Mrs. Murphy
went over these things. We practiced taking the test. It
was multiple choice, and we hadn’t taken one of those
to that point. I took the exam. I was not the smartest of
those 8, but I got into the Catholic High School in
Bayonne and none of the others did. I was very
fortunate. I’m convinced that I never would have
gotten in if Mrs. Murphy hadn’t worked with us. She
was at the stage of her life where she had a mission.
Her own children were grown. Her life was teaching
208 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
us kids. She was a spectacular schoolteacher. In seventh grade, sometimes our teacher would come in and
say, “Open your book and read this chapter.” There
was a lot of free study time. Mrs. Murphy was very
structured. I never did get a chance to thank her. She
was delighted that somebody made it. Only 1 person
also made it the year before.
WCR: You started Catholic high school in the ninth
grade. Was there a big change? Was it much harder
than the public grammar school and Catholic junior
high that you’d gone to?
AND: Yes. Where I lived it was heavily Catholic.
In Bayonne alone there were 7 parish schools. In the
Catholic high school, my classmates were on average
smarter and often came from families where the parents were schoolteachers, doctors, lawyers, and businessmen. That made it tougher for me. They had a bit
of an edge coming in.
WCR: They also had all come from parish schools?
AND: Pretty much.
WCR: Were you active athletically?
AND: I was a good long-distance runner. In my
freshman year in high school I finished second in the
conference. I won a lot of races. As I got older, I didn’t
improve as much as the others. There were a lot I
could beat as a freshman, but by the time we were
seniors, they had just gotten better than me. I achieved
some success at track and I played basketball for a
couple of years. It was clear that I was not as good at
basketball as I was at track. I never made the starting
team.
WCR: What races did you run?
AND: Cross-country in the fall and I was a miler or
half-miler in the spring.
WCR: Cross-country was usually how many miles?
AND: It was 2.5 miles. We’d run it in a park. They
would set up a course so that you had to go up and
down some hills and through some fields. It wasn’t
run around a track. In the winter and spring, we ran on
the track.
WCR: You were second in the long-distance run as
a freshman.
AND: That was my pinnacle. That was in our
region of the state. In Bayonne, I won the mile and
half-mile a couple of times. Once a year there would
be a track meet for Bayonne High School, Marist (the
Catholic High School I went to), and for other people
who went to other schools who lived in Bayonne.
Being the champion of Bayonne wasn’t that great.
WCR: How many were in your high school per
class?
AND: About 90.
WCR: How did you finish in the standings at the
end of your senior year in high school?
AND: I did okay.
WCR: You were first in your class?
AND: No. They never gave out a ranking exactly.
At my graduation they gave awards to 7 people. To
get the top award you had to have gotten an A in every
single course for 4 years. To get the second award,
you had to get at least a B⫹ in every course for 4
JANUARY 15, 2005
years. I got the second award. I missed getting the top
award by 2.
WCR: Did you have to study hard for those grades
or did they come easy for you?
AND: I had to study hard.
WCR: Were there teachers in high school that had
a major impact on you?
AND: Yes, there were a number of them. I had
religious brothers. They are like priests, but they don’t
say mass. One of the brothers was my track coach and
he helped me a lot. He was a good friend, but got
transferred after a year. In my senior year, I had a
brother, Hugh Arthur, a French/Canadian guy who
scared everybody to death. He ran a tight ship. His
field was mathematics. He also taught religion and
something else. I was assigned to his homeroom class
in my senior year. He was hard as nails, but he was
actually a very warm person underneath it all. He gave
me a lot of good advice. As seniors in high school, we
would get distracted and get into a bit of mischief. I
would never claim that I was the best student in the
world.
WCR: Was it a co-ed high school?
AND: No.
WCR: All boys. Where were the girls?
AND: They were down the street at Holy Family.
WCR: You dated them in high school?
AND: Yes. I went to mixers and dances.
WCR: It sounds like high school was a good experience.
AND: Yes. I enjoyed it.
WCR: When it came time to go to college, you
really couldn’t get much advice from your parents or
from your father’s brothers because they’d never been
themselves. Were there many members of your mother’s family in the immediate neighborhood? Did you
know your mother’s family?
AND: They lived in Elizabeth, New Jersey. We saw
them from time to time. She had a brother and a sister.
Each of them had 1 child, a boy. We didn’t see much
of them. We weren’t close. My grandmother on my
mother’s side died relatively young from a subarachnoid hemorrhage. My maternal grandfather lived
about 90 years. He couldn’t speak English very well,
although he worked for decades for Esso (Standard
Oil) in Elizabeth. I couldn’t communicate with him
much, because I didn’t speak Polish.
WCR: You didn’t know either of your grandparents
on your father’s side?
AND: I did. They lived in Bayonne.
WCR: You had 4 grandparents.
AND: Yes. But it was hard to communicate.
WCR: On your mother’s side?
AND: On my mother’s side and my father’s side.
My paternal grandmother spoke halting English. My
paternal grandfather could speak a fair bit, and I could
communicate with him.
WCR: When it came time to pick a college, how did
that work out?
AND: The brothers pushed us toward Catholic colleges. Many of my classmates went to St. Peter’s
College in Jersey City. A batch went to Seton Hall.
Some went to Rutgers. I can’t remember anybody
going to an Ivy League school. I wanted to get away
from home, but not too far. I told the brothers that
maybe I’d go to New England or down south. They
came back with Holy Cross or Georgetown. I applied
to St. Peters and Seton Hall; everybody did. While I
was sifting through that, I learned about Dartmouth.
That seemed to me like it might fit the bill. I wrote off
for their catalog, but I was roundly discouraged from
pursuing that by everybody. For whatever reason, they
said it wouldn’t be a good alternative. I wound up
going to Holy Cross.
WCR: You went to Holy Cross in Worcester, Massachusetts. Your parents must have been extremely
proud of you for having gotten into such a good
college?
AND: Yes. It was a good college to get into. Most
of my classmates stayed locally. I had a lot of really
smart classmates. There was a school called the
Stevens College of Engineering in Hoboken. It’s still
there, but now it’s called something else. A number of
my classmates went there. It too was considered quite
a good school.
WCR: Hoboken? That’s where Frank Sinatra was
from.
AND: That’s right. My dad knew Frank Sinatra.
They dated cousins before my father married.
WCR: What did your father say about Frankie?
AND: He said he was a nice guy. My father was
dating a woman and she had a cousin. Her cousin was
dating Frank Sinatra. They double-dated a number of
times. They got to be friends. He wasn’t anything
then, just a skinny Italian kid from Hoboken. My dad
lived in Hoboken before his family moved to
Bayonne. Frank Sinatra married Nancy, but my father
didn’t go with the other cousin anymore. They drifted
apart. By the time my father was dating someone else,
Frank Sinatra had started to make it big. He was
playing a nightclub in New York City. My father
wanted to do something special for New Year’s Eve.
On a whim he called the nightclub and asked to speak
to Frank Sinatra and told them it was Tony DeMaria
calling. According to my father, Frank got on the
telephone and asked how he was doing? My father
asked Frank, ”Do you think you can get me a ticket to
your show?“ And Frank did it; he got him a table for
his show on New Year’s Eve in a nightclub in New
York City. In Bayonne, it just doesn’t get any better
than that.
WCR: You are a junior.
AND: The tradition was that you name the first son
after the grandfather. My grandfather’s name was
Nicolas. My mother just wouldn’t abide by naming
me Nicholas. They compromised on Anthony, which
was my father’s name, but Nicolas is my second
name, after my grandfather. Of course, I went back to
the tradition so my firstborn was named Anthony after
my father. The fact that it happened to be my name
was a side bonus.
WCR: Did you go to Holy Cross to interview before
you enrolled?
AND: They didn’t have interviews. Maybe they
INTERVIEW/ANTHONY NICHOLAS DEMARIA
209
did, but they didn’t call me for an interview. I just
applied. I remember that the brothers, when I told
them which college I was interested in, advised me to
apply to Fairfield in Fairfield, Connecticut, Providence
College in Providence, Rhode Island, and at Holy
Cross. One weekend my father took off from work.
(I’ll carry this to my grave because my father never
took off from work.) He got one of his brothers to
mind the store. We left on a Friday evening and we
drove to visit Fairfield, Providence, and Holy Cross in
Massachusetts. That drive was the most time I’d ever
spent with my father alone. Shortly thereafter, I got a
letter of acceptance from all 3 of them. We decided I’d
go to Holy Cross, because it had the best reputation at
the time.
WCR: Holy Cross is how far from Bayonne?
AND: Two hundred miles. It was a 5-hour drive.
WCR: How many students were at Holy Cross when
you were there?
AND: Sixteen hundred.
WCR: About 400 in each class. It’s a beautiful
college.
AND: It was a nice school. It was on a hill overlooking the city. Surprisingly, for a small steel belt
town in the middle of Massachusetts, Worcester had
quite a number of colleges: Worcester Poly Tech,
Clark University, Anna Maria (the Catholic women’s
college), Holy Cross, and Assumption College. It was
a nice setting, but it was very strict. I had to go to mass
every morning for the first 2 years I was there. They
took attendance.
WCR: You started college what year?
AND: In 1960.
WCR: You were 17. How did college strike you?
That was the first time you’d ever been away from
home, right?
AND: Yes.
WCR: How did it work out?
AND: I was nervous. It was like a big filtering
process. Going from Washington Public School to
Marrist High School, they only took so many into my
high school. Going from my high school to college,
there were guys from all over the country. The major
Catholic colleges in the USA then were Notre Dame,
Georgetown, and Holy Cross. There were a lot of
valedictorians there. Holy Cross was known for premed education. Even in high school I understood that
getting into medical school was tough. One reason I
went to Holy Cross was because they had a reputation
for getting people into medical schools. When I got
there, there were 150 of us who entered as premed
(150 of 400 were designated as premed). They called
us together at the beginning of school and gave premed orientation in a large auditorium. They told us
that of the 150 students who were there, there were
lots of valedictorians of their respective high schools.
As I got further along, there were fewer guys like me
from Bayonne and more who had come from professional families. It was very strict. They told us that,
they “never had a person that they gave an unqualified
recommendation for medical school that wasn’t accepted to an American medical school.” They fol210 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
lowed that up by saying, they “had never recommended more than 55 or 60.” You didn’t need to be a
rocket scientist to realize that there were 150 of us,
and the most they ever recommended was 60, so there
were a lot of us who would not be recommended.
Once that was clear, everybody looked at each other
and said, “Who is going to be left standing at the end
of 4 years?” That was the atmosphere that permeated
the place. Folklore had it that at one time, they had put
all the premeds in 1 dormitory on campus that the rest
of the student body named “The House of Death.”
They didn’t do that when I was there. It was very
competitive as a premed. But it was a good education.
It prepared me for medical school. Medical school
was quite a bit easier than Holy Cross.
WCR: It sounds like you had already decided that
you wanted to be a physician before you entered
college. How did that come about? You had no physicians in your family.
AND: I’ll tell you a true story. My mother and
father had been targeting me to go to college from the
time I could remember and telling everybody I was
going to go to college. They weren’t shy about it; they
told everybody. It got to be senior year and the whole
neighborhood (half of which were my family) was
wondering what I would do. Everybody knew each
other. Sure enough, I got accepted. My grandmother
and my aunts and uncles were asking, “What are you
going to take up in college?” I didn’t know. How do
you know? I liked history. It was easy and I did well
in it in school. The first thing I said was, “Maybe I’m
going to do history.” That drew a blank expression.
They didn’t know any historians in Bayonne. They
were saying, “History. History. What is that?” That
bombed. I rummaged around a bit and decided biology was pretty interesting and I had always enjoyed
that. So I told them “Biology.” That didn’t ring any
bells either. Honest Bill. This is the truth. I felt the
pressure here. One day my grandmother came to me.
“Anthony, Anthony. What are you going to take up in
college?” And I said, “Maybe I’ll be a doctor.” “Aahhhh,” was her response. All my worries were gone. I
just said “doctor” and they were happy. I didn’t know
if I wanted to be a doctor, but I figured I could change
along the way. It got everybody off my back telling
them what I was going to do. Nobody would believe
for a minute that you would go to college to get an
education. What good was that? My dad said to me
one time when I was flailing around trying to think
what I wanted to do, “Think of college like a train.
You are going to get on the train. When you get off
you want to be somewhere. That train is going to take
you. You have to figure out where it is you want to be
when the train stops.”
WCR: That’s beautiful.
AND: Once I got to Holy Cross, the question was
no longer what I wanted to do. The question was could
I do it? It became a challenge. Was I going to be
standing at the end of 4 years to get their recommendation? During orientation week they said, “Everybody needs to go to confession to get in the proper
spirit, to take communion, etc.” You went to mass
JANUARY 15, 2005
every morning so you might as well go to communion,
too. I went to confession with one of the Jesuits from
New England. I say my thing, “Bless me Father, etc.”
He said, “What did you say? What did you say?” I said
it again. This went on for 3 or 4 minutes. He told me,
“Son, I want to tell you something. I can’t understand
a word you’re saying. You’re going to waste your
money on an education. Nobody’s going to understand you. You’ve got to get rid of this accent.” I had
a heavy New Jersey accent. There weren’t a lot of us
there at the time.
WCR: Did you go on scholarship?
AND: Partial.
WCR: Partial scholarship from the very beginning?
AND: Yes. They didn’t offer me a scholarship right
off when I was accepted, but my father wrote them a
letter. He asked them if there was any chance that I
could get some financial help. They sent back an
application form where he had to put down his income
and expenses. Sure enough, they came back and they
gave me a 50% scholarship.
WCR: That was tuition?
AND: Maybe room and board and tuition. They
also got me a government loan.
WCR: How did Holy Cross work out? You said it
was very rigid. You were scared because the competition was vigorous. You hadn’t met all these students
who had come from professional households, many of
whom, I presume, were fairly wealthy also.
AND: I can’t say it was a wealthy place, but many
of the others had more varied experiences than I had.
They had often been exposed to more things. However, there were a number of guys like me. Holy Cross
wasn’t a ritzy school compared to Georgetown. College is a filtering process. To get into the place, nearly
everybody was in the top 20% of their high-school
class. If you’re in the top 20% of your class, you are
better than 80% of the other guys. When you get into
college, everybody is in the top 20% and things are
tougher.
WCR: You enjoyed the challenge though? You really poured it on from a studying standpoint.
AND: It was my first time away from home. For the
first time I was footloose and fancy free. I goofed off
a bit and was slow academically. I caught up. It was a
good college life and I’ve no complaints. I had to
work very hard. When I got into medical school,
compared with Holy Cross, it was easier.
WCR: What was your major in college?
AND: Liberal arts. If you were premed, you either
took a bachelor of science in biology or a bachelor of
arts. I was a bachelor of arts. I didn’t have a major. I
took the science courses that I needed for medical
school. It was pretty prescribed and I took all liberal
arts courses—languages, philosophy, history, etc.
WCR: Were there any teachers in college that had
a major impact on you?
AND: In my freshman year I had a faculty advisor,
Father Leo McCarthy. I talked to him. He provided
some encouragement. That was really important for
me early on to get me through what was a fairly
challenging transition. Many of the other students
came from elite Jesuit high schools. They had come
from an environment that was similar to Holy Cross.
Maybe they had a bit of a running start. It took me a
little longer to get going. Father McCarthy transitioned me through that pretty well. I did pretty well. I
obviously was in the top 50 premeds because I was
recommended and got accepted to medical school.
WCR: Did you enjoy Worcester? Did you go into
Boston much?
AND: You could only go off campus until 11 P.M.
If you were leaving the campus, you had to sign out,
you had to sign back in, and they did bed check.
Somebody came around to make sure you were in
your room. (This was a long time ago.) During the
week, the lights had to be out at a certain time. That
was it. If you had to cram for a test or something, you
had to go to the bathroom and get in one of the stalls
and sit in there. On the weekends, you had to sign out
on Friday and Saturday. I don’t think you could go off
campus on Sunday. Most of the time students stayed
on campus. It was an all-male school; there were no
women. We were not allowed to have cars until we
were seniors. The first year, all premeds just studied.
Taking Saturday night off meant you went to the
dining area and watched a movie on campus. If you
wanted to go to Boston, you had to hitchhike. You
could sign out for the weekend, but you had to give
them an address. I didn’t have any addresses. It wasn’t
a big social school.
WCR: Who did you date up there?
AND: There was a women’s college. There were a
lot of students there in Worcester. When you got to
Boston, there were even more students. A lot of us had
girlfriends back home and we corresponded with our
girlfriends. When you got a chance, you’d go back
home. That was that.
WCR: Did you go back home much during the
school year? Christmas or Easter?
AND: Always Christmas and Easter and usually
Thanksgiving. That was a long weekend. There was a
guy from my hometown of Bayonne who was also at
Holy Cross. He was a senior in my first year. He was
the son of one of the orthopedic surgeons in Bayonne.
Whenever he was going to Bayonne, he’d call me and
if I could do it, then I’d ride in his car to Bayonne and
give him some money for the cost of the gas.
WCR: Were you sick much as a child? Did you go
to a doctor very often? Do you remember any doctor
visits before you went away to school?
AND: Yes. I used to have a lot of bronchitis. In high
school, my mother took me to a chest doctor in Jersey
City. He gave me what was the equivalent of puffers
in those days and some nose drops. In retrospect, it
may have been allergies. In college I had pneumothoraxes. That put me in the infirmary or the hospital.
WCR: These were spontaneous?
AND: Yes. I was a skinny kid. It scared the hell out
of me the first time I had it.
WCR: How did you end up in college? Do you have
any idea? You mentioned that of the 150 premeds, you
finished in the top 50. Did they give you any idea
where you stood in your class at Holy Cross?
INTERVIEW/ANTHONY NICHOLAS DEMARIA
211
AND: I made the Dean’s list a number of times. I
don’t know where I finished, but I would say of the 50
guys who went to medical school, I was probably in
the middle of that group.
WCR: It’s also fair to say that those 150 premeds
probably finished up higher than the 250 who started
freshman year and weren’t premed.
AND: Oh yes. Of course, there were examples of
very bright guys in other areas, but, as a group, the
premeds were generally at the top of the list.
WCR: What did you do during the summers when
you were in college?
AND: I worked in my dad’s store for a year, I
collected tolls on the New Jersey Turnpike for 2 years,
and I drove a truck for a dairy distributor.
WCR: That paid pretty well?
AND: It paid more than I’d ever made working in
the store. If you got a summer job collecting tolls on
the New Jersey Turnpike, you got the worst shifts in
the worst places. I’d wind up working the Lincoln
Tunnel booth on Sunday when everybody was coming
back from the Jersey shore. I worked a lot of nights.
Those shifts screwed up my summers. My father
called me and said that the guy who distributed dairy
products to him was looking for a truck driver, somebody to distribute to different stores. I did that for a
couple of summers. I worked Monday through Friday,
8 to 5 P.M.
WCR: When it came time to apply to medical
school, did you apply to several medical schools?
What happened?
AND: Everybody was really worried, even the top
guys. It has always been competitive to get into medical school and we all had that anxiety. I applied to
Seton Hall University, the local medical school in
New Jersey. A bunch of my colleagues had applied as
well. The Dean of Admissions at Seton Hall actually
made a trip to Holy Cross. I’m sure he went to other
schools in New England. He was there for 1 day and
interviewed 6 of us. This happened in the first couple
of weeks we were back in school as seniors. About a
week after that, I got a letter in the mail saying that I’d
been accepted to medical school. In the first month of
my senior year I knew I had a place. My focus
thereafter drifted a bit. You asked me about Boston
before. I spent a whole lot more time in Boston. I
applied to a couple of other places, too, but frankly I
couldn’t have afforded to go elsewhere. Seton Hall
was perfect for me because I lived with my folks
again. It was the only sensible thing to do. They didn’t
give me a scholarship, but they gave me a big loan for
school. Between living at home and having a loan, it
was financially feasible.
WCR: How far from home was Seton Hall?
AND: I didn’t have to stop at a red light from my
house to there. Seton Hall was at the Jersey City
Medical Center. I was in one of the first classes. The
medical school didn’t last all that long. Ultimately, the
state took it over and it became the New Jersey College of Medicine.
WCR: It was private when you went there?
AND: Yes. It was run by Seton Hall University,
212 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
which is a diocesan school. I went there for 2 years. It
was too expensive for the diocese to run Seton Hall.
The state of New Jersey took it over as a state medical
school. I graduated from the New Jersey College of
Medicine.
WCR: You were back living at home after 4 years at
very sophisticated (in a way) Holy Cross College.
How did you like being back at home even though you
were going to medical school at the time?
AND: It worked out well for me because I was back
in my old attic room. I was able to focus. When I was
away at college, some of the distractions of college
life got to me. The reason I bring up Ned Weyman is
that we were very similar. He was 2 years ahead of
me. I got to know Ned because his roommate and I
were hospitalized together once. I had a pneumothorax and Ned’s roommate had a thoracic injury playing
lacrosse. Here we were, these 2 college students in this
hospital with a bunch of student nurses. That provided
some distraction for a good part of the year after that.
It seemed like in medical school I’d better not be
distracted. Living at home wasn’t like being at college, where someone was playing cards all the time or
somebody was going down to get a beer or any of
those things. It worked pretty well. My mother made
all the meals and did all the laundry for me. I must
admit, in retrospect, that it was a good thing. I suspect
that I had stayed at home for college, I might have
wound up at the very top of my class. I didn’t mind
living at home at all. The first couple of years of
medical school were pretty rigorous. I had a girlfriend
at the time. It was convenient. I’d call her and we’d go
to movies. Living at home worked out well.
WCR: How many students were in your medical
school class?
AND: One hundred.
WCR: How many medical schools are in New Jersey?
AND: There was only 1 then. I think there are 3 now.
Rutgers has 1. They are all part of the same system.
There’s 1 in New Brunswick called Rutgers. There’s 1 in
Jersey City and there may be 1 in Camden.
WCR: What is your medical school called now?
AND: The College of Medicine and Dentistry of
New Jersey. It was Seton Hall when I started and the
New Jersey College of Medicine when I finished.
WCR: It moved from Jersey City to Newark. Were
there any surprises in medical school for you? When
you first entered medical school did anything really
strike you that struck you as very unusual?
AND: Candidly, when I finished grammar school, I
was one of the smartest kids. When I finished high
school I was one of the smartest, right near the top. I
went to college and I didn’t finish with the top grades.
I dropped down a bit. That shook me up a little. For
the first time coming out of college, I wasn’t coming
out in the top 5% of the premed class. My batteries
were a little charged for medical school. I was rip
roaring ready to go. I spent the first 2 years of medical
school making sure I was in the top 10 or so students
in medical school. However, I did go through a period,
which started at the end of my first year medical
school and hit home in the second year of medical
JANUARY 15, 2005
school, when I really wasn’t sure that I wanted to
study medicine. There were not any counselors at that
stage of the game, but I was sent to talk to the Dean by
a professor. In medical school, I had been pulled out
of the standard classes to be in a special group that
was doing special studies on the side. I got to know
Professor Bernard Briody, another hard-ass kind of
guy. Everyone was a little timid around him. He had
come up with this idea to pick some medical students
and not bother them with the usual day-to-day drudgery of memorizing stuff and put them in some special
classes. I was put in that class and got to know him.
We had a special relation. I told him that I wasn’t sure
I wanted to do medicine. He sent me to the Dean. The
Dean was a little bit surprised because most of the
people he was seeing that weren’t sure about going
further didn’t have good grades. That was not a problem for me. To make a long story short, he said,
“Don’t do anything dumb. The first 2 years of medical
school aren’t exactly representative of what medicine
is all about. You finish the second year and then take
the summer off. Do something totally different and
come back in the fall and start again in your clinical
training and see what it’s like. Take another look at
it.” That seemed like a reasonable idea.
The question was what was I going to do in the
summer. One classmate was a lifeguard at the Jersey
shore at one of the state beaches. He said, “There’s a
lifeguard who has helped with the first aid station who
has had the job for a couple of years and he’s not
coming back this year. He’s going to do something
else. His name is Carl Pepine. Why don’t you take his
job?” I did. I took Carl Pepine’s job on the Jersey
Shore. I went down there to work to get away from
medicine and to let things sift out. While working
there, I met a woman who worked in one of the
hamburger stands at the shore.
WCR: Lori?
AND: Yes, she’s now my wife. We started dating.
I returned for my third year of medical school and it
was quite a bit different.
WCR: What were the characteristics of Lori that
attracted you to her when you were a lifeguard?
AND: Here I was on the lifeguard squad in good
physical shape. I worked out every day. I was tanned.
Most people knew that I was a medical student. Even
if you have a bad case of the uglies, you are going to
do okay under those circumstances. How many
bronzed lifeguard medical students are having trouble
finding dates? Not many.
WCR: Were you living at home?
AND: No. I was living at the shore with another
medical student. When I first got there, I worked on
the beach all day and went home at night. I was kind
of bored. He was volunteering at some community
hospital. I wasn’t going to do that. There was a big
singles’ bar called “The Surf Club,” with 14 finger
bars. It was enormous. I was having a beer there and
asked them if they needed any help. I was anxious to
work at night. The owner said, “As a matter of fact, we
need a bouncer.” By this time I weighed about 165
pounds, but I was still not Gunga Din. I asked the guy
to look at me. He said, “Number one, everyone wears
casual clothes except the bartenders and bouncers who
wear white shirt and tie. So, people are going to know
you are a bouncer, and just that fact alone means no
one is going to mess with you even though you don’t
look like anything. They are going to think you must
know karate or something to be a bouncer. Virtually
everybody that you are going to need to take care of is
going to be drunk.” I became a bouncer for about 1
month and then a bartender position opened up. If
there’s one thing better than being a bronzed lifeguard
medical student, it’s being a bronzed lifeguard medical student tending bar at the Surf Club with all these
singles around. I had all of that going on and here’s
this woman I’m taking a shine to. I had noticed Lori
on the beach and she seemed like a pretty classy lady
and a nice person. I tried to chat her up a little bit and
she didn’t give me the time of day. I wanted to tell her,
“Lady, don’t you realize? There are 500 women who
would drop dead to get a date with me. You don’t
seem to give a damn.” Finally, by hook or by crook (I
don’t know how I worked it), I got a date with her and
we hit it off.
WCR: How long after that did you get married?
AND: The next year.
WCR: You got married after your junior year?
AND: We got married in the summer between
junior and senior year. We got engaged as my junior
year in medical school was coming to an end. We
thought we would get married when I graduated.
WCR: Where was Lori going to school?
AND: She was going to Rider College in the middle
of New Jersey. That was the grand scheme. As the
summer wore on, we just didn’t want to wait for
another year. Neither of us wanted a big wedding
anyway, so we decided that we would get married and
live midway between Newark (where my medical
school was) and the Princeton area (where Rider was).
The idea was that we were going to live in between the
2, so I’d commute about 45 minutes and she’d commute about 45 minutes. I’d finish medical school;
she’d finish college. We went to our families and
announced that we were going to get married. We
pulled out all the stops and got married in August. The
problem was that being good Catholics, Lori got pregnant instantaneously. That screwed everything up. She
took a job substitute teaching instead of going back to
college during my senior year at medical school. We
had Christine. (Lori went back and finished college,
when I was on the faculty at UC Davis. Her degrees
are from UC Davis.) When I started my internship we
had a baby already.
WCR: When you went back to medical school to
begin your junior year after the summer of relaxing,
meeting Lori, etc., medical school became very attractive to you again. How did that come about?
AND: The first 2 years of medical school were
really like graduate school in biology. It was heavy
basic science, the only clinical thing we did, of any
kind, was Introduction to Physical Diagnosis, where
we examined patients. The rest was biochemistry,
physiology, and immunology, etc. After getting to the
INTERVIEW/ANTHONY NICHOLAS DEMARIA
213
clinical rotations in the third year, I became very
happy with medicine. It came pretty easily to me. I
didn’t have to work so hard at it. I enjoyed interacting
with people. I liked the medicine part of medical
school.
WCR: When going through your rotations, did you
have any idea what kind of doctor you wanted to be?
As you rotated through surgery, pediatrics, medicine,
and all the rest, was your decision to go into medicine
easy for you?
AND: I knew I didn’t want to be a surgeon and
there was nothing else that appealed to me. The need
for dexterity and being in the operating room to fix
things up just didn’t appeal to me. I flirted a little with
orthopedics. I thought orthopedics might be interesting. I got a telephone call from one of the people in
medical school asking me what I was interested in. I
told them orthopedics. They arranged for me to do
orthopedics for one of my very first rotations. That
dispelled my interest in orthopedics. I knew I wanted
to do medicine, and it was just a question of what kind
of medicine to do.
WCR: Did any clinicians in your last 2 years in
medical school have a particular impact on you?
AND: Carol Leevy, Chairman of Medicine, a hepatologist, par excellence, was a world-class investigator. He’d lived in the Hudson County area and was
very attached to it. When the medical school made the
transition from Jersey City to Newark, there was a lot
of disorder as you might imagine. Although a very
productive researcher, he was called upon to grab the
reins of the Department of Medicine through the transition period and he did.
WCR: Were you first in your graduating class in
medical school?
AND: No. I’m sure I wasn’t first. I wasn’t the
valedictorian speaker. For the life of me, I can’t remember who was. I wasn’t too far from first.
WCR: When it came to picking an internship, how
did you reason that out?
AND: I was married and had a baby. My becoming
a physician exceeded everyone’s wildest expectations.
I not only went to college, I went to medical school. I
became a doctor. The entire neighborhood was abuzz.
When it came to internship, there was a guy who
really was much revered by all the medical students.
His name was Gustave Laurenzi. He was a pulmonologist. He really made an impression. He was flamboyant, had a big mustache, had a lot of dramatic
qualities, and, of course, was very smart. Harold Jeghers, Chairman of Medicine at the New Jersey College
of Medicine, had been recruited to go to Worcester,
Massachusetts, to a hospital called St. Vincent’s to
participate in the planning of the University of Massachusetts Medical School. He recruited Gustave Laurenzi to be the chairman there. Laurenzi went around
to the students that he knew, and made a pitch to
consider going to that hospital for internship. It was, to
a large degree, the power of his personality coupled
with my goal of going into practice. I figured I’d do an
internship and residency and open an office and be a
practicing doctor. Having a wife and child also was a
214 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
significant factor. The idea of going to a hospital
where they had a house-staff apartment building next
door was fairly attractive. Five of my classmates and
I took an internship at that hospital. They filled their
internship program. It was the biggest bolus they ever
had gotten from 1 medical school. For me, there was
a little extra in it because I already knew Worcester.
I’d gone to college there. It was like going back to my
college town. It was very convenient. I figured for
being in private practice, it might even have an advantage because that’s what private practice was all
about. I made the decision and went with Laurenzi.
From the standpoint of clinical medicine, it was a
great internship. We were deluged with patients and
worked every other night and every other weekend.
The hours were long beyond belief. The “clinical
material” was everywhere. They had a solid group of
very busy practitioners and they had a full-time faculty of about 12, a full-time staff in every specialty. It
was a place where sick people were taken care of.
Laurenzi and all the full-time people did some research. The Chief of Cardiology was Richard Myler,
who subsequently went to San Francisco and coronary
angioplasty fame.
At the end of that year, I was drafted. I applied to
go into the United States Public Health Service. I
initially contacted them because I was going to volunteer for an Indian reservation. I figured that would
be an idyllic place. By this time my wife was pregnant
again with our second child. I thought that might be an
interesting way to spend 2 years because Vietnam was
red hot. I had to discharge the service obligation. They
came back with an opening for a medical resident in
the Staten Island Public Service Hospital. I told them,
“You got it.” I was a resident at the Public Health
Service Hospital. That turned out to be phenomenal.
The hospital was busy as could be. There were many
junior faculty from medical schools who had joined
the faculty before the Vietnam war, but were still
eligible for the draft. We were eligible until age 35.
They were pulling guys out of junior faculty positions
to go back in the service. There were a lot of guys who
signed up to do 2 years in the Public Health Service.
We had a ton of patients and we had an interesting
accumulation of young academic people. We didn’t
have a lot of grand professors, but there were some
lifetime guys who were pretty good. It was a really
great experience. Of course, in Staten Island, Tony
Damato had established a cardiac unit, where they did
the first His bundle electrocardiogram. They gave
birth to the field of clinical electrophysiology. There
were a ton of guys going through there: Ken Rosen,
Mike Rosen, Bruce Goldreyer, John Gallagher (a
classmate of mine at Holy Cross), Masoud Aktar, and
Mark Josephson.
WCR: You finished your medical residency as a
Public Health Service officer, which fulfilled your
draft and also paid you reasonably well. When did you
decide you wanted to become a cardiologist?
AND: The pay wasn’t great, but that’s basically
true. There wasn’t any way Staten Island was a bonanza! It was good medical training, I was exposed to
JANUARY 15, 2005
a lot of new developments going on in medicine, and
I met many of the up-and-comers in cardiology. Cardiology was a huge thing at Staten Island because of
all the His bundle work and the electrophysiology, etc.
That tweaked my interest in cardiology. There was
also some thinking that the Public Health Service
might need subspecialists, particularly cardiologists. It
sounded good to me and I thought maybe that was the
way I ought to go.
WCR: How did it come about that you went to
Davis, California, to do your cardiology fellowship?
AND: Another senior resident and I both decided
on cardiology. As fate would have it, we were kind of
desirable properties; because of our being at Staten
Island, we had a fair bit of exposure to intracardiac
electrophysiology when no one else had. My fellow
resident interviewed at Indiana with Charles Fisch
and was accepted into Fisch’s program. Later, Fisch
mentioned to Dean Mason, who had just gone to UC
Davis to be the Division Director of Cardiology, that
he had hired a senior resident from the Public Health
Service and heard that there was another guy looking
for a cardiology fellowship. I got a telephone call from
Dean Mason who said, “I’m starting a cardiology
program out here and I think you might like it. Why
don’t you come out and take a look?” He even offered
to pay my way out. My wife got a baby sitter for the
kids and we both flew out to Davis on a cold, rainy
New York day in March. When we got to Davis it was
sunny and warm. Dean Mason was an impressive
person and we said, “Fine.” We were there for 36
hours and even rented a house for our coming in July.
Because of being with Damato and all those electrophysiology guys early on, we were pretty desirable.
I actually had the opportunity to go to Duke, to Yale,
and to Emory. I went to Davis. That was one of the
best decisions I ever made. But it was truly uninformed. A knowledgeable person probably would not
have turned down the others. Willis Hurst also called
me and offered me a job on the telephone. He’d
written a big cardiology textbook by that time. I was
on rounds at the Public Health Service Hospital and a
nurse came up to me and said, “Dr. DeMaria, there’s
a Dr. Hurst from Atlanta on the phone.” I thought,
Holy Mackerel! But, I said “no” to all of that and went
to this fledgling embryonic program in Davis. I was in
the first fellowship class. Some of it was the charisma
of Dean Mason. A lot of it was the attraction of being
in California for the first time. It was just a gut feeling.
I sat down with Dean Mason and he said to me, “What
do you want to do in cardiology?” I said, “I’m going
into practice. I’m going to open an office and practice
on the Jersey shore where my wife lives because we
like it there. We thought it might be fun to come to
California for a couple of years and be trained, before
we settled back East and I go into practice.” That’s
exactly what I told him.
WCR: This was during the interview?
AND: Yes. It shows you what bad shape he was in;
he took me anyway.
WCR: How did he respond to that?
AND: He didn’t say much of anything. As I say, I
had a zero degree of sophistication in terms of what I
was doing. I turned down all the prestigious places to
go to Davis. No one even knew where Davis was. I
had no idea you were supposed to say that you wanted
to do research. Nobody told me that. I just told him
what I thought. I figured honesty was the best policy.
WCR: Had you written any papers by that time?
AND: No.
WCR: Research hadn’t even occurred to you yet?
AND: Correct, not in the least. I got the intern’s
award during my internship. There was a competitive
test in the Public Health Service Hospital and I finished on top of that also.
WCR: Who had Dean Mason hired by the time you
arrived at Davis?
AND: He’d hired Ezra Amsterdam from Boston,
Rashid Massumi from George Washington, and Jim
Spann and Bob Zelis, both former colleagues of
Dean’s at the NIH.
WCR: Were you the only fellow?
AND: No. There were 3 of us.
WCR: Who were the other 2?
AND: Richard Miller, who went on to be chief of
cardiology at Baylor College of Medicine in Houston
and Louis Vismara.
WCR: How did it work out for you at Davis?
AND: I went there with the intention of learning
how to be a cardiologist for 2 years while enjoying
northern California and exploring the West Coast. Of
course, the environment I was exposed to was totally
research oriented. With Dean setting the agenda, I’d
no sooner arrived there than I realized that the most
important thing was that in September (a couple of
months later) there was a deadline for research submissions to the Annual Scientific Sessions of the
American College of Cardiology and what was I going
to do to get some research submissions. The next most
important thing was a grant deadline. I was entrained
by the whole momentum of the place. Zelis and Amsterdam were junior guys who were very ambitious.
Massumi, the electrophysiologist, was a very harddriving individual. I just got swept up in it. I couldn’t
help but being swept up in it. I enjoyed it.
I got derailed a little bit early on. I had a basic
research project where we tried to study the effect of
asynchrony of contraction on intrinsic myocardial performance. I’d go to the dog lab every day. It was a
brutally difficult protocol to implement. We did cardiopulmonary bypass and put a button in the mitral
valve position in the dog and put a balloon in the left
ventricle. I didn’t like this very much. Dean Mason
called me in one day and said, “I’ve just gotten this
machine to work with that I think is going to be an
important tool in cardiology. We are one of the few
places that have it. I want you to take it on as your
project. It’s called the radarkymogram. ” The radarkymogram! Do you remember that? It was a device that
could track an epicardial border on fluoroscopy. I had
to do a fluoroscopy and get the epicardial border and
put a marker on it and it would trace it. From that you
were supposed to derive information about cardiac
function and everything under the sun. He told me,
INTERVIEW/ANTHONY NICHOLAS DEMARIA
215
“That’s your project.” I said, “All right, Chief.” I
worked like crazy for about 3 months on the project.
One of my first publications was a review publication
on radarkymography, a little known field. I told him
that the radarkymography was just a bunch of bologna. It was the epicardial area and I couldn’t make
much out of it, despite what the manuscript said.
While I was trying to make it work, I read a paper
written by Harvey Feigenbaum in Indianapolis about
trying to do the same thing with ultrasound. I went to
Dean Mason, which is a hutzpah kind of thing, and
told him that I didn’t think the radarkymogram was
ever going to make it, but there was a technique with
ultrasound that might make it, so that we could see the
inside, not the outside of the heart. I wanted to learn
that. I wanted to go to Indianapolis where my buddy
from the Public Health Service was. I could stay at his
house and find out how to do this from Feigenbaum.
Dean Mason said, “If I were you, I wouldn’t waste my
time with this ultrasound thing. I can tell how bad
mitral stenosis is and that’s all it is used for.” I
prevailed upon him, telling him I really wanted to do
this. I felt it could amount to something. He was going
to humor me and said, “I’ll make a deal with you. I
have a drug study that I need to get done. If you agree
to do the drug study for me and get it done, then you
can go to Indianapolis and I’ll take some of the money
from the drug study and I’ll buy the echo machine.”
An M-mode echo only cost $10,000 in those days. I
told him we had a deal. The drug study turned out to
be perhexiline, which came and died because nobody
could figure out how it worked. Subsequently, somebody figured out that it was a calcium channel blocker.
The drug was probably a decent drug all the while.
In those days, Harvey Feigenbaum was “the king.”
He had people come to his lab for 1 week. He claims
that I gave the title “1-week wonder” to his training
program. Walter Henry from the NIH went there. Rich
Popp had been a fellow under Feigenbaum. Almost
anybody who was anybody in echocardiography had
made a pilgrimage to Indianapolis. I went for 1 week
to learn echocardiography. During the breaks, when
we weren’t examining patients, I read papers. Harvey
had assembled every manuscript that had ever been
written on cardiac ultrasound in 1 loose-leaf notebook.
There were only about 10 of them plus a big symposium in Acta Medica Scandinavia. That was the
world’s literature on echocardiography! I read the
world’s literature in 1 week. In addition to that, Harvey had managed to get the first connection of an echo
machine with a strip chart recorder. For the first time,
instead of taking 1 Polaroid picture and getting a
couple of beats on the picture, now he could get a strip
chart that he could scan around. Holy mackerel, this
was really going to amount to something!
I went back to Davis on fire with all this stuff and
talked to Dean Mason. He still thought I was a bit
nuts, but he was a man of his word. He got the
Smith-Kline echo machine for me. The only place we
could find to do echocardiography in the hospital was
in physical medicine and rehabilitation. Dean managed to get room in the basement, way down in God’s
216 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
never-never land, to put the machine in to do echocardiography. I felt like the lone ranger. I was trying to
find patients to perform echo on to get some examples
and to get some practice. Everybody summarily ignored me. One day during professor’s rounds, Dean
saw a woman in the coronary care unit. This elderly
woman had been admitted to the hospital 3 times over
the course of about 8 weeks with severe unremitting
chest pain, profound ST T-wave changes, and mitral
regurgitation. She was old and didn’t want catheterization. They were reluctant to catheterize her, because to do a coronary angiogram in 1971 was a big
deal. After rounds, I got my echo machine and took it
up to the coronary care unit to get an example of what
an echo looks like in a patient with mitral regurgitation. Of course she had the most incredible florid
flagrant hypertrophic cardiomyopathy. The next day
on rounds I showed the echo and told them what she
had. Dean Mason became a believer! He said, “My
God. This is unbelievable.” The echo showed clearly
what she had and everybody else had missed. There
was still a lot of skepticism. I was struggling to
convince people, and 1 day Zelis called me up and
said, “I know you are looking for patients. I’ve got a
good mitral stenosis over here. It’s not quite classic,
but it would be a nice one for you to echo. I’ve
scheduled her for cath.” I told him to send her over.
She had an atrial septal defect. From that point on,
Zelis never catheterized another patient who hadn’t
had an echo first.
Then we applied for a big grant on disease of the
muscle. I was going to do echoes in patients with
muscular dystrophy. We got the NIH grant. Now I had
enough money to hire a technician. Up to this point I
had been everything—the sonographer, the machine
washer, everything. Alex Newman, a PhD student, was
washing glassware in the cath lab. Everybody said he
was really good but was just trying to figure out what
he wanted to do with his life. I didn’t know him very
well, but I talked to him saying, “I’ve got a technician
position in ultrasound. It pays more than washing
glassware in the cath lab.” He immediately became
my echo tech. The 2 of us started doing research. We
wrote early papers on what nitroglycerin does to the
heart, what mitral prolapse does, and Wolff-Parkinson-White, and we performed the first myocardial
contrast echocardiography. We were spinning and
spinning. Those were magic years. With that, Dean
Mason called me and said, “You need to be on the
faculty.” He got me a full-time faculty position. I went
home and talked to Lori, “This is kind of fun. I like
this stuff. I think we ought to keep doing it.” She said,
“I like living here in California.” I told her, “Let’s
keep doing it and if worse comes to worse, I can
always go into practice. I’ll do it for a couple of
years.” It was a magic time in Davis. It was a really
productive place. Rich Miller was publishing like
crazy. Dean was working 24 hours a day. There was a
ton of stuff going on. I kept thinking that maybe I
ought to go into practice. One day a guy called from
Palos Verdes, California, looking for somebody to run
cardiology at their hospital and asked if I would be
JANUARY 15, 2005
FIGURE 4. AND receiving a plaque from Dr. Dean T. Mason, at
that time President of the American College of Cardiology, for
organizing the program of the Annual Scientific Sessions the
year that Dr. Mason was the College’s president.
interested. I went down and interviewed for the job
and they offered it to me. I was very ambivalent. I
decided that I would ask for the sky—a high salary,
help with my mortgage, all kinds of stuff. I would ask
for the moon and if they gave it to me, then of course
I couldn’t turn it down. If they said they wouldn’t do
it, I’d stay in academics. That was the way I could
resolve my ambivalence. I asked for the moon. After
a while they called me to apologize for the time that
had gone by. They were pleased to tell me that they
would grant me everything I requested, 100% of the
things I requested. At that point I realized that practice
was just not what I wanted to do. I never looked back
after that. I was very lucky, because I got to know
Harvey Feigenbaum by being a 1-week wonder. As he
taught courses and presided over meetings, I was one
of the people he would include. He helped me a lot.
Of course, I owe everything to Dean (Figure 4). He
was mentor par excellence. Dean was just spectacular.
Harvey helped a lot. Lo and behold, after a time Dean
was president of the American College of Cardiology
and somebody had to run the Annual Scientific Session. He was going to have Richard Miller run it, but
Rich left for Baylor to be Chief of Cardiology. Here
he was, a couple of months into his presidency and
didn’t have anybody to run the meeting. He grabbed
me and said, “You’ve got to be Chairman of the
Annual Scientific Session for the American College of
Cardiology.” What the hell did I know? I was still wet
behind the ears. We did it together. That year the
whole program was laid out on the floor of my house.
I had stacks of papers in different categories. I met a
lot of people and got to know them very well. That
helped.
WCR: How did the move to Lexington come about?
AND: That was kind of a “2 for 1.” I was a pretty
young guy. I was 37. They were talking to me about
being Chief of Cardiology. I liked the man who was
recruiting me, John Thompson, Chairman of Medicine. I was fascinated by the idea, so I visited there.
Then Davis had a problem with the cardiac surgical
program. There was a fair amount of tension in the air
at Davis. At some point in time, Dean was no longer
Chief of Cardiology and they asked me to take the job.
I couldn’t do it. I left. Lexington was a good place to
go. The Chairman of Medicine was young and energetic. He was rebuilding a department of medicine.
They were enormously generous in providing a recruitment package. It was a time for me to see if I
could do things my way.
WCR: That was 1981?
AND: Yes. Lexington is a nice town, a great place
to live. By this time, my wife and I had 3 kids and we
liked living in small towns. I had good reason to go
there. I was flattered. I was arrogant enough to think at
37 I could run my division. I was too stupid to realize
that I was probably too young to do that.
WCR: You recruited some good people there.
AND: Yes. I took Steve Nissen with me from Davis.
He was a resident at Davis and had already signed up
for the cardiology fellowship program there. When I
left, he came with me. John Waters came also. After
I got to Lexington, I recruited some really good people—Tom Wisenbaugh, a really good hemodynamist
Cindy Grines and Jon Elion who did really well. A
guy called me during Christmas vacation. My secretary said, “There’s some fellow candidate who wants
to talk to you. He happens to be visiting his in-laws in
Louisville. He might want to do a fellowship at
Kentucky. He wants to know if he can come over.” I
said, “What the hell. Things are slow. It’s Christmas.”
Paul Grayburn came over to visit. I was blown away.
I thought he was terrific. I quickly sent him to talk to
a couple of other guys who were around. I offered him
a job on the spot.
WCR: He told me you called his home and his wife
answered the telephone. She, being from Louisville,
accepted for him. When he got home that night she
told him where he was going for his cardiology fellowship.
AND: That’s actually the way it went down. He
knew darn well when he left that he was getting the
fellowship. Usually, after the interview you wait and
see all the candidates and make a decision. When he
was leaving, I told him not to take any other job unless
he talked to me first. He must have known that he was
accepted. The way, we selected fellows was that we’d
get all of our applications and the faculty would sit
around to discuss them at length and offer jobs. In this
particular instance, I made it fairly clear to Paul that I
was going to have the faculty sit around and deliberate
this, but all the votes weren’t going to be equal on that.
WCR: Paul did a little bit like you did. He interned
at a private hospital in Dallas, thinking he was going
to go into practice.
AND: He reminded me of myself. The thing about
Paul that really gets me is that when he applied to
Kentucky, he was exactly the way I was when I
applied to Davis. He hadn’t done a lick of research. I
just knew that there was no way in God’s earth he
could ever get an interview at University of Texas
Southwestern. Jim Willerson was running the program
INTERVIEW/ANTHONY NICHOLAS DEMARIA
217
FIGURE 5. AND as a Master in Cardiology in 1988.
there and would have looked at his resume just the
way he would have looked at mine. This guy comes to
Lexington and I say, “You’ve got a job here.” He
comes and he trains and we work with him for 3 years,
teaching him everything we knew. I said this guy is
going to stay here and he’s going to be a star and
really build this program. Willerson, with 1 telephone
call, recruited him back to Dallas. I thought that Willerson wouldn’t have given him the time of day 3 or 4
years earlier. Now all of a sudden Willerson says,
“You need to come back to Dallas.” Grayburn went
back. We lost him.
WCR: Your experience in Lexington turned out to
be a very good one?
AND: It was a terrific experience (Figure 5). We
made a name for ourselves in digital angiography.
Steve Nissen started his intravascular ultrasonic imaging career there. He and I worked together very
closely for 11 years. In the best tradition of mentorstudent relations, the student has now surpassed the
mentor. Cindy Grines did a couple of really important
studies and then moved on to work with Bill O’Neill
again. We had a pretty good clinical volume. We
trained a lot of good doctors. Many went into academics for only a few years and then drifted off regrettably. That was one of my regrets. It was very hard to
leave Lexington.
218 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
WCR: How did the San Diego opportunity come
about?
AND: John Ross called me. The institution had
recognized that they needed to strengthen the clinical
program in cardiology. They were looking for someone to be Chief of Clinical Cardiology. John asked me
about it. We liked Lexington, but San Diego was a
bigger and more visible program and I admired John
very much. The university had a lot of attraction about
it, and being back in California, where we had spent
10 years earlier, helped. I went and looked at the
position. I was leaning toward not going. I was happy
in Lexington and was going to stay put. While I was
looking, John made a decision to step down as Chief
of Cardiology. That left the division head open and
they made me a good offer. It was time. I stayed at
Davis for 10 years; I stayed in Lexington for 11 years.
It was time to move. So I recruited my assistant Kate
Greathouse, and my lab manager, Oi Ling Kwan, both
of whom I have worked with for ⬎20 years and
depend on greatly, to go west (Figure 6). Some people
move, some people don’t. I moved.
WCR: You went to San Diego in 1992. How many
faculty do you have there now in cardiology?
AND: It’s a relatively small program. We have 28
people on full-time faculty.
WCR: How many fellows do you have?
JANUARY 15, 2005
FIGURE 6. AND with his wife Lori and devoted assistants.
AND: It varies from year to year. We have 3 regular
training positions.
WCR: Per year?
AND: Per year. That’s 9. Then we have 2 interventional and 2 electrophysiology fellows. We have a
number of people who are there just on research
fellowships.
WCR: If you take all the faculty positions and all
the fellows, the secretaries, and the technicians, etc.,
you are talking about 150 people?
AND: I’d think closer to 125. As of July 1, I’ve
transitioned.
WCR: You’ve stepped down as head the division of
cardiovascular medicine. You are now director of the
Cardiovascular Center. What is that?
AND: Shortly after I got there, about 7 years ago,
Stuart Jamieson, Ken Chien, and I wanted to have a
cardiovascular center that would bring together all the
people at our institution who were working in cardiovascular disease. We were spread out all over the
place. We are not only in different areas—La Jolla
versus the Hillcrest area of San Diego— but even
within La Jolla at the Basic Science building and VA
Hospital. We had approached the dean about a Cardiovascular Center to make it a comprehensive entity.
WCR: That’s across departments?
AND: Yes. Along the way, Ken actually became
head of the Institute for Molecular Medicine. There’s
an institute at UCSD dedicated to molecular medicine—all aspects—not only cardiology (Ken is mostly
cardiology), but cancer, immunology, the whole ball
of wax. He’s still associated with the cardiovascular
center, but not in a primary position. We formed the
group. The triumvirate of Stuart Jamison, Chief of
Cardiac Surgery and me. As a research director, we
have Kirk Knowlton. He bridged the gap between Ken
Chien and me very nicely. Kirk is an excellent research guy, but he has clinical activities as well. We
were working on the center raising money.
About 8 to 10 months ago, Ed Holmes, the Dean,
called me and said, “Listen, if you can raise $30
million, I can put in $70 million. The hospital will
assume debt. Let’s build a cardiovascular center and
make it inpatient and outpatient. We’ll focus on clin-
ical and translational research. We’ll house cardiac
surgery, adult cardiology, the pulmonary vascular program, pulmonary hypertension of all types a stroke
program (Pat Lyden), all cardiovascular radiology,
preventive cardiology (Michael Criqui from the department of preventive medicine), peripheral vascular
surgery, and peripheral vascular medicine. The idea
was that we would construct a building that would
have inpatient and outpatient facilities and clinical
research facilities (but not basic science laboratories,
not hoods, not even small animals). The catheterization and electrophysiology labs would have adequate
backup space so investigators could tie them up doing
a research study. It’s not just a hospital. It will have all
the clinical facilities that you would want to have
available to do research projects—appropriate space
for noninvasive imaging, cardiovascular, radiology,
electrophysiology cath, or whatever you want for clinical research.
Ed Holmes thought it was a full-time job. It’s
probably close to that. More importantly, this position
will oversee adjudication resources— how is space
allotted, how are offices allocated, how is priority for
hospital beds decided. Those issues on space and
money required, in Ed Holmes’ opinion, and which I
couldn’t disagree with, that someone had to be free of
a vested interest in those decisions. The question was
what did I want to do? I thought it was a neat challenge to build a cardiovascular center, and one that
would fit in very well with my lifestyle at this point in
time. Being Editor of the Journal of the American
College of Cardiology obviously occupies a lot of my
effort. The Dean and I negotiated a bit and came to an
amicable agreement. As of July, I have accepted that
responsibility. Kirk Knowlton will assume the duties
of Chief of Cardiology. The goal is to have the building up and running in 4 years. It should take us 1 year
to plan it out and another 2 to 3 to get it built.
WCR: How are you enjoying the editorship of the
Journal of the American College of Cardiology?
AND: I love it. It’s been a great privilege to have
that opportunity. It is a unique opportunity to influence what people read, and therefore, perhaps what
they think. We have a sacred responsibility to the
authors to treat their work with the respect that they
put into it. I feel like I have my finger on the pulse of
cardiology. Anything that is going on is being submitted, and usually substantially before it surfaces as
something important. It’s a learning experience. It’s
an opportunity to influence my profession. It’s gives
me a certain creative opportunity in writing editor’s
pages and making decisions on how to put manuscripts together and organize them in certain ways. It’s
been a challenge in handling this transition from a
purely paper journal to one that’s now paper and
on-line, and will be progressively on-line.
WCR: What were some surprises you as you took
over the editorship?
AND: I really hadn’t anticipated the degree of serendipity that’s involved in what gets accepted for
publication and what doesn’t. We always joked from
my earliest days with Dean Mason that the biggest
INTERVIEW/ANTHONY NICHOLAS DEMARIA
219
variable was the reviewer. I fully anticipated that we
were going to have to allot for variability on the part
of a reviewer, but even with that I have been amazed
at the frequency with which critiques are opposite. In
addition to reviewer variation are other things that
come into play. We went back for a period of 1 year
and found that in just ⬍40% of our manuscripts, 1
reviewer ranked it in the top 2 (of 5) categories and the
other rejected it. That was in almost 40% of the
manuscripts. I was totally unprepared for that much
divergence in the way 2 learned clinicians would
assess a given manuscript. The reality of the volume
and timelines of work have also been striking. Every
day, more manuscripts turn up. It’s got to be handled
in a timely fashion. It’s not something you can put on
the side. The camaraderie with the associate editors
has been an unexpected side benefit. Working with
Ori Ben-Yehuda, Wilbur Lew, Sam Tsimikas, Greg
Feld, Kirk Knowlton, and Barry Greenberg has
bonded us closely.
WCR: I guess you get the question a lot and I’ve
gotten it a lot through the years: How much time do
you spend a week on the Journal of the American
College of Cardiology?
AND: About 25 hours a week. It varies up and
down. This week I won’t spend any. In an average
week I usually dedicate 11/2 entire weekend days at
least, and usually a bit more to just doing journal
things. I dedicate about 1 working day during the
week to preparing and handling our weekly meeting,
and then there’s a couple of hours each day depending
on different days. Nobody works a 40-hour week in
our business. We’re in a competitive business being
editors in this day and age. If you and I are competing
for the best papers, then one of the things we sell is
rapid response. One thing we have not talked about is
probably worth talking about. Part of my years in
Lexington was being President of the American College of Cardiology. I had come to that position by
Dean Mason making me the Scientific Sessions program chairman and then I got involved a lot in the
Government Relations Committee and private sector
relations. At a relatively young age (I think I was the
youngest president ever), I got the opportunity to be
president. When they called me and asked me if I was
willing to be nominated, I said, “I’m too young for
this, but, on the other hand, you may never ask me
again.” I went and did it.
WCR: You were how old?
AND: I was 45 when I was president (1988 to
1989), but I was nominated at age 42. That was a
marvelous experience. The year I was president we
laid the foundation in a major way for some of the
advocacy programs, which I think have been very
meaningful to clinical cardiologists, while preserving
the important educational mission and the support for
scholarship. I had the opportunity to travel for the
college. It’s a bit of a bully pulpit to tell people what
you think cardiology should be like in those President’s Pages. I found that thoroughly enjoyable. If you
had told me when I sat in front of Dean Mason, or
even when I graduated medical school, “You know
220 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
what? You’re graduating from New Jersey College of
Medicine. Someday you’re going to be president of
the echo society, president of the American College of
Cardiology, and the Editor of the Journal of the American College of Cardiology.” I would have told you,
“You’ve got to be out of your mind. That could never
happen.”
WCR: Of all the things you’ve done professionally,
what are you most proud of?
AND: The people I’ve trained. Of the people I’ve
had a chance to influence, some are famous—Steve
Nissen and Paul Grayburn. Many are important people
in medical schools—Bill Bommer at Davis, Michael
Smith at Kentucky, Ari Ben Yehuda, Ajit Raisinghani,
Shami Mahmud, and Sam Tsimikas here at UCSD.
I’ve trained many foreign physicians. I’ve had the
opportunity to have colleagues from countries around
the world who’ve worked in the lab for 1 or 2 years
and gone on to be major investigators. Far and away,
that’s been the thing I’ve enjoyed the most. I most
enjoy when fellows write their first papers and hold
them in their hands. It doesn’t get much better than
that.
WCR: What happened to your 3 sisters?
AND: Two are single and still live in New Jersey.
Adele is in business and Carol is a schoolteacher. My
oldest sister Barbara Hupe lives in Houston, Texas
now. She’s also a schoolteacher. Her husband died an
untimely death of a brain tumor.
WCR: So 2 of the 3 are teachers?
AND: Right.
WCR: Did all 3 go to college?
AND: Yes.
WCR: That’s quite remarkable.
AND: Especially for my sisters, because the philosophy at the time was men needed to go to college, but
for women, it was a luxury. They went.
WCR: Tell me about your immediate family.
AND: My wife is the best thing about me. Everybody likes Lori. She’s marvelous. We have been partners in everything I’ve done, including the new UCSD
Cardiovascular Center. She’s wise, insightful, and a
tower of strength.
WCR: What was her premarital name?
AND: Horn. We’ve been married 37 years now.
I’ve been very fortunate. That’s the best decision I
ever made in my life. Going to Davis was the second
best. Doing echocardiography was probably my third
best decision. Our oldest daughter Christine describes
herself as a recovering lawyer, who is now getting a
PhD in sociology. She and her husband Matt have a
son, Jackson, and are having a second child. Christine
went to the University of Kentucky Law School. Our
middle child Anthony is a very successful lawyer in
Fresno, California. He and his wife Kim have 3 children, Alec, Blake, and Cameron. He’s phenomenally
successful. He wins almost all his cases. He did very
well in law school at USD Law School. Our son
Jonathan is involved in a number of different activities. He’s very bright but hasn’t quite decided where
he wants his career path to go yet (Figures 7 and 8).
JANUARY 15, 2005
FIGURE 7. AND and his wife and three offspring.
Anaheim. They flew my mother to Los Angeles and
put her in a limo, and brought her to the hotel in
Anaheim. I got a call that my mother was at the hotel.
I went to the lobby to meet her. I said, “Mom, welcome. How’re you doing?” She said, “I’m fine. I’m
fine.” Very soon she asked, “Where’s the telephone?
I’ve got to make a telephone call.” I was taken aback.
She called her best friend, Rose Yazwinski, and said,
“Rose, you’re not going to believe this. They had a car
just for me, a very big one! Rose, you could put 4 or
5 people in the car. I was the only one in it.” I honestly
believe that that limo made my mother think that
maybe I’d done all right. And all those years they
suspected that I just couldn’t quite make it out of
medical school!
WCR: Tony, many thanks for pouring your soul out.
It was great.
MOST IMPORTANT PUBLICATIONS SELECTED BY AND
FIGURE 8. AND and his wife with their grandchildren.
WCR: Is there anything you’d like to talk about that
we have neglected to talk about?
AND: Two of my favorite stories. The first story
happened when I was at Davis, my father called me
and said, “Anthony, I’ve got to talk to you man to
man. The people around the neighborhood are beginning to ask questions. How come Anthony is hanging
around the medical school so long? It’s getting embarrassing. They wonder if you haven’t finished?” I
told him, “Dad, I’m not going to do practice. I’m
going to be an academic.” He said, “Anthony, you’ve
got a wife and 2 kids. You’ve got to stop fooling
around here. I’m calling you to tell you this very day
that Dr. Fabrielle, who used to be on 46th and Avenue
C, died, and his wife came in to tell me that you can
have the practice if you want it.” I told him, “Dad, I’m
going to stay at the medical school.”
The second story. When as President of the American College of Cardiology, I invited my mother to
come to the annual meeting. (The College lets the
president bring their family to see their child conduct
the meeting at the Convocation.) The meeting was in
1. Massumi RA, Mason DT, Amsterdam EA, DeMaria AN, Miller RR, Scheinman MM, Zelis R. Ventricular fibrillation and tachycardia after intravenous
atropine for treatmentof bradycardias. N Engl J Med 1972;287:336 –338.
14. DeMaria AN, King JF, Bogren HG, Lies JE, Mason DT. The variable
spectrum of echocardiographic manifestations of the mitral valve syndrome.
Circulation 1974;50:33– 41.
15. DeMaria AN, Vera Z, Amsterdam EA, Mason DT, Massumi RA. Disturbances of cardiac rhythm and conduction inducted by exercise: diagnostic,
prognostic and therapeutic implications. Am J Cardiol 1974;33:732–736.
16. DeMaria AN, Vismara LA, Auditore K, Amsterdam EA, Zelis R, Mason DT.
Effects of nitroglycerin on left ventricular cavitary size and cardiac performance
determined by ultrasound in man. Am J Med 1974;57:754 –760.
18. DeMaria AN, King JF, Salel AF, Caudill CC, Miller RR, Mason DT.
Echography and phonography of acute aortic regurgitation in bacterial endocarditis. Ann Intern Med 1975;82:329 –335.
20. DeMaria AN, Lies JE, King JF, Miller RR, Amsterdam EA, Mason DT.
Echocardiographic assessment of atrial transport, mitral movement, and ventricular performance following electroversion of supraventricular arrhythmias. Circulation 1975;51:273–282.
21. DeMaria AN, Vismara LV, Miller RR, Neumann A, Mason DT. Unusual
echographic manifestations of right and left heart myxomas. Am J Med 1975;59:
713–720.
31. DeMaria AN. The syndrome of mitral valve prolapse: problems and perspectives. Ann Intern Med 1976;85(4):525–526.
32. DeMaria AN, Amsterdam EA, Vismara LA, Neumann A, Mason DT.
Arrhythmias in the mitral valve prolapse syndrome: prevalence, nature and
frequency. Ann Interm Med 1976;84:656 – 660.
33. DeMaria AN, Miller RR, Amsterdam EA, Markson W, Mason DT. Mitral
valve early diastolic closing velocity in the echocardiogram: relation to sequential
diastolic flow and ventricular compliance. Am J Cardiol 1976;37:693–700.
36. DeMaria AN, Vera Z, Neumann A, Mason DT. Alterations in ventricular
contraction pattern in the Wolff-Parkinson-White syndrome: detection by echocardiography. Circulation 1976;53:249 –257.
59. DeMaria AN, Neumann A, Lee G, Mason DT. Echocardiographic identification of the mitral valve prolapse syndrome. Am J Med 1977;62:819 –29.
97. DeMaria AN, Neumann A, Lee G, Fowler W, Mason DT. Alterations in
ventricular mass and performance induced by exercise training in man evaluated
by echocardiography. Circulation 1978;57:237–244.
101. DeMaria AN, Oliver LE, Bogren HG, George L, Mason DT. Apparent
reduction of aortic and left heart chamber size in atrial septal defect. Am J Cardiol
1978;42:545–550.
119. Sahn DJ, DeMaria AN, Kisslo J, Weyman A. Recommendations regarding
quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978;58(6):1072–1083.
125. DeMaria AN, Bommer W, Neumann A, Weinert L, Bogren H, Mason DT.
Identification and localization of aneurysms of the ascending aorta by crosssectional echocardiography. Circulation 1979;59:755–761.
129. DeMaria AN, Neumann A, Bommer W, Weinert L, Grehl T, Amsterdam
EA, Mason DT. Left ventricular thrombi by cross-sectional echocardiography.
Ann Inten Med 1979;90:14 –18.
130. DeMaria AN, Neumann A, Schubart PJ, Lee G, Mason DT. Systemic
correlation of cardiac chamber size and ventricular performance determined with
echocardiography and alterations in heart rate in normal persons. Am J Cardiol
1979;43:1–9.
INTERVIEW/ANTHONY NICHOLAS DEMARIA
221
148. DeMaria AN, Bommer W, Joye JA, Lee G, Bouteller J, Mason DT. Value and
limitations of cross-sectional echocardiography of the aortic valve in the diagnosis
and quantification of valvular aortic stenosis. Circulation 1980;62:304 –312.
150. DeMaria AN, Bommer W, Lee G, Mason DT. Value and limitations of
two-dimensional echocardiography in assessment of cardiomyopathy. Am J Cardiol 1980;46:1224 –1231.
185. Low RI, Takeda P, Lee G, Mason DT, Awan NA, DeMaria AN. Effects of
diltiazem-induced calcium blockade upon exercise capacity in effort angina due
to chronic coronary artery disease. Am Heart J 1981;101:713–718.
195. Woythaler JN, Singer SL, Kwan OL, Meltzer RS, Reubner B, Bommer W,
DeMaria AN. Accuracy of echocardiography versus electrocardiography in detecting left ventricular hypertrophy: comparison with post-mortem mass measurements. J Am Coll Cardiol 1982;2:305–311.
200. Nissen SE, Booth D, Waters J, Fassas T, DeMaria AN. Evaluation of left
ventricular contractile pattern by intravenous digital subtraction ventriculography: comparison with cineangiography and assessment of interobserver variability. Am J Cardiol 1983;52(10):1293–1298.
204. DeMaria AN, Bommer W, Kwan OL, Riggs K, Smith MD, Waters J. In vivo
correlation of thermodilution cardiac output and videodensitometric indicatordilution curves obtained from contrast two-dimensional echocardiograms. J Am
Coll Cardiol 1984;3:999 –1004.
207. Johnson GL, Kwan OL, Cottrill CM, DeMaria AN. Detection and quantitation of right ventricular outlet obstruction secondary to aneurysm of the membranous ventricular septum by combined two-dimensional echocardiography:
continuous-wave Doppler ultrasound. Am J Cardiol 1984;53:1476 –1477.
208. Johnson GL, Kwan OL, Handshoe S , Noonan J, DeMaria AN. Accuracy of
combined two-dimensional echocardiography and continuous wave Doppler recording in the estimation of pressure gradient in right ventricular outlet obstruction. J Am Coll Cardiol 1984;3:1013–1018.
213. Smith MD, Kwan OL, Reiser HJ, DeMaria AN. Superior intensity and
reproducibility of SHU-454, a new right heart contrast agent. J Am Coll Cardiol
1984;3:992–998.
219. Friedman BJ, Waters JS, Kwan OL, DeMaria AN. Comparison of magnetic
resonance imaging and echocardiography in determination of cardiac dimensions
in normals. J Am Coll Cardiol 1985;5:1369 –1376.
221. Smith MD, Dawson PL, Elion JL, Booth DC, Handshoe R, Kwan OL, Earle
GF, DeMaria AN. Correlation of continuous wave Doppler velocities with
cardiac catheterization gradients: an experimental model of aortic stenosis. J Am
Coll Cardiol 1985;6:1306 –1314.
229. Friedman BJ, Drinkovic N, Miles H, Shih WJ, Mazzoleni A, DeMaria AN.
Assessment of left ventricular diastolic function: comparison of Doppler echocardiography and gated blood pool scintigraphy. J Am Coll Cardiol 1986;8(6):
1348 –1354.
230. Grayburn PA, Smith MD, Handshoe R, Friedman BJ, Handshoe S, DeMaria
AN. Comparative accuracy of pulsed Doppler, echo and auscultation in the
detection of aortic insufficiency. Ann Int Med 1986;l04:599 – 605.
235. Nissen SE, Elion JL, Booth DC, Evans J, DeMaria AN. Value and limitations of computer analysis of digital subtraction angiography in the assessment of
coronary flow reserve. Circulation 1986;73:562–571.
240. Smith MD, Handshoe R, Handshoe S, Kwan OL, DeMaria AN. Comparative accuracy of 2-D echocardiography and Doppler pressure half-time methods
in assessing severity of mitral stenosis in patients with and without prior commissurotomy. Circulation 1986;73:100 –107.
241. Smith MD, Kwan OL, DeMaria AN. Value and limitations of continuouswave Doppler echocardiography in estimating severity of valvular stenosis.
JAMA 1986;255:3145–3151.
244. Wisenbaugh T, Nissen S, DeMaria AN. Mechanics of postextrasystolic
potentiation in normal subjects and patients with valvular heart disease. Circulation 1986;74:10 –20.
248. DeMaria AN, Wisenbaugh TW. Identification and treatment of diastolic
dysfunction: role of transmitral Doppler recordings. J Am Coll Cardiol 1987;9:
1106 –1107.
251. Grayburn PA, Handshoe R, Smith MD, Harrison MR, DeMaria AN. Quantitative assessment of the hemodynamic consequences of aortic regurgitation by means
of continuous wave Doppler recordings. J Am Coll Cardiol 1987;10:135–141.
252. Grayburn PA, Nissen SE, Elion JL, Evans J, DeMaria AN. Quantitation of
aortic regurgitation by computer analysis of digital subtraction angiography. J Am
Coll Cardiol 1987;10:1122–1127.
253. Grayburn PA, Smith MD, Gurley JC, Booth DC, DeMaria AN. Effect of
aortic regurgitation on the assessment of mitral valve orifice area by Doppler
pressure half-time. Am J Cardiol 1987;60:322–326.
260. Nissen SE, Elion JL, Grayburn P, Booth DC, Wisenbaugh TW, DeMaria
AN. Determination of left ventricular ejection fraction by computer densitometric
analysis of digital subtraction angiography: experimental validation and correlation with area-length methods. Am J Cardiol 1987;59:675– 680.
263. Rovai D, Nissen SE, Elion JL, Smith MD, L’Abbate A, Kwan OL, DeMaria
AN. Contrast echo washout curves application of basic principles of indicator
dilution theory and calculation of ejection fraction. J Am Coll Cardiol 1987;10:
125–134.
273. Gillespie MN, Booth DC, Friedman BJ, Cunningham MR, Jay M, DeMaria
AN. fMLP provokes coronary vasoconstriction and myocardial ischemia in
rabbits. Am J Physiol 1988;254:H481–H486.
274. Grayburn PA, Smith MD, Harrison MR, Gurley JC, DeMaria AN. Pivotal
role of aortic valve area calculation by the continuity equation for Doppler
222 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
assessment of aortic stenosis in patients with combined aortic stenosis and
regurgitation. Am J Cardiol 1988;61:376 –381.
275. Gurley JC, Nissen SE, Booth DC, Harrison M, Grayburn P, Elion JL,
DeMaria AN. Comparison of simultaneously performed digital and film-based
angiography in the assessment of coronary artery disease. Circulation 1988;
78(6):1411–1420.
279. Harrison MR, Smith MD, Nissen SE, Grayburn PA, DeMaria AN. Use of
exercise Doppler to evaluate cardiac drugs: effects of propranolol and verapamil
on aortic blood flow velocity and acceleration. J Am Coll Cardiol 1988;1(5):
1002–1009.
289. DeMaria AN, Smith MD, Harrison MR. Clinical significance of in-vitro and
in-vivo experimental findings regarding Doppler flow velocity recordings. J Am
Coll Cardiol 1989;13:1682–1686.
291. Grines CL, Nissen SE, Booth DC, Branco MC, Gurley JC, Bennett KA,
DeMaria AN, KAMIT Study Group. A new thrombolytic regimen for acute
myocardial infarction using half dose tissue plasminogen activator with full dose
streptokinase. J Am Coll Cardiol 1989;14(3):573–580.
293. Harrison MR, Clifton GD, Sublett KL, DeMaria AN. Effect of heart rate on
Doppler indexes of systolic function in humans. J Am Coll Cardiol 1989;14(4):
929 –935.
296. Rovai D, L’Abbate A, Lombardi M, Nissen SE, Marzilli M, Distante A,
Ferdeghini EM, DeMaria AN. Nonuniformity of the transmural distribution of
coronary blood flow during the cardiac cycle. In-vivo documentation by contrast
echocardiography. Circulation 1989;79(1):179 –187.
297. Smith MD, Kwan OL, Elion JL, McClure R, DeMaria AN. Left heart
opacification with peripheral venous injection of a new saccharide echo contrast
agent in dogs. J Am Coll Cardiol 1989;13:1622–1628.
299. Spain MG, Smith MD, Graybum PA, Harlamert EA, O’Brien M, DeMaria
AN. Quantitative assessment of mitral regurgitation by Doppler color flow
imaging: angiographic and hemodynamic correlations. J Am Coll Cardiol 1989;
13:585–590.
300. Wisenbaugh TW, Harlamert EA, DeMaria AN. Relation of left ventricular
filling dynamics to alterations in load and compliance in patients with and without
pressure-overload hypertrophy. Circulation 1989;1(1):101–106.
303Berk MR, Xie G, Kwan OL, Knapp C, Evans J, Kotchen T, Kotchen JM,
DeMaria AN. Reduction of left ventricular preload by lower body negative
pressure alters Doppler transmitral filling patterns. J Am Coll Cardiol 1990;16(6):
1387–1392.
308. Garrahy PG, Kwan OL, Booth DC, DeMaria AN. Assessment of abnormal
systolic intraventricular flow patterns by Doppler imaging in patients with left
ventricular dyssynergy. Circulation 1990;82:95–104.
310. Grines CL, Booth DC, Nissen SE, Gurley JC, Bennett KA, O’Connor WN,
DeMaria AN. Mechanism of acute myocardial infarction in patients with prior
coronary artery bypass grafting and therapeutic implications. Am J Cardiol
1990;65:1292–1296.
312. Gurley JC, Nissen SE, Elion JL, Booth DC, DeMaria AN. Determination of
coronary flow reserve by digital angiography: validation of a practical method not
requiring power injection or electrocardiographic gating. J Am Coll Cardiol
1990;16:190 –197.
313. Nissen SE, Grines CL, Gurley JC, Sublett KL, Haynie DP, Diaz C, Booth
DC, DeMaria AN. Application of a new phased-array ultrasound imaging catheter in the assessment of vascular dimensions: in-vivo comparison to cineangiography. Circulation 1990;81:660 – 666.
320. Grines CL, Booth DC, Nissen SE, Gurley JC, Bennett KA, DeMaria AN.
Acute effects of parenteral beta-blockade on regional ventricular function of
infarct and noninfarct zones after reperfusion therapy in humans. J Am Coll
Cardiol 1991;17:1382–1387.
323. Grines CL, Nissen ST, Booth DC, Gurley JC, Chelliah N, Wolf R, Blankenship J, Branco MC, Bennett K, DeMaria AN, Kentucky Acute Myocardial
Infarction Trial (KAMIT) Group. A prospective, randomized trial comparing
combination half-dose tissue-type plasminogen activator and streptokinase with
full-dose tissue-type plasminogen activator. Circulation 1991;84:540 –549.
325. Harrison MR, Clifton GD, Pennell AT, DeMaria AN. Effect of heart rate on
left ventricular diastolic transmitral flow velocity patterns assessed by Doppler
echocardiography in normal subjects. Am J Cardiol 1991;67:622– 627.
329. Nissen SE, Gurley JC, Grines CL, Booth DC, McClure R, Berk M, Fischer
C, DeMaria AN. Intravascular ultrasound assessment of lumen size and wall
morphology in normal subjects and patients with coronary artery disease. Circulation 1991;84:1087–1099.
330. Smith MD, Harrison MR, Pinton R, Kandil H, Kwan OL, DeMaria AN.
Regurgitant jet size by transesophageal compared with transthoracic Doppler
flow imaging. Circulation 1991;83:79 – 86.
338. King DL, Harrison MR, King D Jr, Gopal AS, Kwan OL, DeMaria AN.
Ultrasound beam orientation during standard two-dimensional imaging: assessment by three-dimensional echocardiography. J Am Soc Echocardiogr 1992;5(6):
569 –576.
339. King DL, Harrison MR, King D Jr, Gopal AS, Martin R, DeMaria AN.
Improved reproducibility of left atrial and left ventricular measurements by
guided three-dimensional echocardiography. J Am Coll Cardiol 1992;20(5):
1238 –1245.
342. Rovai D, Ghelardini G, Lombardi M, Trivella MG, Nevola E, Taddei L,
Michelassi C, Distante A, DeMaria AN. Myocardial washout of sonicated iopamidol reflects coronary blood flow in the absence of autoregulation. J Am Coll
Cardiol 1992;20(6):1417–1424.
JANUARY 15, 2005
355. Brotheron T, Pollard T, Simpson P, DeMaria AN. Classifying tissue and
structure in echocardiograms: hierarchy of fuzzy-logic based neural networks
automate analysis. Engineering Med Bio 1994;13(5):754 –759.
359. Xie GY, Berk MR, Smith MD, DeMaria AN. A simplified method for
determining regurgitant fraction by Doppler echocardiography in patients with
aortic regurgitation. J Am Coll Cardiol 1994;24(4):l041–1045.
360. Xie GY, Berk MR, Smith MD, Gurley JC, DeMaria AN. Prognostic value
of Doppler transmitral flow patterns in patients with congestive heart failure.
J Am Coll Cardiol 1994;24(I):132–139.
368. Nozaki S, DeMaria AN, Helmer GA, Hammond HK. Detection of regional
left ventricular dysfunction in early pacing-induced heart failure using ultrasonic
integrated backscatter. Circulation 1995;92:2676 –2682.
378. DeMaria AN, Lee TH, Leon DF, Ullyot DJ, Wolk MJ, Mills PS, Fay SC,
Brown JH, Flatau CN, Bodycombe DP. Effect of managed care on cardiovascular
specialists: involvement, attitudes and practice adaptations. J Am Coll Cardiol
1996;28(7):1884 –1895.
379. Giordano F, Ping P, McKirnan MD, Nozaki S, DeMaria AN, Dillman W,
Mathieu-Costello O, Hammond HK. Intracoronary gene transfer of fibroblast
growth factor-5 increases blood flow and contractile function in an ischemic
region of the heart. Nature Med 1996;2(5):534 –539.
382. Kimura BJ, Russo RJ, Bhargava V, McDaniel M, Peterson KL, DeMaria
AN. Atheroma morphology and distribution in proximal left anterior descending
coronary artery: in vivo observations. J Am Coll Cardiol 1996;27(4):825– 831.
386. Yamagishi M, Nissen SE, Booth DC, Gurley JC, Koyama J, Kawano S,
DeMaria AN. Coronary reactivity to nitroglyerin: intravascular ultrasound evidence for the importance of plaque distribution. J Am Coll Cardiol 1996;25:224 –
230.
398. Ohmori K, Cotter B, Kwan OL, Mizushige K, DeMaria AN. Relation of
contrast echo intensity and flow velocity to the amplification of contrast opacification produced by intermittent ultrasound transmission. Am Heart J 1997;
134(6):1066 –1074.
406. Galiuto L, Ignone G, DeMaria AN. Heterogeneity of contraction and
relaxation velocities of normal human myocardium. A quantitative study using
pulsed wave tissue Doppler echocardiography. Am J Cardiol 1998;81(5):609 –
614.
415. Yasu T, Schmid-Schonbein GW, Cotter B, DeMaria AN. Flow dynamics of
QW7437, a new dodecafluoropentane ultrasound contrast agent, in the microcirculation: microvascular mechanisms for persistent tissue echo enhancement. J Am
Coll Cardiol 1999;34(2):578 –586.
423. Kimura BJ, Scott R, Willis CL, DeMaria AN. Accuracy and cost-effectiveness of single-view echocardiographic screening for suspected mitral valve prolapse. Am J Med 2000;108:331–333.
424. Masugata H, Cotter B, Peters B, Ohmori K, Mizushige K, DeMaria AN.
Assessment of coronary stenosis severity and transmural perfusion gradient by
myocardial contrast echocardiography: comparison of gray scale B-mode with
power Doppler imaging. Circulation 2000;102:1427–1433.
432. Kimura BJ, Bocchicchio M, Willis CL, DeMaria AN. Screening cardiac
ultrasonographic examination in patients with suspected cardiac disease in the
emergency department. Am Heart J 2001;142(2):324 –330.
434. Lafitte S, Masugata H, Peters B, Togni M, Strachan M, Kwan OL, DeMaria
AN. Comparative value of dobutamine and adenosine stress in the detection of
coronary stenosis with myocardial contrast echocardiography. Circulation 2001;
103:2724 –2730.
435. Laffite S, Masugata H, Peters B, Togni M, Strachan M, Yao B, Kwan OL,
DeMaria AN. Accuracy and reproducibility of coronary flow rate assessment by
real-time contrast echocardiography: in vitro and in vivo studies. J Am Soc
Echocardiogr 2001;14:1010 –1019.
436. Leistad E, Ohmori K, Peterson TA, Christensen G, DeMaria AN. Quantitative assessment of myocardial perfusion during graded coronary artery stenoses
by intravenous myocardial contrast echocardiography. J Am Coll Cardiol 2001;
37:624 – 631.
437. Mahmud E, Raisinghani A, Keramati S, Auger W, Blanchard DG, DeMaria
AN. Dilation of the coronary sinus on echocardiogram: prevalence and signifi-
cance in patients with chronic pulmonary hypertension. J Am Soc Echocardiogr
2001;14:44 – 49.
438. Maisel AS, Koon J, Krishnaswamy P, Kasenegra R, Clopton P, Gardetto N,
Morrissey R, DeMaria A. Utility of B-natriuretic peptide (BNP) as a rapid, point
of care test for screening patients undergoing echocardiography for left ventricular dysfunction. Am Heart J 2001;141:367–374.
440. Masugata H, Lafitte S, Peters B, Strachan GM, DeMaria AN. Comparison
of real-time and intermittent triggered myocardial contrast echocardiography for
quantification of coronary stenosis severity and transmural perfusion gradient.
Circulation 2001;104:1550 –1556.
442. Masugata H, Peters B, Lafitte S, Strachan GM, Ohmori K, DeMaria AN.
Quantitative assessment of myocardial perfusion during graded coronary stenosis by
rReal-time myocardial cContrast echo refilling curves. J Am Coll Cardiol 2001;37:
262–269.
443. Ohmori K, Cotter B, Leistad E, Bhargava V, Wolf PL, Mizushige K,
DeMaria AN. Assessment of myocardial reperfusion by intravenous myocardial
contrast echocardiography: analysis of the intensity and texture of opacification.
Circulation 2001;103:2021–2027.
458. Kimura BJ, Amundson SA, Willis CL, Gilpin EA, DeMaria AN. Usefulness
of a hand-held ultrasound device for bedside examination of left ventricular
function. Am J Cardiol 2002;90:1038 –1039.
460. Lafitte S, Higashiyama A, Masugata H, Peters B, Strachan M, Kwan OL,
DeMaria AN. Contrast echocardiography can assess risk area and infarct size
during coronary occlusion and reperfusion: experimental validation. J Am Coll
Cardiol 2002;39(9):1546 –1554.
461. Lubien E, DeMaria AN, Krishnaswamy P, Clopton P, Koon J, Kazanegra R,
Gardetto N, Wanner E, Maisel AS. Utility of B-natriuretic peptide in detecting
diastolic dysfunction: comparison with Doppler velocity recordings. Circulation
2002;105:595– 601.
464. Raisinghani A, Mahmud E, Sadeghi M, Peters B, Strachan GM, Huynh T,
Blanchard D, DeMaria AN. Paradoxical inferior-posterior wall systolic expansion
in patients with end-stage liver disease. Am J Cardiol 2002;89:626 – 629.
477. Kimura BJ, DeMaria AN. Time requirements of the standard echocardiogram: implications regarding limited studies. J Am Soc Echocardiogr 2003;
16(10):1015–1018.
478. Kimura BJ, Fowler SJ, Nguyen DT, Amundson SA, DeMaria AN. Detection
of early carotid arterial atherosclerosis by briefly trained physicians using a
hand-held ultrasound device. Am J Cardiol 2003;92(2):239 –240.
480. Kunichika H, Peters B, Cotter B, Masugata H, Kunichika N, Wolf PL,
DeMaria AN. Visualization of risk-area myocardium as a high-intensity, hyperenhanced “hot spot” by myocardial contrast echocardiography following coronary reperfusion. J Am Coll Cardiol 2003;42(3):552–557.
482. Masugata H, Fujita N, Kondo I, Peters B, Ohmori K, Mizushige K, Kohno
M, DeMaria AN. Assessment of right ventricular perfusion after right coronary
artery occlusion. J Am Coll Cardiol 2003;41(10):1823–1830.
484. Raisinghani A, Wei K, Crouse LJ, Villanueva F, Feigenbaum H, Schiller
NB, Weiss J, Zaqvi TZ, Siegel R, Monaghan M, Goldman JH, DeMaria AN.
Myocardial contrast echocardiography with triggered ultrasound does not cause
ventricular premature beats: evidence from PB127 studies. J Am Soc Echocardiogr 2003;10:1037–1042.
485. Wei K, Crouse L, Weiss J, Villanueva F, Schiller NB, Naqvi TZ, Siegel R,
Monaghan M, Goldman J, Aggarwal P, Feigenbaum H, DeMaria A. Comparison
of usefulness of dipyridamole stress myocardial contrast echocardiography to
technetium-99m sestamibi single-photon emission computed tomography for
detection of coronary artery disease (PB127 multicenter Phase 2 trial Results).
Am J Cardiol 2003;91:1293–1298.
488. Nissen SE, Tuzeu EM, Schoenhagen P, Brown BG, Ganz P, Vogel RA,
Crowe T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria AN. Effect of
intensive compared with moderate lipid-lowering therapy on progression of
coronary atherosclerosis—a randomized controlled trial. JAMA 2004;291(9):
1071–1080.
INTERVIEW/ANTHONY NICHOLAS DEMARIA
223
Relation of Impaired Thrombolysis In Myocardial
Infarction Myocardial Perfusion Grades to Residual
Thrombus Following the Restoration of Epicardial
Patency in ST-Elevation Myocardial Infarction
Ajay J. Kirtane, MD, Aaron Weisbord, MD, Dimitrios Karmpaliotis, MD,
Sabina A. Murphy, MPH, Robert P. Giugliano, MD, SM, Christopher P. Cannon, MD,
Elliott M. Antman, MD, E. Magnus Ohman, MD, Matthew T. Roe, MD, MHS,
Eugene Braunwald, MD, and C. Michael Gibson, MS, MD, for the TIMI Study Group
Clinical and angiographic data were analyzed from
929 patients who had ST-elevation myocardial infarction and open epicardial arteries after fibrinolytic
therapy. Residual angiographically evident thrombus
was associated with more frequent Thrombolysis In
Myocardial Infarction (TIMI) grade 2 flow (33.6% vs
26.8%, p ⴝ 0.03), higher corrected TIMI frame counts
(34 vs 31 frames, p ⴝ 0.0003), and lower TIMI
myocardial perfusion grades (43.0% vs 32.0% TIMI
myocardial perfusion grades 0/1, p ⴝ 0.001) among
all patients and among patients who had TIMI grade
3 flow (33.5% vs 26.0% TIMI myocardial perfusion
grades 0/1, p ⴝ 0.043). In multivariate analyses,
angiographically evident thrombus was associated
with higher corrected TIMI frame counts and worsened myocardial perfusion independent of clinical
and angiographic covariates, including TIMI grade 3
flow. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:224 –227)
atients who have acute coronary syndromes may
have nonocclusive and angiographically evident
P
thrombus (AET) in the culprit coronary artery despite
reperfusion therapy. In patients who have non–STelevation acute coronary syndromes, visualization of
thrombus by angioscopy and by angiography have
been associated with high levels of troponin, and AET
has been associated with impaired myocardial perfusion.1,2 Residual thrombus after coronary stenting has
been associated with recurrent ischemic events.3 One
potential pathophysiologic link between AET and adFrom the Department of Medicine, Beth Israel Deaconess Medical
Center and the TIMI Study Group, Brigham & Women’s Hospital,
Harvard Medical School, Cambridge, Massachusetts; the University
of North Carolina, Chapel Hill, North Carolina; and the Duke Clinical
Research Institute, Durham, North Carolina. This study was supported
in part by grants from Millennium Pharmaceuticals, Cambridge,
Massachusetts, and Schering-Plough Research Institute, Kennilworth,
New Jersey; (Integrilin and Tenecteplase in Acute Myocardial Infarction); Merck and Co., Blue Bell, Pennsylvania (FASTER); Aventis
Pharma, Antony, France (ENTIRE Thrombolysis In Myocardial Infarction 23); Centocor and Eli Lilly Inc., Malvern, Pennsylvania; and
Indianapolis, Indiana (Thrombolysis In Myocardial Infarction 14). Dr.
Gibson address is: 350 Longwood Avenue, First Floor, Boston,
Massachusetts 02115. E-mail: mgibson@timi.org. Manuscript received June 8, 2004; revised manuscript received and accepted
September 7, 2004.
224
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
verse outcomes is microvascular dysfunction secondary to distal embolization and the release of vasoactive
substances that may arise in patients who have a
greater thrombus burden.4 We hypothesized that the
presence of AET in patent arteries after fibrinolytic
therapy in patients who have ST-elevation myocardial
infarction would be associated with impaired myocardial perfusion, independent of epicardial flow.
•••
Clinical and angiographic data were drawn from
patients with ST-elevation myocardial infarction who
had been enrolled in the Thrombolysis In Myocardial
Infarction (TIMI) 14,5 Integrilin and Tenecteplase in
Acute Myocardial Infarction,6 Enoxaparin and
Thrombolysis Reperfusion for Acute Myocardial Infarction Treatment [ENTIRE] TIMI 23,7 and Fibrinolytics and Aggrastat with ST segment resolution
[FASTER]8 trials of fibrinolytic therapy for ST-elevation myocardial infarction. Angiography was performed at 60 minutes after fibrinolytic administration.
Angiograms were analyzed off-line in a core laboratory (PERFUSE Core Laboratory). Analysis was restricted to patients who had been enrolled in these
trials in whom the TIMI myocardial perfusion grade
(TMPG) could be ascertained and in whom the culprit
artery was angiographically patent as defined by TIMI
flow grade 2 or 3 at 60 minutes after administration of
fibrinolytic therapy (929 patients of 1,268 patients
who had open arteries). Quantitative coronary angiography and assessments of corrected TIMI frame
count9 and TMPG10 were performed as previously
described.
All analyses were performed with STATA 7.0.11
All continuous variables are reported as medians with
25% and 75% interquartile values unless otherwise
specified. Nonparametric Wilcoxon’s rank sum test
and Student’s t test were used as appropriate. Chisquare test was used for analysis of categorical variables. Multivariate linear and logistic regression models were constructed using impaired epicardial flow
and myocardial perfusion by corrected TIMI frame
count and TMPG, respectively, as outcome variables,
and incorporating AET, potential confounders
(p ⬍0.10 in univariate analyses), other known covariates of corrected TIMI frame count, and TMPG in the
dataset, and treatment with a glycoprotein IIb/IIIa
inhibitor.
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.007
lesion (Table 1). The minimal luminal diameter of the culprit lesion did
Thrombus
not differ significantly between paNo. of
tients who had AET and those who
Characteristic
Patients
No
Yes
p Value
did not. Among patients who had
Left anterior descending
AET, there was a larger percentage
culprit vessel
of TIMI grade 2 flow (33.6% vs
⫹
325
39.3%
27.9%
⬍0.001
26.8%, p ⫽ 0.03; Table 2) and cor⫺
604
60.7%
72.1%
No. of narrowed
rected TIMI frame counts were
coronary arteries
higher (slower). When analysis was
1
486
51.2%
56.2%
0.34
restricted to the 656 patients who had
2
288
32.5%
29.8%
TIMI grade 3 flow, corrected TIMI
3
141
16.3%
14.0%
Lesion length (mm)
891
9 (7–13)
11 (7–15)
0.004
frame counts remained higher among
Reference segment
894
2.7 (2.3–3.2)
2.9 (2.5–3.4)
0.0002
patients who had AET (median
(mm)
[25%, 75%] of 29 frames [22, 34] vs
Minimum luminal
894
1.0 (0.8–1.3)
1.0 (0.7–1.3)
0.67
26 frames, p ⫽ 0.02).
diameter (mm)
The presence of AET in the inStenosis (%)
894
62.0 (52.5–71.6) 64.8 (54.2–73.1) 0.009
farct-related
artery after fibrinolytic
Values are percentages or medians (25% to 75% interquartile ranges).
therapy was associated with impaired myocardial perfusion as assessed by TMPG (43.0% vs 32.0%
TMPG 0/1, p ⫽ 0.001; Figure 1).
TABLE 2 Angiographic Characteristics of Flow and Perfusion
Similarly, any abnormality in TMPG
Thrombus
(grades 0/1/2) was observed more
frequently among patients who had
No. of
No
Yes
AET (46.7% vs 34.6%, p ⬍0.001).
Patients
(n ⫽ 566)
(n ⫽ 341)
Chi-square
p Value
When analysis was further restricted
TIMI frame count at
31 (22–41)
34 (25–50)
0.0003
to patients who had TIMI grade 3
60 min*
flow, the association between AET
TIMI flow grade at
60 min
and impaired myocardial perfusion
2
273
26.8%
33.6%
4.9
0.03
remained, with TMPG 0/1 observed
3
656
73.1%
63.4%
in 33.5% of patients who had AET
TMPG at 60 min
versus 26.0% of patients who did not
0
208
21.5%
23.9%
18.3
⬍0.001
1
128
10.5%
19.1%
have AET (p ⫽ 0.043; Figure 2).
2
28
2.6%
3.7%
AET remained independently as3
565
65.4%
53.3%
sociated
with a higher corrected
0/1
336
32.0%
43.0%
11.5
0.001
TIMI frame count (p ⫽ 0.01) after
2/3
593
68.0%
57.0%
adjustment for TIMI risk index, time
0/1/2
364
34.6%
46.7%
13.5
⬍0.001
3
565
65.4%
53.3%
to treatment, gender, history of hypercholesterolemia, culprit location
*Values are medians (25% to 75% interquartile ranges).
in the left anterior descending coronary artery, reference segment diameter, percent stenosis of the culprit
AET 60 minutes after fibrinolytic administration lesion, lesion length, treatment with a glycoprotein
was detected in 37.8% of patients (351 of 929). A IIb/IIIa inhibitor, and TIMI grade 3 flow. In a similar
history of hypercholesterolemia was more frequent model restricted to patients who had TIMI grade 3
among patients who had AET (27.6% vs 21.5%, p ⫽ flow, AET was independently associated with a higher
0.03). There was a slightly larger percentage of men corrected TIMI frame count (p ⫽ 0.03).
AET was also independently associated with imwho had AET among the patients (82.9% vs 77.8%, p
⫽ 0.06). There were no significant associations be- paired TMPG after adjustment for clinical and angiotween AET and other baseline characteristics, includ- graphic covariates, including TIMI risk index, treatment
ing age, active smoking, previous aspirin use, history with a glycoprotein IIb/IIIa inhibitor, and TIMI grade 3
of hypertension, history of diabetes, family history of flow (odds ratio 1.5 for TMPG 0/1, 95% confidence
coronary artery disease, previous coronary artery dis- interval 1.1 to 2.0, p ⫽ 0.004; odds ratio 1.6 for TMPG
ease (angina, revascularization, or myocardial infarc- 0/1/2, 95% confidence interval 1.2 to 2.1, p ⫽ 0.002).
tion), time to treatment, white blood cell counts (n ⫽ When reference segment diameter, percent stenosis of
584), C-reactive protein levels (n ⫽ 415), or TIMI risk the culprit lesion, lesion length, and corrected TIMI
index [heart rate ⫻ (age/10)2/systolic blood pressure]. frame count were added to the model, AET remained
Angiographic characteristics associated with AET independently associated with impaired TMPG (odds
were noninvolvement of the culprit left anterior de- ratio 1.4 for TMPG 0/1, 95% confidence interval 1.0 to
scending artery, longer lesions, reference segment di- 1.9, p ⫽ 0.026; odds ratio 1.5 for TMPG 0/1/2, 95%
ameters, and percent diameter stenosis of the culprit confidence interval 1.1 to 2.0, p ⫽ 0.010).
TABLE 1 Angiographic Lesion Characteristics
BRIEF REPORTS
225
In patients who have ST-elevation
myocardial infarction treated with fibrinolytic therapy or percutaneous coronary intervention, thrombus formation can result in distal embolization of
microthrombi and platelet/fibrin aggregates into the microcirculation.13 The
role of thrombus in distal embolization
during percutaneous coronary intervention has been demonstrated by
Fukuda et al,14 who described an association between the presence of thrombus as assessed by intravascular ultrasound and angiographically observed
distal embolization during percutaneous coronary intervention. However,
even before percutaneous coronary inFIGURE 1. Association of AET with TMPG among all patent arteries. Percentages of patervention or fibrinolytic therapy, mitients who had (A) TMPG 0/1 or (B) TMPG 0/1/2.
croembolization may occur among patients who have acute coronary
syndromes.15 Thus, the finding of an
association between residual AET and
impaired TMPG suggests that residual
AET after fibrinolytic therapy may be
a surrogate of a larger preexisting
thrombotic burden or a more organized, thromboresistant clot that may
result in worsening microvascular dysfunction.
Although distal embolization may
be 1 mechanistic explanation for our
observations, residual intracoronary
thrombus after fibrinolytic therapy
may be a marker of a more intense
prothrombotic and inflammatory
process with resultant microcirculatory dysfunction. Fibrinolytic therapy results in thrombin generation
and platelet activation in patients
FIGURE 2. Association of AET with TMPG among patients who had TIMI flow grade 3.
who have ST-elevation myocardial
Percentages of patients who had (A) TMPG 0/1 or (B) TMPG 0/1/2.
infarction.16,17 These 2 processes can
contribute to thrombus propagation
•••
and to the release of procoagulant and vasoconstrictor
AET after fibrinolytic therapy for ST-elevation substances that may further perpetuate endothelial
myocardial infarction has been associated with lower dysfunction and microvascular dysfunction.4 In anirates of epicardial patency and slower epicardial mal models, arterial thrombus has been associated
flow.12 This study extends previous observations and with impaired release of endothelium-dependent redemonstrates that, even among patients who have had laxing factor and with higher indexes of arterial resissuccessful restoration of epicardial patency after fi- tance.18 Although we did not observe any associations
brinolytic therapy, residual AET is associated with between AET and white blood cell counts or C-reacimpaired myocardial perfusion in addition to slower tive protein levels, this analysis was restricted to paepicardial flow. These associations were independent tients who had patent epicardial arteries ⱕ60 minutes
of severity of residual stenosis and other characteris- of fibrinolytic administration, a subgroup previously
tics of the culprit lesion. Moreover, after adjustment demonstrated to have a less intense systemic inflamfor TIMI grade 3 flow and corrected TIMI frame count matory response than patients who had occluded arin addition to baseline and lesion characteristics, AET teries.19
remained independently associated with impaired
The combined effects of microembolization, inmyocardial perfusion. These findings support the hy- flammation, and vasoconstriction may contribute to
pothesis that AET is associated with increased micro- the observed associations across thrombus, impaired
vascular obstruction/vasoconstriction, perhaps related corrected TIMI frame count, and TMPG. An increase
to greater distal embolization, even among patients in “downstream” microvascular resistance may also
who have achieved successful epicardial reperfusion. explain the independent association between AET and
226 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
higher (slower) corrected TIMI frame count, even
when adjusted for lesion characteristics that affect
epicardial flow. Insofar as AET is associated with
impaired myocardial perfusion, strategies aimed at
decreasing the risks of distal embolization, vasoconstriction, and impaired myocardial perfusion warrant
further evaluation. Although the negative preliminary
results of the Enhanced Myocardial Efficacy and Recovery by Aspiration of Liberalized Debris [EMERALD] trial of a distal protection device in ST-elevation myocardial infarction may temper enthusiasm,20
distal protection devices do not offer protection
against embolization into proximal arterial branches
and do not decrease vasoconstrictor release. Further,
embolization can occur with deployment of these devices, particularly if passing the device beyond the
culprit lesion is difficult.
1. Okamatsu K, Takano M, Sakai S, Ishibashi F, Uemura R, Takano T, Mizuno
K. Elevated troponin T levels and lesion characteristics in non-ST-elevation acute
coronary syndromes. Circulation 2004;109:465– 470.
2. Wong GC, Morrow DA, Murphy S, Kraimer N, Pai R, James D, Robertson
DH, Demopoulos LA, DiBattiste P, Cannon CP, Gibson CM. Elevations in
troponin T and I are associated with abnormal tissue level perfusion: a TACTICSTIMI 18 substudy. Treat Angina with Aggrastat and Determine Cost of Therapy
with an Invasive or Conservative Strategy-Thrombolysis in Myocardial Infarction. Circulation 2002;106:202–207.
3. Ross MJ, Herrmann HC, Moliterno DJ, Blankenship JC, Demopoulos L,
DiBattiste PM, Ellis SG, Ghazzal Z, Martin JL, White J, Topol EJ. Angiographic
variables predict increased risk for adverse ischemic events after coronary stenting with glycoprotein IIb/IIIa inhibition: results from the TARGET trial. J Am
Coll Cardiol 2003;42:981–988.
4. Willerson JT, Golino P, Eidt J, Campbell WB, Buja LM. Specific platelet
mediators and unstable coronary artery lesions. Experimental evidence and potential clinical implications. Circulation 1989;80:198 –205.
5. Antman EM, Giugliano RP, Gibson CM, McCabe CH, Coussement P, Kleiman
NS, Vahanian A, Adgey AA, Menown I, Rupprecht HJ, et al. Abciximab
facilitates the rate and extent of thrombolysis: results of the Thrombolysis In
Myocardial Infarction (TIMI) 14 trial. The TIMI 14 Investigators. Circulation
1999;99:2720 –2732.
6. Giugliano RP, Roe MT, Harrington RA, Gibson CM, Zeymer U, Van de Werf
F, Baran KW, Hobbach HP, Woodlief LH, Hannan KL, et al. Combination
reperfusion therapy with eptifibatide and reduced-dose tenecteplase for STelevation myocardial infarction: results of the Integrilin and Tenecteplase in
Acute Myocardial Infarction (INTEGRITI) Phase II Angiographic Trial. J Am
Coll Cardiol 2003;41:1251–1260.
7. Antman EM, Louwerenburg HW, Baars HF, Wesdorp JC, Hamer B, Bassand
JP, Bigonzi F, Pisapia G, Gibson CM, Heidbuchel H, et al. Enoxaparin as
adjunctive antithrombin therapy for ST-elevation myocardial infarction: results of
the ENTIRE-Thrombolysis in Myocardial Infarction (TIMI) 23 Trial. Circulation
2002;105:1642–1649.
8. Ohman EM, Van de Werf F, Antman EM, Califf RM, Lemos JAD, Gibson
CM, Oliverio RL, Harrelson L, McCabe C, DiBattiste P, et al. Tenecteplase and
tirofiban in ST-segment elevation acute myocardial infarction: results of a randomized trial. Am Heart J 2005; in press.
9. Gibson CM, Cannon CP, Daley WL, Dodge JT Jr, Alexander B Jr, Marble SJ,
McCabe CH, Raymond L, Fortin T, Poole WK, Braunwald E. TIMI frame count:
a quantitative method of assessing coronary artery flow. Circulation 1996;93:
879 – 888.
10. Gibson CM, Cannon CP, Murphy SA, Ryan KA, Mesley R, Marble SJ,
McCabe CH, Van De Werf F, Braunwald E. Relationship of TIMI myocardial
perfusion grade to mortality after administration of thrombolytic drugs. Circulation 2000;101:125–130.
11. STATA Statistical Software. College Station, TX: STATA Corp., 2000.
12. Gibson CM, Murphy S, Menown IB, Sequeira RF, Greene R, Van de Werf
F, Schweiger MJ, Ghali M, Frey MJ, Ryan KA, et al. Determinants of coronary
blood flow after thrombolytic administration. TIMI Study Group. Thrombolysis
in Myocardial Infarction. J Am Coll Cardiol 1999;34:1403–1412.
13. Saber RS, Edwards WD, Bailey KR, McGovern TW, Schwartz RS, Holmes
DR Jr. Coronary embolization after balloon angioplasty or thrombolytic therapy:
an autopsy study of 32 cases. J Am Coll Cardiol 1993;22:1283–1288.
14. Fukuda D, Tanaka A, Shimada K, Nishida Y, Kawarabayashi T, Yoshikawa
J. Predicting angiographic distal embolization following percutaneous coronary
intervention in patients with acute myocardial infarction. Am J Cardiol 2003;91:
403– 407.
15. Heusch G, Schulz R, Baumgart D, Haude M, Erbel R. Coronary microembolization. Prog Cardiovasc Dis 2001;44:217–230.
16. Merlini PA, Bauer KA, Oltrona L, Ardissino D, Spinola A, Cattaneo M,
Broccolino M, Mannucci PM, Rosenberg RD. Thrombin generation and activity
during thrombolysis and concomitant heparin therapy in patients with acute
myocardial infarction. J Am Coll Cardiol 1995;25:203–209.
17. Gurbel PA, Serebruany VL, Shustov AR, Bahr RD, Carpo C, Ohman EM,
Topol EJ. Effects of reteplase and alteplase on platelet aggregation and major
receptor expression during the first 24 hours of acute myocardial infarction
treatment. GUSTO-III Investigators. Global Use of Strategies to Open Occluded
Coronary Arteries. J Am Coll Cardiol 1998;31:1466 –1473.
18. Reil TD, Moore WS, Kashyap VS, Nene SS, Gelabert HA, QuinonesBaldrich WJ. The effects of thrombus, thrombectomy and thrombolysis on
endothelial function. Eur J Vasc Endovasc Surg 2000;19:162–168.
19. Barron HV, Cannon CP, Murphy SA, Braunwald E, Gibson CM. Association
between white blood cell count, epicardial blood flow, myocardial perfusion, and
clinical outcomes in the setting of acute myocardial infarction: a Thrombolysis In
Myocardial Infarction 10 substudy. Circulation 2000;102:2329 –2334.
20. Stone GW, Webb J, Cox DA, Brodie BR, Quershi M, Dulas D, Kalynch A,
Turco M, Schultheiss HP, Rutherford B, et al. Primary angioplasty in acute
myocardial infarction with distal protection of the microcirculation: principal
results from the prospective, randomized EMERALD trial. J Am Coll Cardiol
2004;43(5 suppl A):285A.
BRIEF REPORTS
227
Angiography and Revascularization in Patients With
Heart Failure Following Fibrinolytic Therapy for
ST-Elevation Acute Myocardial Infarction
Amir Kashani, MS, MD, C. Michael Gibson, MS, MD, Sabina A. Murphy, MPH,
Marc S. Sabatine, MD, MPH, David A. Morrow, MD, MPH, Elliott M. Antman, MD, and
Robert P. Giugliano, MD, SM
We evaluated the use of coronary angiography and
clinical outcomes among patients who had heart failure and were enrolled in the Intravenous Novel Plasminogen Activator (NPA) for the Treatment of Infarcting Myocardium Early study, a large international
trial of fibrinolytic therapy in ST-elevation myocardial
infarction. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:228 –233)
he Second Intravenous Novel Plasminogen Activator (NPA) for the Treatment of Infarcting MyoT
cardium Early trial included patients ⱖ18 years of
1
age who presented within 6 hours of symptom onset
with ST elevation in 2 contiguous leads or new left
bundle branch block. Patients in cardiac shock (Killip’s class IV) were excluded. Other major exclusions
were increased bleeding risk, previous stroke, systolic
blood pressure ⱖ180 mm Hg, or diastolic blood pressure ⱖ110 mm Hg. A total of 15,078 patients was
enrolled from 1997 to 1998 in 855 hospitals in 35
countries.
•••
Patients from the trial were categorized into 1 of
4 mutually exclusive groups based on the development of heart failure (HF) and use of angiography
during the index hospitalization: group 1, no HF
and angiogram; group 2, no HF and no angiogram;
group 3, HF and angiogram; and group 4, HF and no
angiogram. Twelve patients did not have angiographic data and were excluded from analyses in
these groups. A secondary analysis explored the
severity of HF categorized in 4 groups2: (1) shock,
defined as Killip’s class IV, systolic blood pressure
⬍90 mm Hg despite adequate ventricular filling or
systolic blood pressure ⱖ90 mm Hg that required
inotropic agents or intra-aortic balloon pump after
volume expansion; (2) severe HF, defined as HF
(not satisfying shock criteria) that required new use
of, or increase in the dose of, a diuretic and ⱖ1
objective signs of HF (i.e., radiographic evidence of
pulmonary edema, rales on examination that did not
From the Rochester General Hospital, Rochester, New York; and the
TIMI Study Group, Brigham & Women’s Hospital, Boston, Massachusetts. Dr. Giugliano’s address is: TIMI Study Group, 350 Longwood
Avenue, 1st Floor Offices, Boston, Massachusetts 02115. E-mail:
rgiugliano@partners.org. Manuscript received August 5, 2004; revised manuscript received and accepted August 26, 2004.
228
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
resolve with coughing, or mean pulmonary capillary wedge pressure ⱖ18 mm Hg and cardiac index
ⱕ2.2 L/min/m2); (3) mild HF, defined as not meeting HF criteria 1 or 2 and objective signs of HF at
presentation that resolved with therapy before discharge, required diuretics after randomization, or
had objective findings of HF (but not the 2 together); and (4) no HF, defined as no evidence of
HF on presentation or during index hospitalization.
Data were obtained from trial case record forms.
The log-rank test was used to compare cumulative
mortality rates through 1 year. Multivariate logistic
models were constructed using backward selection,
with statistical significance set at a p value ⬍0.05.
The z score was used to identify variables with the
greatest predictive power. A multivariate model to
evaluate the association between use of angiography and 1-year mortality rate was developed. The
Thrombolysis In Myocardial Infarction risk score
for ST-elevation myocardial infarction (STEMI), a
simple bedside tool that produces an integer score
that provides risk stratification with respect to mortality rate in patients who have STEMI,3 was used
to adjust for differences in baseline risk. A propensity score to predict the use of angiography was
developed using baseline patient and hospital characteristics associated with performance of an angiogram before discharge. Thrombolysis In Myocardial
Infarction risk score, angiographic propensity score,
and baseline hospital characteristics were used in a
stepwise fashion to calculate the adjusted hazard
decrease in 1-year mortality rate associated with the
use of angiography in the population stratified by
the presence/absence of HF during the index admission.
Of 15,078 patients enrolled, 3,420 (23%) developed HF during the index admission after fibrinolysis, and 5,667 (38%) underwent angiography before discharge. Among patients who had HF during
the index admission, 75% developed HF on hospital
days 0 to 1, and the remaining 25% developed HF
between day 2 and discharge (median 9 days). Baseline characteristics for the 4 groups of patients who
were stratified by the presence of HF and use of
angiography during the index admission are presented in Tables 1 (patient characteristics) and 2
(hospital characteristics). Patients who underwent
angiography before discharge were more often
younger, were men, had previous percutaneous cor0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.092
TABLE 1 Baseline Characteristics
HF During Index Admission
(n ⫽ 3,420, 23%)
No HF During Index Admission
(n ⫽ 11,658, 77%)
Angiogram
Variable
Age ⱖ65 yrs
Women
Weight ⬍67 kg
Systemic hypertension
Diabetes mellitus
Current smoker
Previous PCI
Previous angina pectoris
Previous myocardial infarction
Previous heart failure
Previous renal insufficiency
Previous medications
␤ Blockers
Diuretics
Angiotensin-converting enzyme inhibitors/
angiotensin receptor blockers
Aspirin
Findings at presentation
Creatinine ⬎1.5 mg/dl
Atrial fibrillation
Anterior myocardial infarction
Peak creatine kinase ⫾ SD
Heart rate ⬎100 beats/min
Systolic blood pressure ⬍100 mm Hg
Angiogram
Present
(n ⫽ 1,010)
Absent
(n ⫽ 2,410)
All
(n ⫽ 3,420)
Present
(n ⫽ 4,657)
Absent
(n ⫽ 6,989)
All*
(n ⫽ 11,646)
50%
27%
19%
36%
20%
43%
8%
24%
21%
5%
2%
64%
33%
25%
37%
17%
37%
3%
32%
24%
11%
2%
60%
31%
23%
37%
18%
38%
4%
30%
23%
9%
2%
33%
20%
17%
30%
13%
46%
7%
17%
15%
1%
0.9%
39%
25%
21%
28%
13%
48%
3%
20%
14%
2%
0.8%
37%
23%
19%
29%
13%
47%
5%
19%
14%
1%
0.8%
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.001
0.67
⬍0.001
⬍0.001
⬍0.001
⬍0.001
19%
14%
16%
19%
20%
18%
19%
18%
17%
17%
8%
13%
13%
9%
12%
15%
8%
12%
⬍0.001
⬍0.001
⬍0.001
24%
26%
25%
21%
17%
18%
⬍0.001
8%
6%
53%
2,247 ⫾ 2,152
13%
4%
13%
8%
53%
2,383 ⫾ 2,131
14%
4%
11%
7%
53%
2,343 ⫾ 2,138
13%
4%
5%
2%
39%
1,485 ⫾ 1,563
6%
2%
6%
3%
39%
1,755 ⫾ 1,651
6%
2%
5%
3%
39%
1,646 ⫾ 1,621
6%
2%
BRIEF REPORTS
Data are percentage of patients, unless otherwise indicated.
*Twelve patients did not have data regarding the use or nonuse of angiography and were excluded from analyses.
†
HF present versus absent. All 4-way (yes HF/yes angiogram, yes HF/no angiogram, no HF/yes angiogram, no HF/no angiogram) p values are ⬍0.001 except peak creatine kinase ⫾ SD (p ⫽ 0.001).
PCI ⫽ percutaneous coronary intervention.
p Value†
⬍0.0001
⬍0.001
⬍0.001
⬍0.0001
⬍0.001
⬍0.001
229
TABLE 2 Hospital Characteristics
Variable
Geographic region
North America
Latin America
Eastern Europe
Western Europe
Hospital size (no. of beds)
⬍300 beds
300–700 beds
⬎700 beds
City (⬎100,000 population)
hospital
On-site angiography
available
24-h PCI available
On-site coronary bypass
available
Teaching hospital
Lytic by cardiologist‡
HF During Index Admission
(n ⫽ 3,420, 23%)
No HF During Index Admission
(n ⫽ 11,658, 77%)
Angiogram
Angiogram
Present
(n ⫽ 1,010)
Absent
(n ⫽ 2,410)
All HF
(n ⫽ 3,420)
Present
(n ⫽ 4,657)
Absent
(n ⫽ 6,989)
All No HF*
(n ⫽ 11,646)
31%
2%
8%
59%
12%
2%
27%
60%
18%
2%
21%
59%
27%
3%
7%
62%
14%
3%
26%
60%
20%
3%
19%
59%
25%
47%
29%
72%
20%
47%
33%
84%
22%
47%
31%
81%
25%
47%
28%
71%
23%
49%
29%
79%
24%
48%
28%
76%
⬍0.001
65%
46%
52%
69%
47%
56%
⬍0.0001
31%
32%
15%
21%
20%
24%
37%
34%
16%
23%
24%
27%
⬍0.001
0.0001
68%
60%
65%
56%
66%
57%
68%
65%
64%
60%
66%
62%
0.84
⬍0.001
p Value†
⬍0.001
0.001
*Twelve patients did not have data regarding the use or nonuse of angiography and were excluded from analyses.
†
HF present versus absent. All 4-way (yes HF/yes angiogram, yes HF/no angiogram, no HF/yes angiogram, no HF/no angiogram) p values ⬍0.001 except
teaching hospital (p ⫽ 0.002).
‡
Hospitals in which the cardiologist is the physician who most commonly administers fibrinolytic therapy.
Abbreviations as in Table 1.
TABLE 3 Multivariate Analyses Predicting Patients Who
Underwent Angiography
Variable
On-site PCI availability
24 h (vs none)
Day only (vs none)
Geographic region
Eastern Europe (vs Western Europe)
North America (vs Western Europe)
Latin America (vs Western Europe)
Rural/local hospital (⬍100,000
population)
Age ⬍65 yrs
Hospitals with lytic administered by
cardiologist*
Previous PCI
No previous HF
Hospital size
Medium 300–700 beds (vs large
700 beds)
Small 300 beds (vs large 700
beds)
Previous hypertension
Men
Creatinine ⬍1.5 mg/dl
OR
95% CI
z Score
5.0
3.2
4.5–5.6
2.8–3.5
28.6
19.3
0.2 0.18–0.23
2.2
1.9–2.4
0.6
0.5–0.8
2.1
1.9–2.4
23.7
13.9
4.0
14.4
1.6
1.6
1.4–1.7
1.4–1.7
10.2
9.9
2.1
2.4
1.7–2.5
1.8–3.2
7.2
5.9
1.3
1.1–1.4
4.5
1.2
1.0–1.4
2.4
1.2
1.2
1.3
1.1–1.3
1.1–1.3
1.1–1.6
4.0
3.8
3.3
Only variables with p values ⱕ0.001 are shown. The model was also adjusted
for the following variables: teaching hospital (p ⫽ 0.002), previous aspirin use (p
⫽ 0.007), anterior myocardial infarction (p ⫽ 0.02), current smoking (p ⫽ 0.02),
atrial fibrillation (p ⫽ 0.03), previous angina (p ⫽ 0.045), weight ⬎67 kg (p ⫽
0.08), and previous myocardial infarction (p ⫽ 0.08).
*Hospitals in which the cardiologist is the physician who most commonly
administers fibrinolytic therapy.
CI ⫽ confidence interval; OR ⫽ odds ratio. Other abbreviation as in Table 1.
onary intervention, did not have HF, and presented
at North American hospitals with on-site interventional facilities (Table 3).
230 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
FIGURE 1. Rates of angiography and revascularization before
discharge in patients who had HF and those who did not. The
third pair of bars represents revascularization rates among patients who had an angiogram.
A total of 3,538 patients (23%) was revascularized;
among these patients, 2,969 (20%) previously had
percutaneous coronary intervention only, 610 (4%)
previously had coronary artery bypass grafting only,
and 53 (0.4%) previously had both procedures. Patients who had HF were less likely to have undergone
angiography (30% vs 40%) and revascularization
(19% vs 25%) than patients who did not have HF (p
⬍0.001 for the 2 groups). However, once an angiogram had been performed, rates of revascularization
were similar (63% vs 62%, p ⫽ 0.93; Figure 1). Rates
JANUARY 15, 2005
FIGURE 2. Geographic variation in rates of angiography and revascularization.
of invasive procedures varied widely and depended on
hospital characteristics (Table 2) and geographic location (Figure 2). Procedures were performed fourfold
to 10-fold more frequently in the United States (angiography 79%, revascularization 59%) than in Canada and Europe. There was a greater difference in
rates of angiography and revascularization among patients who had HF than among those who did not who
had been enrolled in Europe compared with patients in
North America.
Mortality was fourfold to sevenfold higher at 1
year in patients who had HF than in those who did not
(Figure 3). Across the HF spectrum, mortality rates at
1 year were lower among patients who underwent
angiography before discharge: shock (41% vs 75%),
severe HF (11% vs 21%), mild HF (6% vs 16%), and
no HF (2% vs 7%).
Use of angiography before discharge was associated with a lower mortality at 1 year in an unadjusted analysis among patients who did not have HF
(hazard ratio 0.33, 95% confidence interval 0.27 to
0.40) and among those who had HF during the
index admission (hazard ratio 0.50, 95% confidence
interval 0.42 to 0.60). After adjustment for baseline
risk with the Thrombolysis In Myocardial Infarction risk score for STEMI, a propensity score for
undergoing angiography before discharge, and dif-
ferences in hospital characteristics, use of angiography before discharge continued to demonstrate a
strong and independent association with lower
1-year mortality, regardless of the presence of HF
during the index admission (Table 4).
Causes of death at 1 year in patients who had been
stratified by use of angiography are listed in Table 5.
Among patients who died, recurrent myocardial infarction and cardiac rupture were more frequent
causes of death in patients who did not undergo angiography during the index admission.
Twenty-three percent of patients developed HF
during the index hospitalization in this recent, large,
international trial of fibrinolytic therapy for STEMI.
Despite a fourfold higher mortality rate, patients
who had HF were less likely to undergo angiography and revascularization than were patients who
did not have HF. This observation may indicate
physicians’ reluctance to use early and aggressive
management secondary to the presence of greater
co-morbidity in this patient population. However,
our data demonstrate that patients who had HF and
those who did not develop HF had higher survival
rates if angiography had been performed before
discharge, even after adjustment for baseline patient
risk, a propensity score adjusting for the likelihood
of angiography, and differences in hospital characBRIEF REPORTS
231
had been admitted to North American hospitals with
on-site interventional facilities were most likely to
undergo angiography before discharge. Although
patients who have HF tend to have more co-morbidities, they also appear to benefit from an early
invasive approach after fibrinolysis. We previously
published a detailed comparison of patients who
underwent angiography with patients who did not.4
•••
Our findings support recommendations based on
previous studies5–7 and the 2004 STEMI guidelines8
that advocate an early invasive approach in HF after
STEMI. We also associated an early invasive approach with longer survival rates across the HF
spectrum. This observation is consistent with a similar analysis of primary percutaneous intervention
for different degrees of HF.9 A German study associated routine early stenting in all patients after
fibrinolysis (regardless of HF) with a 50% decreased rate of ischemic complications (51% vs
26%, p ⫽ 0.001).10
Our analysis of geographic practice variation has
confirmed those of previous studies11,12 that demonstrated an apparent underuse of invasive procedures after fibrinolysis, particularly outside the
United States. Even in American hospitals with
on-site facilities, there appears to be a paradoxic
tendency for higher rates of angiography in patients
FIGURE 3. Cumulative mortality rates over 1 year in patients
who have lower risk.4,13–15 We observed a similar
who had HF versus those who did not (A) and among those who
practice pattern among patients who had been strathad different degrees of HF during the index admission (B). Anified by the presence of HF in our analyses, espegio ⴝ angiography.
cially in Europe, where patients who had HF were
25% to 40% less likely to undergo
an invasive procedure than were
patients who did not have HF.
TABLE 4 Use of Inpatient Angiography to Predict 30-day Mortality Rate
Our data suggest that such a conHazard Ratio
servative approach is associated with
for In-patient
an increased incidence of deaths due
Models Predicting Death Over 30 Days
Angiography
95% CI
p Value
to reinfarction and cardiac rupture.
Unadjusted (univariate) model
Percutaneous coronary intervention
HF absent
0.33
0.27–0.40
⬍0.0001
after fibrinolysis has been associated
HF present
0.50
0.42–0.60
⬍0.0001
with a decrease in recurrent myocarAdjusted for TIMI risk score
dial infarction,16 possibly due to the
HF absent
0.36
0.29–0.44
⬍0.0001
HF present
0.57
0.48–0.69
⬍0.0001
additional mechanical stability that
Adjusted for TIMI risk score and propensity
the intervention provides to the culscore for angiography before discharge
prit lesion. The observed decrease in
HF absent
0.36
0.28–0.46
⬍0.0001
fatal cardiac rupture in patients who
HF present
0.59
0.48–0.73
⬍0.0001
underwent revascularization is conAdjusted for TIMI risk score, propensity
score, and hospital characteristics
sistent with the notion that percutaHF absent
0.36
0.28–0.46
⬍0.0001
neous coronary intervention restores
HF present
0.57
0.46–0.70
⬍0.0001
myocardial perfusion and improves
Unadjusted and adjusted mortality at 1 year among patients who underwent angiography before
tissue turgor, thus increasing the tendischarge. Results are shown stratified by the presence/absence of HF during the index admission.
sile strength of the myocardium. This
Patients who did not undergo angiography before discharge represent the referent group (OR 1.0). See
mechanism has been postulated as an
text for details.
explanation for the very low rate of
TIMI ⫽ Thrombolysis In Myocardial Infarction; other abbreviations as in Table 3.
rupture observed with primary percutaneous coronary intervention.17–19
Improved myocardial perfusion into
the healing infarct zone may also deteristics. In this worldwide clinical trial, young men crease the adverse effects of remodeling and has
who previously had successful percutaneous coro- been postulated as a potential late benefit of an open
nary intervention, had fewer co-morbidities, and artery.20
232 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
executive summary: a report of the ACC/AHA Task
Force on Practice Guidelines (Committee to Revise the
1999 Guidelines on the Management of Patients With
Acute Myocardial Infarction). Circulation 2004;110;
Angiogram No Angiogram
Excess Deaths/
e82– e292.
9. Zahn R, Schiele R, Schneider S, Gitt AK, WienberCause of Death
(282 deaths) (1,238 deaths)
p Value
1,000 Patients
gen H, Seidl K, Voigtlander T, Gottwik M, Berg G,
HF (n ⫽ 385)
24.8%
25.4%
0.88
⫹21
Altmann E, et al. Primary angioplasty versus intravenous thrombolysis in acute myocardial infarction: can
Reinfarction (n ⫽ 305)
15.6%
21.1%
0.04
⫹20
we define subgroups of patients benefiting most from
Rupture (n ⫽ 173)
6.4%
12.5%
0.002
⫹13
primary angioplasty? Results from the pooled data of
Intracranial hemorrhage
2.8%
7.0%
0.006
⫹8
the Maximal Individual Therapy in Acute Myocardial
(n ⫽ 95)
Infarction Registry and the Myocardial Infarction RegOther cardiovascular (n ⫽ 133)
9.2%
8.6%
0.73
⫹7
istry. J Am Coll Cardiol 2001;37:1827–1835.
Dysrhythmia (n ⫽ 105)
8.2%
6.6%
0.36
⫹5
10. Scheller B, Hennen B, Hammer B, Walle J, Hofer
Other noncardiovascular
17.0%
7.9%
⬍0.0001
⫹2
C, Hilpert V, Winter H, Nickenig G, Bohm M. Bene(n ⫽ 146)
ficial effects of immediate stenting after thrombolysis in
acute myocardial infarction. J Am Coll Cardiol 2003;
Unobserved (n ⫽ 75)
6.4%
4.6%
0.22
⫹3
42:634 – 641.
Unknown (n ⫽ 103)
9.6%
6.1%
0.048
⫹3
11. Gupta M, Chang WC, Van de Werf F, Granger CB,
Midodzi W, Barbash G, Pehrson K, Oto A, Toutouzas
Numbers in the second and third columns represent the percentage of deaths due to the indicated
P, Jansky P, Armstrong PW. International differences in
causes. The last column represents the excess number of deaths among 1,000 patients who did not
in-hospital revascularization and outcomes following
undergo angiography versus those who did undergo angiography before discharge for each cause of
acute myocardial infarction: a multilevel analysis of
death.
patients in ASSENT-2. Eur Heart J 2003;24:1640 –
1650.
12. Fox KA, Goodman SG, Anderson FA Jr, Granger
CB, Moscucci M, Flather MD, Spencer F, Budaj A,
Dabbous OH, Gore JM. From guidelines to clinical
1. The In TIME-II Investigators. Intravenous NPA for the Treatment of Infarcting
practice: the impact of hospital and geographical characteristics on temporal
Myocardium Early; InTIME-II, a double-blind comparison of single-bolus
trends in the management of acute coronary syndromes. The Global Registry of
lanoteplase vs accelerated alteplase for the treatment of patients with acute
Acute Coronary Events (GRACE). Eur Heart J 2003;24:1414 –1424.
13. Stenestrand U, Wallentin L. Early revascularisation and 1-year survival in
myocardial infarction. Eur Heart J 2000;21:2005–2013.
2. Kashani A, Giugliano RP, Altman E, Morrow DA, Gibson CM, Murphy S,
14-day survivors of acute myocardial infarction: a prospective cohort study.
Braunwald E. Severity of heart failure, treatments, and outcomes after fibrinoLancet 2002;359:1805–1811.
14. Magid DJ, Calonge BN, Rumsfeld JS, Canto JG, Frederick PD, Every NR,
lysis. Eur Heart J 2004;25:1702–1710.
3. Morrow DA, Antman EM, Charlesworth A, Cairns R, Murphy SA, de Lemos
Barron HV. Relation between hospital primary angioplasty volume and mortality
JA, Giugliano RP, McCabe CH, Braunwald E. TIMI risk score for ST-elevation
for patients with acute MI treated with primary angioplasty vs thrombolytic
myocardial infarction: a convenient, bedside, clinical score for risk assessment at
therapy. JAMA 2000;284:3131–3138.
15. Pilote L, Miller DP, Califf RM, Rao JS, Weaver WD, Topol EJ. Determinants
presentation: an intravenous nPA for treatment of infarcting myocardium early II
of the use of coronary angiography and revascularization after thrombolysis for
trial substudy. Circulation 2000;102:2031–2037.
4. Llevadot J, Giugliano RP, Antman EM, Wilcox RG, Gurfinkel EP, Henry T,
acute myocardial infarction. N Engl J Med 1996;335:1198 –1205.
16. Gibson CM, Karha J, Murphy SA, James D, Morrow DA, Cannon CP,
McCabe CH, Charlesworth A, Thompson S, Nicolau JC, et al. Availability of
Giugliano RP, Antman EM, Braunwald E. Early and long-term clinical outcomes
on-site catheterization and clinical outcomes in patients receiving fibrinolysis for
associated with reinfarction following fibrinolytic administration in the ThromST-elevation myocardial infarction. Eur Heart J 2001;22:2104 –2115.
5. Hochman JS, Sleeper LA, Webb JG, Sanborn TA, White HD, Talley JD, Buller
bolysis in Myocardial Infarction trials. J Am Coll Cardiol 2003;42:7–16.
17. Moreno R, Lopez-Sendon J, Garcia E, Perez de Isla L, Lopez de Sa E, Ortega
CE, Jacobs AK, Slater JN, Col J, et al. Early revascularization in acute myocarA, Moreno M, Rubio R, Soriano J, Abeytua M, Garcia-Fernandez MA. Primary
dial infarction complicated by cardiogenic shock. SHOCK Investigators. Should
angioplasty reduces the risk of left ventricular free wall rupture compared with
We Emergently Revascularize Occluded Coronaries for Cardiogenic Shock.
thrombolysis in patients with acute myocardial infarction. J Am Coll Cardiol
N Engl J Med 1999;341:625– 634.
6. Dauerman HL, Goldberg RJ, White K, Gore JM, Sadiq I, Gurfinkel E, Budaj
2002;39:598 – 603.
18. Sugiura T, Nagahama Y, Nakamura S, Kudo Y, Yamasaki F, Iwasaka T. Left
A, Lopez de Sa E, Lopez-Sendon J. Revascularization, stenting, and outcomes of
ventricular free wall rupture after reperfusion therapy for acute myocardial
patients with acute myocardial infarction complicated by cardiogenic shock. Am J
infarction. Am J Cardiol 2003;92:282–284.
Cardiol 2002;90:838 – 842.
7. Hochman JS, Sleeper LA, White HD, Dzavik V, Wong SC, Menon V, Webb
19. Ohishi F, Hayasaki K, Honda T. Effect of thrombolysis on rupture of the left
JG, Steingart R, Picard MH, Menegus MA, et al. One-year survival following
ventricular free wall following acute myocardial infarction. J Cardiol 1996;28:
early revascularization for cardiogenic shock. JAMA 2001;285:190 –192.
27–32.
8. Antman EM, Armstrong PW, Bates ER, Green LA, Hand M, Hochman JS,
20. Sadanandan S, Buller C, Menon V, Dzavik V, Terrin M, Thompson B, Lamas
Krumholz HM, Kushner FG, Lamas GA, Mullany CJ, et al. ACC/AHA guideG, Hochman JS. The late open artery hypothesis—a decade later. Am Heart J
lines for the management of patients with ST-elevation myocardial infarction:
2001;142:411– 421.
TABLE 5 One-year Causes of Death in Patients Who Had Angiography Before
Discharge and Those Who Did Not
BRIEF REPORTS
233
Impact of ST-Segment Depression Resolution on
Mortality After Successful Mechanical Reperfusion in
Patients With ST-Segment Elevation Acute
Myocardial Infarction
Giuseppe De Luca, MD, PhD, Arthur C. Maas, MD, Arnoud W.J. van’t Hof,
Jan Paul Ottervanger, MD, PhD, Jan C.A. Hoorntje, MD, PhD,
A.T. Marcel Gosselink, MD, PhD, Jan-Henk E. Dambrink, MD, PhD,
Menko-Jan de Boer, MD, PhD, and Harry Suryapranata, MD, PhD
The aim of the present study was to evaluate the
additional prognostic effect of ST-depression resolution in 610 patients who had ST-elevation myocardial
infarction and underwent successful primary angioplasty (postprocedural Thrombolysis In Myocardial
Infarction 3 flow and complete resolution of ST-segment elevation). Incomplete resolution of ST-segment
depression (<70%) was observed in 50 patients
(8.2%). These patients were older, had a higher Killip’s class at presentation, had larger infarcts, and
had an increased 1-year mortality (10% vs 2%, p ⴝ
0.0004). At multivariate analysis, incomplete resolution of ST-segment depression was an independent
predictor of 1-year mortality (p ⴝ 0.028). 䊚2005
by Excerpta Medica Inc.
(Am J Cardiol 2005;95:234 –236)
he main goal in the control of ST-segment elevation acute myocardial infarction is an early and
T
sustained patency of the infarct-related artery.
1–3
However, several studies have shown that patency of
the epicardial vessel does not guarantee optimal myocardial perfusion.4 – 6 Electrocardiography represents a
inexpensive and easy method to evaluate myocardial
reperfusion.5–7 We5 previously showed that, among
patients who had postprocedural Thrombolysis In
Myocardial Infarction (TIMI) grade 3 flow, resolution
of ST-segment elevation is an independent prognostic
determinant after primary angioplasty. Despite the
negative prognostic implication of additional ST-segment depression in patients who have ST-segment
elevation acute myocardial infarction,8 –10 the prognostic role of resolved ST-segment depression has not
been investigated. This study investigated the prognostic role of resolved ST-segment depression in patients whose ST-segment elevation acute myocardial
infarction was successfully managed with primary
angioplasty.
•••
From the Department of Cardiology, ISALA Klinieken De Weezenlanden
Hospital, Zwolle, The Netherlands. Dr. Suryapranata’s address is:
ISALA Klinieken, De Weezenlanden Hospital, Department of Cardiology, Groot Wezeland 20, 8011 JW Zwolle, The Netherlands. Email: h.suryapranata@diagram-zwolle.nl. Manuscript received June
14, 2004; revised manuscript received and accepted September 1,
2004.
234
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MD, PhD,
A total of 1,548 patients who had ST-segment
elevation acute myocardial infarction underwent primary angioplasty at our institution between April
1997 and October 2001. Of 1,072 patients who had
complete data on resolution of ST-segment elevation
myocardial infarction, 900 showed baseline ST-segment depression. Our final patient population consisted of 610 patients who had baseline ST-segment
depression and successful reperfusion, defined as
postprocedural TIMI grade 3 flow and complete resolution of ST-segment elevation (ST-segment resolution ⱖ70%).7 Informed consent was obtained from
each patient before angiography. All subjects presented ⱕ6 hours from symptom onset or between 6
and 24 hours if they had continuous symptoms and
signs of ischemia (persistent or recurrent chest pain
and/or persistent elevation or reelevation of ST segment). All patients received aspirin (500 mg intravenously) and heparin (10,000 IU intravenously) before
the procedure. Because the benefits of glycoprotein
IIb/IIIa inhibitors have been proved only recently,18
⬍5% of subjects received this additional therapy. Coronary angiography and ST-segment resolution were
analyzed by an independent core laboratory (Diagram,
Zwolle, The Netherlands). TIMI flow grade and myocardial blush grade were assessed after primary angioplasty, as previously described.6 Enzymatic infarct
size was calculated by cumulative release of lactate
dehydrogenase from serial measurements ⱕ48 hours
after symptom onset. Analysis of ST-segment resolution was performed by comparison between baseline
and 3-hour 12-lead electrocardiograms, as previously
described.7 Patients were classified as having complete (ⱖ70%) and incomplete (⬍70%) resolution of
ST-segment depression. Records of patients who visited our outpatient clinic were reviewed. For all other
patients, information was obtained from the patient’s
general physician or by direct telephone interview
with the patient. For patients who died during followup, hospital records and necropsy data were reviewed.
One-year follow-up data were available in all patients.
Statistical analysis was performed with SPSS 10.0
(SPSS, Inc., Chicago, Illinois). Analysis of variance
and chi-square test were used appropriately for continuous and categorical variables. Differences in event
rates between groups during the follow-up period
were assessed by the Kaplan-Meier method using the
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.008
TABLE 1 Demographic, Clinical, and Angiographic
Characteristics According to Resolution of ST-segment
Depression
ST-depression
Resolution
Age (yrs)
Men
Diabetes mellitus
Hypertension
Previous myocardial
infarction
Killip’s class ⬎1
Anterior wall acute
infarction
Ischemia time ⬎4 h
Multivessel coronary
disease
TIMI 0/1 flow before
Myocardial blush
grade 2/3
Stent
Maximal baseline
ST-depression
(mm)
Complete
(n ⫽ 560)
Incomplete
(n ⫽ 50)
p Value
59 ⫾ 11
433 (77.3%)
40 (7.1%)
125 (22.9%)
49 (8.8%)
68 ⫾ 9
33 (66.0%)
5 (10%)
15 (30%)
8 (16.0%)
⬍0.0001
0.071
⬎0.1
⬎0.1
0.091
31 (5.5%)
215 (38.4%)
48 (16.0%)
16 (32.0%)
0.004
⬍0.1
177 (31.6%)
300 (53.6%)
17 (34.0%)
33 (66.0%)
⬎0.1
0.091
369 (65.9%)
505 (90.2%)
37 (74%)
46 (92.0%)
⬎0.1
⬎0.1
317 (56.6%)
2.22 ⫾ 1.56
29 (58.0%)
2.37 ⫾ 1.31
⬎0.1
⬎0.1
FIGURE 1. Kaplan-Meier survival curves according to resolution
of ST-segment depression.
TABLE 2 Predictors of One-year Mortality at Multivariate
Analysis
Relative risk (95% CI)
log-rank test. Multivariate analysis was performed
with Cox’s proportional hazard method to identify
independent predictors of 1-year mortality rate. Stepwise selection of variables and estimation of significant probabilities were computed by a maximal likelihood ratio test. The chi-square value was calculated
from the log of the ratio of maximal partial likelihood
functions. The additional value of each category of
variables added sequentially was evaluated on the
basis of the increases in the overall likelihood statistical ratio.
Incomplete resolution of ST-segment depression
was observed in 50 patients (8.2%). These patients
were older, had a higher Killip’s class at presentation,
showed a higher prevalence of women, and had more
previous infarctions (Table 1). Incomplete resolution
of ST-segment depression was associated with larger
infarcts (2,058 ⫾ 1,658 vs 1,417 ⫾ 1,253 U/L, p
⬍0.0001) and higher 1-year mortality (10% vs 2%,
relative risk 5.41, 95% confidence interval 1.88 to
15.58, p ⫽ 0.0004). Kaplan-Meier survival curves
according to resolution of ST-segment depression are
shown in Figure 1. At multivariate analysis, incomplete resolution of ST-segment depression was an
independent predictor of 1-year mortality (p ⫽ 0.028;
Table 2).
•••
The main finding of the present study is that,
among patients whose ST-segment elevation acute
myocardial infarction was successfully managed with
primary angioplasty, resolved ST-segment depression
does provide additional prognostic information to resolved ST-segment elevation. Despite optimal restoration of anterograde flow in the infarct-related artery
(TIMI grade 3 flow), suboptimal myocardial reperfusion may occur in a large percentage of patients.5,6
Evaluation of ST-segment resolution has been shown
to be a reliable, cheap, and user-friendly method to
Killip’s class ⬎1
Anterior infarction
Incomplete ST-depression
resolution
Age ⬎65 yrs
Myocardial blush grade 0/1
p Value
11.01 (4.92–30.2)
⬍0.0001
3.52 (1.21–10.2)
0.021
3.58 (1.15–11.15)
0.028
3.69 (1.15–11.8)
1.83 (0.55–6.12)
0.028
⬎0.1
CI ⫽ confidence interval.
analyze myocardial perfusion in patients who had
acute myocardial infarction that was controlled by
pharmacologic or mechanical reperfusion.5–7 Our previous report5 showed that, in patients who had postprocedural TIMI 3 flow, ST-segment resolution significantly added prognostic information in terms of
long-term mortality rate. The mechanism of concomitant ST-depression in patients who have acute myocardial infarction remains a matter of debate. It may
represent a “reciprocal” image of ST-segment elevation in the infarct zone,11,12 more extensive infarction,13,14 or additional ischemia beyond the infarct
zone.15,16 Previous reports have shown that ST-segment depression has significant prognostic implications in patients who have ST-segment elevation acute
myocardial infarction.8 –10 Thus, it is conceivable that
resolution of ST-segment depression would provide
additional prognostic information about resolution of
ST-segment elevation. This is the first study to explore
this issue in patients who had ST-segment elevation
acute myocardial infarction and underwent primary
angioplasty. To address this issue, we performed our
analysis in patients who had postprocedural TIMI
grade 3 flow and complete ST-segment resolution. We
found that postprocedural incomplete resolution of
ST-segment depression was associated with larger infarcts and subsequently with a higher mortality rate.
This finding was confirmed at multivariate analysis.
Analysis of resolved ST-segment elevation at a
fixed period after reperfusion therapy (3 hours in our
BRIEF REPORTS
235
study) may have limitations. Myocardial reperfusion
is a dynamic process during which alternating episodes of ST-segment resolution may occur.17 Fluctuation of the ST segment has been described during or
shortly after thrombolysis and primary angioplasty,
and continuous ST-segment monitoring has been
shown to be a valid alternative.17,18
1. The GUSTO Angiographic Investigators. The effects of tissue plasminogen
activator, streptokinase, or both on coronary-artery patency, ventricular function, and survival after acute myocardial infarction. N Engl J Med 1993;329:
1615–1622.
2. Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group. Indications for
fibrinolytic therapy in suspected acute myocardial infarction: collaborative overview of early mortality and major morbidity results from all randomised trials of
more than 1000 patients. Lancet 1994;343:311–322.
3. Rentrop KP. Restoration of anterograde flow in acute myocardial infarction:
the first 15 years. J Am Coll Cardiol 1995;25:1S–2S.
4. van ’t Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead
electrocardiogram after successful reperfusion therapy for acute myocardial infarction. Zwolle Myocardial infarction Study Group. Lancet 1997;350:615– 619.
5. van ’t Hof A, Liem A, Suryapranata H, Hoorntje JCA, de Boer MJ, Zijlstra F,
on the behalf of the Zwolle Myocardial Infarction Study Group. Angiographic
assessment of myocardial reperfusion in patients treated with primary angioplasty
for acute myocardial infarction. Myocardial Blush Grade. Circulation 1998;97:
2302–2306.
6. Ito H, Tomooka T, Sakai N, Yu H, Higashino Y, Fujii K, Masuyama T,
Kitabatake A, Minamino T. Lack of myocardial perfusion immediately after
successful thrombolysis. A predictor of poor recovery of left ventricular function
in anterior myocardial infarction. Circulation 1992;85:1699 –1705.
7. Schroder R, Dissmann R, Bruggemann T, Wegscheider K, Linderer T, Tebbe
U, Neuhaus KL. Extent of early ST segment elevation resolution: a simple but
strong predictor of outcome in patients with acute myocardial infarction. J Am
Coll Cardiol 1994;24:384 –391.
8. Croft CH, Woodward W, Nicod P, Corbett JR, Lewis SE, Willerson JT, Rude
RE. Clinical implications of anterior S-T segment depression in patients with
acute inferior myocardial infarction. Am J Cardiol 1982;50:428 – 436.
9. Shah A, Wagner GS, Califf RM, Boineau RE, Green CL, Wildermann NM,
Trollinger KM, Pope JE, Krucoff MW. Comparative prognostic significance of
simultaneous versus independent resolution of ST segment depression relative to
ST segment elevation during acute myocardial infarction. J Am Coll Cardiol
1997;30:1478 –1483.
10. Lembo NJ, Starling MR, Dell’Italia LJ, Crawford MH, Chaudhuri TK,
O’Rourke RA. Clinical and prognostic importance of persistent precordial (V1V4) electrocardiographic ST segment depression in patients with inferior transmural myocardial infarction. Circulation 1986;74:56 – 63.
11. Camara EJ, Chandra N, Ouyang P, Gottlieb SH, Shapiro EP. Reciprocal ST
change in acute myocardial infarction: assessment by electrocardiography and
echocardiography. J Am Coll Cardiol 1983;2:251–257.
12. Mukharji J, Murray S, Lewis SE, Croft CH, Corbett JR, Willerson JT, Rude
RE. Is anterior ST depression with acute transmural inferior infarction due to
posterior infarction? A vectorcardiographic and scintigraphic study. J Am Coll
Cardiol 1984;4:28 –34.
13. Wasserman AG, Ross AM, Bogaty D, Richardson DW, Hutchinson RG, Rios
JC. Anterior ST segment depression during acute inferior myocardial infarction:
evidence for the reciprocal change theory. Am Heart J 1983;106:516 –520.
14. Haraphongse M, Tanomsup S, Jugdutt BI. Inferior ST segment depression
during acute anterior myocardial infarction: clinical and angiographic correlations. J Am Coll Cardiol 1984;4:467– 476.
15. Pichler M, Shah PK, Peter T, Singh B, Berman D, Shellock F, Swan HJ. Wall
motion abnormalities and electrocardiographic changes in acute transmural myocardial infarction: implications of reciprocal ST segment depression. Am Heart J
1983;106(5 pt 1):1003–1009.
16. Tzivoni D, Chenzbraun A, Keren A, Benhorin J, Gottlieb S, Lonn E, Stern S.
Reciprocal electrocardiographic changes in acute myocardial infarction. Am J
Cardiol 1985;56:23–26.
17. Shah PK, Cercek B, Lew A, Ganz W. Angiographic validation of bedside
markers of reperfusion. J Am Coll Cardiol 1993;21:55– 61.
18. Krucoff MW, Croll MA, Pope JE, Granger CB, O’Connor CM, Sigmon
KN, Wagner BL, Ryan JA, Lee KL, Kereiakes DJ. Continuous 12-lead
ST-segment recovery analysis in the TAMI-7 study. Performance of a noninvasive method for real-time detection of failed myocardial reperfusion.
Circulation 1993;88:437– 446.
Effects of Tirofiban and Statins on High-Sensitivity
C-Reactive Protein, Interleukin-6, and Soluble CD40
Ligand Following Percutaneous Coronary Interventions
in Patients With Stable Coronary Artery Disease
Rabih R. Azar, MD, MSc, Georges Badaoui, MD, Antoine Sarkis, MD,
Roland Kassab, MD, Elie Salamé, MD, Samira Klaymé, PhD, Roger Naman, MD, and
Mirna Germanos, PharmD
This study assessed the effects of tirofiban and statins
on high-sensitivity C-reactive protein, interleukin-6,
and soluble CD40 ligand after percutaneous coronary
intervention in patients who had stable coronary artery disease. Tirofiban insignificantly limited the increase of soluble CD40 ligand after revascularization,
especially in patients who had high levels of this
marker at baseline (p ⴝ 0.06), whereas statins significantly inhibited increases in interleukin-6 and, to a
From the Division of Cardiology, the Laboratory of Histocompatibility,
and the Division of Immunology, Hotel Dieu de France Hospital and
the St-Joseph University School of Medicine, Beirut, Lebanon. This
study was supported by research grant FM 65 from the Council of
Research of St-Joseph University, Beirut, Lebanon. Dr. Azar’s address
is: Division of Cardiology, Hotel Dieu de France Hospital, Achrafieh,
Beirut, Lebanon. E-mail: razar@usj.edu.lb. Manuscript received July 9,
2004; revised manuscript received and accepted August 31, 2004.
236
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
lesser extent, high-sensitivity C-reactive protein without affecting the soluble CD40 ligand. 䊚2005 by
Excerpta Medica Inc.
(Am J Cardiol 2005;95:236 –240)
ercutaneous coronary intervention leads to activation of platelets and white cells and triggers
P
an acute inflammatory response that plays a major
role in the pathogenesis of complications after this
procedure.1,2 Antiplatelet therapy with tirofiban decreases ischemic complications of percutaneous intervention, especially in the setting of acute coronary syndromes, when inflammation is at its peak.3,4
However, effects of tirofiban on inflammation and platelet activation and their release of the proinflammatory
mediator, soluble CD40 ligand (sCD40l), remain unknown. In addition, statins improve the outcome of patients who undergo percutaneous revascularization. The
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.093
TABLE 1 Baseline Clinical and Procedural Characteristics of
Tirofiban and Saline Groups
Variable
Age (yrs)
Men
Diabetes mellitus
Hyperlipidemia
Current smoker
Hypertension
Family history of coronary artery
disease
Healed myocardial infarction
Previous coronary bypass
Current statin therapy
3-Vessel coronary artery disease
Percent baseline stenosis
Percent residual stenosis
Average length of stent
Maximum inflation pressure
Tirofiban
(n ⫽ 35)
Saline
(n ⫽ 35)
59 ⫾ 9
30 (86%)
13 (37%)
29 (83%)
10 (29%)
19 (54%)
16 (46%)
63 ⫾ 10
29 (83%)
8 (23%)
30 (86%)
15 (43%)
24 (69%)
10 (29%)
9 (26%)
5 (14%)
22 (63%)
4 (11%)
85 ⫾ 12
1⫾5
18 ⫾ 7
13 ⫾ 3
5 (14%)
5 (14%)
22 (63%)
4 (11%)
88 ⫾ 10
2⫾5
18 ⫾ 4
14 ⫾ 2
Data are numbers of patients (percentages) or mean ⫾ SD.
TABLE 2 Baseline Clinical and Procedural Characteristics in
Statin and Control Groups
Variable
Age (yrs)
Men
Diabetes mellitus
Hyperlipidemia
Current smoker
Hypertension
Family history of coronary artery
disease
Healed myocardial infarction
Previous coronary bypass
3-Vessel coronary artery disease
Tirofiban during PCI
Percent baseline stenosis
Percent residual stenosis
Average length of stent
Maximum inflation pressure
Statin
(n ⫽ 44)
Control
(n ⫽ 26)
59 ⫾ 10
35 (79%)
14 (32%)
41 (93%)
17 (39%)
24 (54%)
18 (41%)
63 ⫾ 8
24 (92%)
7 (27%)
8 (69%)*
8 (31%)
19 (73%)
8 (31%)
10 (23%)
6 (14%)
5 (11%)
22 (50%)
86 ⫾ 10
1⫾5
18 ⫾ 6
14 ⫾ 3
4 (15%)
4 (15%)
3 (11%)
13 (50%)
86 ⫾ 12
1⫾5
19 ⫾ 5
14 ⫾ 2
*p ⫽ 0.008.
Data are numbers of patients (percentages) or mean ⫾ SD.
PCI ⫽ percutaneous coronary intervention.
mechanism of this benefit was hypothesized to be an
anti-inflammatory activity,5–7 but this has not been directly demonstrated. This study investigated the effects
of tirofiban and statins on serum levels of high-sensitivity C-reactive protein (hs-CRP), interleukin-6 (IL-6), and
sCD40l after percutaneous coronary intervention in patients who have stable coronary artery disease.
•••
Patients who had stable coronary artery disease and
who were referred for elective percutaneous coronary
intervention for a de novo lesion in a native coronary
artery were screened to participate in this study. Patients who had proinflammatory conditions or were
taking anti-inflammatory drugs were excluded. Informed consent was obtained from all patients, and the
study protocol was approved by the institutional review board of the Hotel Dieu de France Hospital
(Beirut, Lebanon).
FIGURE 1. Mean changes ⴞ SD in levels of sCD40l according to
sCD40l quartiles at baseline (p <0.001 by 1-way analysis of
variance).
TABLE 3 Pre- and Postprocedural Levels of All Markers in the
Tirofiban and Saline Groups
Tirofiban
(n ⫽ 35)
hs-CRP (mg/L)
Baseline
48 h
Mean change
IL-6 (pg/ml)
Baseline
48 h
Mean change
sCD40l (ng/ml)
Baseline
48 h
Mean change
7 ⫾ 10
17.3 ⫾ 16
10.3 ⫾ 10
Saline
(n ⫽ 35)
7.7 ⫾ 6.9
19.6 ⫾ 17.2
11.9 ⫾ 14.7
3.3 ⫾ 2.5
6 ⫾ 4.5
2.7 ⫾ 4
4.5 ⫾ 4
7.2 ⫾ 6.4
2.7 ⫾ 4.4
4 ⫾ 2.9
6.3 ⫾ 3.1*
2.2 ⫾ 3.6
4.5 ⫾ 2.9
8⫾4
3.5 ⫾ 4.2
Data are mean ⫾ SD.
*p ⫽ 0.05 versus saline; p ⬍0.01 for all values at 48 hours versus baseline.
The study was randomized, open label, and
prospective. At the time of percutaneous coronary
intervention and before crossing the lesion with a
guidewire, patients were randomized to receive an
intravenous bolus and a 24-hour infusion of tirofiban
or normal saline. Percutaneous coronary intervention
was performed in the standard fashion and consisted
of balloon angioplasty and stenting. Tirofiban was
given at a dose of 10 ␮g/kg as an intravenous bolus
and as a 0.15-␮g/kg/min continuous infusion for 24
hours. Patients who were receiving statins at the time
of the procedure were defined as statin users. Levels
of hs-CRP, IL-6, and sCD40l were measured immediately before and 48 hours after the procedure, and
cardiac troponin T was measured 24 hours after percutaneous coronary intervention. Study end points
were changes in marker levels and in levels 48 hours
after the procedure.
BRIEF REPORTS
237
(Table 1) and between the statin and control groups
(Table 2). Pretreatment with clopidogrel at a loading
dose of 300 mg 18 to 24 hours before the procedure
Statin
Control
was given to 94% of patients (equally between groups:
(n ⫽ 44)
(n ⫽ 26)
tirofiban vs saline and statin vs control). There were
hs-CRP (mg/l)
no complications, and all patients had undetectable
Baseline
7.3 ⫾ 7.1
7.4 ⫾ 10.8
troponin levels 24 hours after revascularization.
48 h
16.8 ⫾ 13.3
21.3 ⫾ 21
Angioplasty resulted in significant increases in hsMean change
9.4 ⫾ 10.6
13.9 ⫾ 15
CRP, IL-6, and sCD40l (150%, 69%, and 65%, reIL-6 (pg/ml)
Baseline
3.9 ⫾ 2.8
3.9 ⫾ 4.2
spectively, vs baseline, p ⬍0.01 for all). Changes in
48 h
5.6 ⫾ 3.9*
8.3 ⫾ 7.3
hs-CRP and IL-6 correlated closely together (r ⫽ 0.5,
†
Mean change
1.6 ⫾ 3.9
4.3 ⫾ 4.1
p ⬍0.001) but did not correlate with changes in
sCD40l (ng/ml)
sCD40l (r ⫽ 0.2 for hs-CRP vs sCD40l and r ⫽ 0.05
Baseline
4.2 ⫾ 3
4.5 ⫾ 2.8
for IL-6 vs sCD40l, p ⫽ NS for the 2 comparisons).
48 h
7⫾4
7.4 ⫾ 3.4
Mean change
2.8 ⫾ 3.8
2.9 ⫾ 4.1
However, the increase in sCD40l correlated strongly
†
with the baseline level of this marker. Patients in the
*p ⱕ0.05 and p ⫽ 0.008 versus controls; p ⬍0.01 for all values at 48
upper quartile of sCD40l did not show a statistically
hours versus baseline.
Data are mean ⫾ SD.
significant change in this marker after the procedure,
whereas those in the lower 3 quartiles showed a significant increase of
this marker (p ⬍0.001 for intergroup
comparison; Figure 1).
Effects of tirofiban are listed in
Table 3. Levels of hs-CRP and IL-6
increased equally in the tirofiban and
saline groups, but sCD40l increased
less (p ⫽ NS) in the tirofiban group,
resulting in a statistically lower level
of sCD40l at 48 hours compared
with the control group (p ⫽ 0.05).
Because sCD40l did not significantly increase after angioplasty in
patients who had high levels of this
marker at baseline (highest quartile),
the effect of tirofiban was tested in
the 3 lower quartiles of sCD40l in
which percutaneous coronary interFIGURE 2. Changes in hs-CRP according to baseline hs-CRP levels and statin use (p ⴝ
vention significantly increased this
0.005 for intergroup comparison in controls and p ⴝ NS in the statin group by
marker. In these patients, the inhibi1-way analysis of variance).
tory effect of tirofiban was more pronounced (mean changes 2.7 ⫾ 3.7 vs
4.6 ⫾ 3.5 ng/ml for tirofiban vs saBlood samples were immediately taken to the labo- line, p ⫽ 0.06), resulting in much lower postproceratory, where serum was frozen and stored at ⫺80°C dural levels of sCD40l in the tirofiban group (5.5 ⫾
until batch analysis. hs-CRP was measured by nephe- 3.1 vs 7.7 ⫾ 3.5 ng/ml, p ⫽ 0.02).
lometry (N Latex CRP Mono, Dade-Behring, Marburg,
Effects of statins are presented in Table 4. Statin
Germany). IL-6 and sCD40l were measured with high- therapy did not affect the increase in sCD40l but did
sensitivity quantitative immunoassays (R&D Systems, attenuate the increase of hs-CRP by 32% (p ⫽ NS)
Abingdon, United Kingdom). Assay sensitivities were and of IL-6 by 63% (p ⫽ 0.008). However, when the
0.7 pg/ml for IL-6 and 4.2 pg/ml for sCD40l. Intra- and change in hs-CRP was measured according to baseline
interassay variabilities were ⱕ6% for all tests.
quartile levels of this marker, statin therapy signifiData were summarized as mean ⫾ SD or as per- cantly attenuated the increase of hs-CRP by 71% in
centages and compared with Student’s t test or chi- the subgroup of patients in the upper quartile (mean
square test. One-way analysis of variance was used to change 9 ⫾ 10 vs 31 ⫾ 22 mg/L for statin vs control,
compare changes in markers and effects of therapies p ⫽ 0.01). As shown in Figure 2, there was a statisin multiple subgroups. Pearson’s test was used to
tically significant interaction between baseline levels
study the correlation across changes in hs-CRP, IL-6,
and sCD40l. All tests were 2-tailed, and a p value of hs-CRP, statin therapy, and increased hs-CRP after
percutaneous revascularization.
⬍0.05 was considered statistically significant.
•••
Baseline clinical and procedural characteristics of
This study demonstrates that, in patients who have
the 70 patients who were enrolled in the study
were similar between the tirofiban and saline groups stable coronary artery disease, percutaneous coronary
TABLE 4 Pre- and Postprocedural Levels of All Markers in
Statin and Control Groups
238 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
interventions induce a systemic inflammatory response similar to that seen in acute coronary syndromes. It also demonstrates platelet activation, as
demonstrated by increased levels of sCD40l. This
molecule is a potent proinflammatory and prothrombotic agent8 –11 and is a major link in interactions
between platelets and white blood cells. However,
there was no correlation between changes in hs-CRP
and sCD40l or between IL-6 and sCD40l, suggesting
that inflammation after percutaneous coronary intervention is independent of platelets and is primarily
driven by activation of white blood cells. The effects
of tirofiban and statins on these inflammatory markers
provide additional support for this statement.
This study suggests that tirofiban inhibits the release
of sCD40l by platelets but does not attenuate the acute
inflammatory response triggered by percutaneous coronary intervention. Several mechanisms may explain
these findings. Although tirofiban blocks in vitro the
release of sCD40l from platelets in a dose-dependent
manner,12 the decrease in sCD40l observed in our trial
was of borderline significance and may have been insufficient to suppress inflammation. In addition, tirofiban
can theoretically attenuate the formation of platelet/leukocyte complexes by blocking the binding of the platelets’ glycoprotein IIb/IIIa receptor to the leukocyte’s
MAC-1 receptor.13 However, these complexes can always form by binding the platelet’s P-selectin to the
leukocyte’s P-selectin glycoprotein ligand-1.14 Leukocytes may also be directly activated by rupture of the
culprit plaque (independent of their interaction with
platelets). Blockage of the leukocyte MAC-1 receptor
seems essential for suppression of inflammation. This
receptor is not affected by tirofiban but is effectively
blocked by abciximab,15,16 which has been shown to
inhibit the increases in hs-CRP, IL-6, and tumor necrosis
factor-␣ by 32% to 100% after percutaneous coronary
intervention.17
Statins attenuated the acute inflammatory response
triggered by percutaneous intervention. The greater
effect of statins on IL-6 compared with hs-CRP most
likely reflects the greater stability of CRP in the circulation (half-life of 19 vs 4 hours for IL-6).18 In
addition, IL-6 is secreted by activated white blood
cells and stimulates CRP synthesis by the liver.18
Thus, IL-6 is the first to decrease after inhibition of
leukocytes, whereas a decrease in hs-CRP is usually
delayed by an additional 24 to 48 hours. The mechanism of statins’ anti-inflammatory effect is independent of platelets, because these drugs did not attenuate
the release of sCD40l and is most likely secondary to
direct inhibition of leukocyte activation and to plaque
stabilization.19
The implications of the present findings should be
considered within the context of recently published
trials. Chan et al6 demonstrated a survival benefit
associated with statin therapy at the time of percutaneous coronary intervention. In their cohort of 5,052
patients, statin therapy before revascularization resulted in 53% and 67% decreases in mortality at 30
days and 6 months, respectively. In a subsequent
report,7 they found that this benefit was mainly ob-
served among patients within the highest hs-CRP
quartile before intervention and hypothesized that it
was secondary to the anti-inflammatory properties of
statins. Our results support their conclusion.
Our study has several limitations. First, the suppressive effect of tirofiban on the increase of
sCD40l was of borderline significance (p ⫽ 0.06
after exclusion of the upper quartile group). However, levels of sCD40l after the procedure were
significantly lower in patients who received tirofiban, whether or not patients in the upper sCD40l
quartile were excluded. Second, we measured
sCD40l levels 48 hours after angioplasty, whereas
tirofiban infusion was given for only 24 hours. The
magnitude of the decrease in sCD40l might have
been greater if measurements had been done at 24
hours. Third, we limited our study to patients who
had stable coronary artery disease. Fourth, statin
use was not randomized. Although statins users and
nonusers shared similar characteristics, there may
have been unmeasurable potential confounders that
affected the results. Fifth, the duration and doses of
statins before revascularization were not collected.
Statins might have been started shortly before percutaneous coronary intervention and/or the dosing
may have been inadequate. However, in that case,
the true anti-inflammatory effect of statins may be
even more important than indicated by our findings.
1. Azar RR, McKay RG, Kiernan FJ, Seecharran B, Feng YJ, Fram DB, Wu
AHB, Waters DD. Coronary angioplasty induces a systemic inflammatory response. Am J Cardiol 1997;80:1476 –1478.
2. Gaspardone A, Crea F, Versaci F, Tomai F, Pellegrino A, Chiariello L, Gioffre
PA. Predictive value of C-reactive protein after successful coronary-artery stenting in patients with stable angina. Am J Cardiol 1998;82:515–518.
3. Platelet Receptor Inhibition in Ischemic Syndrome Management in Patients
Limited by Unstable Signs and Symptoms (RRISM-PLUS) Study Investigators. Inhibition of the platelet glycoprotein IIb/IIIa receptor with tirofiban in
unstable angina and non–Q-wave myocardial infarction. N Engl J Med 1998;
338:1488 –1497.
4. The RESTORE Investigators. Effects of platelet glycoprotein IIb/IIIa
blockade with tirofiban on adverse cardiac events in patients with unstable
angina or acute myocardial infarction undergoing coronary angioplasty. Randomized Efficacy Study of Tirofiban for Outcomes and Restenosis. Circulation 1997;96:1445–1453.
5. Albert MA, Danielson E, Rifai N, Ridker PM. Effect of statin therapy on
C-reactive protein levels: the pravastatin inflammation/CRP evaluation (PRINCE): a
randomized trial and cohort study. JAMA 2001;286:64 –70.
6. Chan AW, Bhatt DL, Chew DP, Quinn MJ, Moliterno DJ, Topol EJ, Ellis SG.
Early and sustained survival benefit associated with statin therapy at the time of
percutaneous coronary intervention. Circulation 2002;105:691– 696.
7. Chan AW, Bhatt D, Chew DP, Reginelli J, Schneider JP, Topol EJ, Ellis SG.
Relation of inflammation and benefit of statins after percutaneous coronary
interventions. Circulation 2003;107:1750 –1756.
8. André P, Nannizzi-Alaimo L, Prasad SK, Phillips DR. Platelet-derived
CD40L. The switch-hitting player of cardiovascular disease. Circulation
2002;106:896 – 899.
9. Schonbeck U, Libby P. The CD40/CD154 receptor/ligand dyad. Cell Mol Life
Sci 2001;58:4 – 43.
10. Henn V, Slupsky J, Grafe M, Anagnostopoulos I, Forster R, Muller-Berghaus
G, Kroczek RA. CD40 ligand on activated platelets triggers an inflammatory
reaction of endothelial cells. Nature 1998;391:591–594.
11. André P, Prasad SK, Denis CV, He M, Papalia JM, Hynes RO, Phillips DR,
Wagner DD. CD40L stabilizes arterial thrombi by a beta-3 integrin-dependent
mechanism. Nat Med 2002;8:247–252.
12. Nannizzi-Alaimo N, Alves VL, Phillips DR. Inhibitory effects of glycoprotein IIb/IIIa antagonists and aspirin on the release of sCD40l during platelet
stimulation. Circulation 2003;107:1123–1128.
13. Weber C, Springer TA. Neutrophil accumulation on activated, surfaceadherent platelets in flow is mediated by interaction of Mac-1 with fibrinogen
bound to alpha-IIb beta-3 and stimulated by platelet activating factor. J Clin
Invest 1997;100:2085–2093.
BRIEF REPORTS
239
14. Sarma J, Laan CA, Alam S, Jha A, Fox KAA, Dransfield I. Increased platelet
17. Lincoff AM, Kereiakes DJ, Mascelli MA, Deckelbaum LI, Barnathan ES,
binding to circulating monocytes in acute coronary syndromes. Circulation
2002;105:2166 –2171.
15. Schwarz M, Nordt T, Bode C, Peter K. The GP IIb/IIIa inhibitor abciximab
(c7E3) inhibits the binding of various ligands to the leukocyte integrin Mac-1
(CD11b/CD18, alphaMbeta2). Thromb Res 2002;107:121–128.
16. Mickelson JK, Ali MN, Kleiman NS, Lakkis NM, Chow TW, Hughes BJ,
Smith CW. Chimeric 7E3 Fab (ReoPro) decreases detectable CD11b on neutrophils from patients undergoing coronary angioplasty. J Am Coll Cardiol 1999;
33:97–106.
Patel KK, Frederick B, Nakada MT, Topol EJ. Abciximab suppresses the rise in
levels of circulating inflammatory markers after percutaneous coronary revascularization. Circulation 2001;104:163–167.
18. Liuzzo G, Buffon A, Biasucci LM, Gallimore JR, Caligiuri G, Vitelli A,
Altamura S, Ciliberto G, Rebuzzi AG, Crea F, et al. Enhanced inflammatory
response to coronary angioplasty in patients with severe unstable angina. Circulation 1998;98:2370 –2376.
19. Bonetti PO, Lerman LO, Napoli C, Lerman A. Statin effects beyond lipid
lowering—are they clinically relevant? Eur Heart J 2003;24:225–248.
Value of Preprocedure Multislice Computed
Tomographic Coronary Angiography to Predict the
Outcome of Percutaneous Recanalization of Chronic
Total Occlusions
Nico R. Mollet, MD, Angela Hoye, MB, ChB, MRCP, Pedro A. Lemos, MD,
Filippo Cademartiri, MD, Georgios Sianos, MD, Eugene P. McFadden, MB,
Gabriel P. Krestin, MD, Patrick W. Serruys, MD, and Pim J. de Feyter, MD, PhD
We performed multislice computed tomographic coronary angiography in 45 patients who had chronic total
occlusions and were scheduled for percutaneous recanalization. Multivariate analysis identified a blunt stump
(by conventional angiography), occlusion length >15
mm, and severe calcification (by multislice computed
tomographic coronary angiography) as independent
predictors of procedural failure. 䊚2005 by Excerpta
Medica Inc.
(Am J Cardiol 2005;95:240 –243)
ixteen-row multislice spiral computed tomographic (MSCT) coronary angiography has reS
cently been shown to allow reliable noninvasive
evaluation of coronary morphology.1–3 In the
present study, we analyzed the potential of preprocedural MSCT coronary angiography to provide
additional information and thus predict the procedural outcome in patients who had chronic total
occlusion (CTO) and were referred for percutaneous coronary recanalization.
•••
Forty-five patients referred for percutaneous recanalization of ⱖ1 CTO lesion underwent MSCT coronary angiography before the coronary procedure (median interval 29 days, interquartile range 9 to 53). The
diagnosis of CTO was made on diagnostic angiograms
that demonstrated complete occlusion of a major epicardial coronary artery, which was deemed to be of
ⱖ3 months’ duration from the date from the previous
angiogram, a clinical history of myocardial infarction,
or onset of or a severe episode of prolonged anginal
From the Department of Cardiology, Thoraxcenter and Department of
Radiology, Erasmus Medical Center, Rotterdam, The Netherlands.
Dr. de Feyter’s address is: Departments of Cardiology and Radiology,
Thoraxcenter, Erasmus Medical Center, Bd 410, P.O. Box
2040, 3000 CA Rotterdam, The Netherlands. E-mail: p.j.defeyter@
erasmusmc.nl. Manuscript received June 28, 2004; revised manuscript received and accepted September 2, 2004.
240
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
chest pain. In addition, inclusion into the study required a serum creatinine level ⬍120 mmol/L, presence of sinus rhythm, and the ability to hold a breath
for 20 seconds. The protocol was approved by the
institutional review board, and all patients gave written informed consent.
Conventional angiographic assessment was performed by observers who were unaware of the results
of MSCT scans. Parameters previously reported to
have prognostic importance for procedural failure
were assessed: absence of anterograde flow through
bridging collaterals, absence of a tapered stump, presence of severe calcification at the occluded segment,
side branch at the occlusion site, and tortuosity of the
vessel proximal to the occlusion (defined as an angle
⬎45° in any projection). Where possible, occlusion
length was measured from the view with the longest
lesion on quantitative coronary angiography as the
distance between a stump and a distal vessel as visualized by anterograde filling through bridging collaterals. In addition, in some other patients, length was
determined from the baseline angioplastic procedure
film using a bilateral coronary injection.
Twenty-two patients who had a heart rate ⬎65 beats/
min before multislice spiral computed tomography received an oral dose of 100 mg of metoprolol 1 hour
before scanning. All examinations were performed with
a 16-row MSCT scanner (Sensation 16, Siemens, Forchheim, Germany; collimation 16 ⫻ 0.75 mm, rotation
time 420 ms, table feed 3.0 mm/rotation, tube voltage
120 kV, tube current 400 to 450 mA). After intravenous
administration of 120 ml of nonionic contrast material
(Visipaque 320, Amersham Health, Little Chalfont,
United Kingdom), an automatic bolus-tracking technique triggered the start of MSCT scanning. Images were
reconstructed with retrospective electrocardiographic
gating during the mid- to end-diastolic phase to provide
nearly motion-free image quality; additional reconstruc0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.009
Patients’ mean age ⫾ SD was
57.0 ⫾ 10.1 years, 40 (89%) were
men, 10 (22%) were diabetic, and 14
(31%) had multivessel disease. ForSuccess
Failure
ty-seven CTO lesions were treated.
(n ⫽ 26 lesions)
(n ⫽ 21 lesions)
p Value
Angiographic measurement of CTO
Mean total procedural time (min)
148 ⫾ 61
148 ⫾ 44
1.0
length was possible in 39 lesions
Mean volume of contrast used (ml)
451 ⫾ 258
453 ⫾ 265
1.0
(83%), 31 (66%) from the diagnostic
Mean no. of wires
2.0 ⫾ 1.2
2.1 ⫾ 0.8
0.9
film and an additional 8 (17%) from
Wire type*
Graphix Intermediate
20 (76.9%)
12 (57.1%)
0.2
bilateral injection and assessment of
Choice PT Plus
5 (19.2%)
3 (14.3%)
0.8
retrograde collateral filling. The
Crosswire NT
8 (30.8%)
10 (47.6%)
0.3
mean length of occlusion was longer
Miracle
5 (19.2%)
12 (57.1%)
⬍0.01
by MSCT coronary angiography
Intraluminal
3 (11.5%)
4 (19.0%)
0.5
than by angiography (21.8 ⫾ 18.6 vs
Over-the-wire balloon support
15 (57.7%)
17 (81.0%)
0.1
14.6 ⫾ 10.9 mm, respectively). Pro*Wire type not mutually exclusive.
cedural data, including type and
number of guidewires used, are presented in Table 1. The only differtion windows (e.g., early diastolic phase) were explored ence between success and failure was in the increased
use of the Miracle wire in the failure group. Overall
when necessary.
All MSCT scans were analyzed off-line by op- mean procedural time was 148 ⫾ 53 minutes, with a
erators who were blinded to angiographic and pro- mean fluoroscopic time of 47 ⫾ 24 minutes. Overall,
cedural data. Parameters similar to those of conven- 45% of interventional procedures failed (Table 2).
tional angiography were evaluated: a blunt rather Success versus failure was not dependent on the opthan tapered stump, severe calcification, side branch erator or choice of interventional strategy. At univarat the occlusion site, proximal tortuosity, and oc- iate analysis, the following were associated with proclusion length. Severe calcification was defined as cedural failure: clinical assessment (occlusion
the presence of high-density plaques (ⱖ130 HU) duration ⱖ9 months), angiographic assessment (lack
involving ⬎50% of the coronary wall on a cross- of anterograde collateral filling, a blunt rather than a
sectional image and localized within the occlusion tapered stump, and side branch at the occlusion site),
and MSCT coronary angiographic assessment (a blunt
stump or occluded segment.
All procedures were performed by operators who rather than a tapered stump, severe calcification, and
were highly experienced in the treatment of CTOs, occlusion length ⬎15 mm; Table 2).
When analyzed separately, the following “tradiwith the interventional strategy left to the discretion of
the operator. Wires were used in a stepwise progres- tional” clinical and angiographic characteristics were
sion, starting with a wire that had a relatively less identified as multivariate predictors of procedural failtraumatic tip (Graphix Intermediate, Boston Scientific ure: occlusion duration ⬎9 months and stump morCorporation, Miami, Florida) or a hydrophilic wire phology (Table 3). Separate multivariate analysis that
(Choice PT Plus, Boston Scientific Corporation, or assessed only MSCT coronary angiographic parameCrosswire NT, Terumo Corporation, Tokyo, Japan) ters identified the following predictors: occlusion
and progressing to stiffer wires (Miracle, Asahi Intec, length ⬎15 mm, severe calcification, and stump morNagoya, Japan) and specialized technologies (Safe- phology. Final best model testing for pooled clinical,
Cross, Intraluminal Therapeutics, Carlsbad, New angiographic, and MSCT parameters identified a blunt
Mexico).4,5 Procedural failure was defined as an in- rather than a tapered stump (by angiography), occlusion length ⬎15 mm (by MSCT coronary angiograability to cross the occlusion with a guidewire.
Multivariate logistic regression analyses were per- phy), and severe calcification (by MSCT coronary
formed to identify angiographic and MSCT parame- angiography) as multivariate independent predictors
ters associated with procedural failure (all univariate of procedural failure (Figure 1).
predictors with a p value ⱕ0.1 were tested for their
•••
multivariate predictive value, and final models were
The current selection process of technically appropribuilt by backward stepwise selection). Angiographic ate candidates for percutaneous recanalization of CTO is
parameters assessed were those identified in previous based on the evaluation of a relatively restricted number
studies6: the occluded artery, duration of occlusion, of clinical and angiographic characteristics. In this study,
multivessel disease, anterograde and retrograde collat- we show that noninvasive evaluation of patients who
eral filling, type of stump, side branch at the site of have CTO by preprocedural MSCT coronary angiograocclusion, calcific deposits, vessel tortuosity, and oc- phy improves the ability to predict the outcome of a
clusion length ⬎15 mm. The predictive strengths of percutaneous recanalization attempt. Our findings indithe models were evaluated by means of the ⫺2 log- cate that MSCT coronary angiography may aid in the
likelihood statistic, and models’ lack of fit with the therapeutic decision making for patients who have CTO.
Hosmer-Lemeshow test, and their global predictive In addition, accurate preprocedural characterization of
accuracy were assessed by the C index (area under the CTO features may assist in outlining the therapeutic
interventional strategy.
receiver-operating characteristic curve).
TABLE 1 Procedural Duration and Use of Contrast Material and Guidewires for
Percutaneous Intervention of a CTO With Respect to Successful Versus
Unsuccessful Recanalization
BRIEF REPORTS
241
TABLE 2 Clinical, Angiographic, and MSCT Coronary Angiography Lesion Characteristics (n ⫽ 47)
Angiography
Frequency
(%)
Right coronary artery
Left anterior descending
artery
Left circumflex artery
Occlusion duration ⱖ9 mo
Multivessel disease
Anterograde filling
Retrograde collateral filling
Bridging collaterals
Tapered stump
Stump morphology not
determinable
Side branch at occlusion site
Severe calcification
Vessel tortuosity
Occlusion length ⬎15 mm
Occlusion length not
determinable
Overall
Failure
Rate
(%)
MSCT Coronary Angiography
OR (95% CI)
p Value
Frequency
(%)
Failure
Rate
(%)
OR (95% CI)
p Value
0.5
0.6
—
—
—
—
—
—
—
—
43
43
50
40
1.45 (0.45–4.66)
0.72 (0.22–2.32)
15
51
34
60
72
38
60
0
43
63
38
32
44
39
25
—
0.92
4.72
0.64
0.27
0.92
0.68
0.12
(0.18–4.64)
(1.36–16.39)
(0.19–2.20)
(0.08–0.94)
(0.26–3.32)
(0.21–2.25)
(0.03–0.45)
—
0.9
0.02
0.5
0.04
0.9
0.5
⬍0.01
—
—
—
—
—
—
—
60
15
—
—
—
—
—
—
32
43
—
—
—
—
—
—
0.16 (0.03–0.73)
0.25 (0.34–1.82)
—
—
—
—
—
—
0.02
0.2
57
34
23
26
17
59
50
64
67
50
4.37
1.39
2.75
3.39
1.29
(1.23–15.54)
(0.41–4.65)
(0.68–11.14)
(0.85–13.48)
(0.28–5.94)
0.02
0.2
0.08
0.7
45
38
21
51
0
52
67
50
63
—
1.76
4.44
1.31
4.72
0.3
0.02
0.7
0.02
—
100
45
—
—
—
—
(0.55–5.64)
(1.27–15.61)
(0.32–5.32)
(1.36–16.39)
—
—
—
CI ⫽ confidence interval; MSCT ⫽ multislice computed tomographic; OR ⫽ odds ratio.
TABLE 3 Angiographic and MSCT Coronary Angiographic Multivariate Predictors of Procedural Failure for Chronic Total
Occlusion
Variable
Clinical/angiographic
predictors
Occlusion duration ⬎9 mo
Tapered stump
Constant
MSCT coronary angiography
predictors
Occlusion length ⬎15 mm
Severe calcification
Stump morphology
Blunt
Tapered
Not determinable
Constant
Clinical/angiographic ⫹
MSCT coronary
angiographic predictors
Tapered stump*
Occlusion length ⬎15 mm
Severe calcification
Constant
Coefficient
1.27
⫺1.93
0.24
1.86
2.49
—
1 (reference)
⫺2.19
⫺2.65
⫺0.45
⫺2.43
2.17
2.03
⫺0.67
Wald’s
Chi-square
3.30
7.46
0.13
5.21
6.51
5.63
—
5.23
3.46
0.26
7.98
6.16
5.18
0.74
DF
p Value
1
1
1
1
1
2
—
1
1
1
1
1
1
1
0.07
⬍0.01
0.7
0.02
0.01
0.06
—
0.02
0.06
0.6
⬍0.01
0.01
0.02
0.4
OR (95% CI)
HosmerLemeshow
Test
⫺2 Log
Likelihood
DF
p Value
C
Index
50.0
2
0.66
0.80
44.2
6
0.99
0.84
41.0
5
0.60
0.85
3.56 (0.90–14.02)
0.15 (0.04–0.58)
—
6.39 (1.30–31.41)
12.01 (1.78–81.1)
—
—
0.11 (0.02–0.73)
0.07 (0.00–1.15)
—
0.09 (0.02–0.48)
8.77 (1.58–48.76)
7.62 (1.33–43.74)
—
*A ⫺2 log-likelihood change in the global model if 1 variable is removed: tapered stump ⫺10.7 (p ⬍0.01 for change), occlusion length ⫺8.2 (p ⬍0.01), and
calcification ⫺6.6 (p ⫽ 0.01 for change).
DF ⫽ degrees of freedom; other abbreviations as in Table 2.
MSCT coronary angiography of CTOs adds important information compared with “conventional” coronary
angiography. The length of the occluded segment has
long been identified as an important predictor of failed
recanalization. However, accurate measurement of lesion length using conventional angiography may be difficult, mainly due to foreshortening, calibration limita242 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
tions, and lack of visualization of the distal vessel in the
absence of collateral filling. In the present study, lesion
length could be measured in only 66% of diagnostic
films. Conversely, MSCT coronary angiography allowed
reliable 3-dimensional length measurement of coronary
segments.7 In the present series, when angiographic and
MSCT coronary angiographic occlusion lengths were
JANUARY 15, 2005
phy for interventions scheduled to be
performed shortly after scanning.
However, the elective nature of CTO
recanalization angioplasty allows a
safe time lag between these procedures. In our study, multislice computed tomography was performed ⬃1
month before coronary intervention.
The relatively high radiation exposure
during MSCT coronary angiography,
reportedly between 6.7 and 13.0
mSv,8 –10 remains a matter of concern.
However, MSCT coronary angiography
may optimize therapeutic strategy (e.g.,
calcifications may require intraluminal
techniques), resulting in shorter procedures. MSCT coronary angiography is
currently feasible for selected patients,
and further studies are needed to evaluate its value in a more general patient
population.
1. Nieman K, Cademartiri F, Lemos PA, Raaijmakers R,
Pattynama PM, de Feyter PJ. Reliable noninvasive coronary angiography with fast submillimeter multislice spiral
computed tomography. Circulation 2002;106:2051–2054.
2. Ropers D, Baum U, Pohle K, Anders K, Ulzheimer S,
Ohnesorge B, Schlundt C, Bautz W, Daniel WG, Achenbach
S. Detection of coronary artery stenoses with thin-slice
multi-detector row spiral computed tomography and multiplanar reconstruction. Circulation 2003;107:664 – 666.
FIGURE 1. (A, inset) CTO of the left anterior descending coronary artery with favor3. Mollet NR, Cademartiri F, Nieman K, Saia F, Lemos
able invasive angiographic CTO characteristics, tapered stump, absence of calcificaPA, McFadden EP, Pattynama PMT, Serruys PW, Krestin
GP, De Feyter PJ. Multislice spiral CT coronary angiogtions, and occlusion length <15 mm (arrowheads) in a patient who had been reraphy in patients with stable angina pectoris. J Am Coll
ferred for percutaneous recanalization. Volume-rendered MSCT image that provides
Cardiol 2004;43:2265–2270.
a 3-dimensional overview of the coronary arteries, and (B) maximum intensity pro4. Ng W, Chen WH, Lee PY, Lau CP. Initial experience
jection of the same MSCT image show a severely calcified occlusion stump (arrows).
and safety in the treatment of chronic total coronary
(C) Curved multiplanar reconstructed MSCT image of the left anterior descending
occlusions with a new optical coherent reflectometryguided radiofrequency ablation guidewire. Am J Cardiol
coronary artery shows a severely calcified stump (arrow) and an occluded segment
2003;92:732–734.
(arrowheads). Collateral filling is clearly visible distal to the occlusion. This percuta5. Hoye A, Onderwater E, Cummins P, Sianos G, Serruys
neous attempt at recanalization of the CTO was unsuccessful.
P. Improved recanalization of chronic total coronary occlusions using an optical coherence reflectometry-guided
guidewire. Catheter Cardiovasc Interv 2004;63:158 –163.
6. Puma JA, Sketch MH Jr, Tcheng JE, Harrington RA,
measured, results of the MSCT coronary angiography
HR, Stack RS, Califf RM. Percutaneous revascularization of chronic
were “longer,” perhaps reflecting the inaccuracy of quan- Phillips
coronary occlusions: an overview. J Am Coll Cardiol 1995;26:1–11.
titative coronary angiography as previously described. 7. Ferencik M, Moselewski F, Ropers D, Hoffmann U, Baum U, Anders K,
Moreover, MSCT coronary angiography allows evalua- Pomerantsev EV, Abbara S, Brady TJ, Achenbach S. Quantitative parameters of
quality in multidetector spiral computed tomographic coronary imaging
tion of the morphology of the occlusion trajectory, in- image
with submillimeter collimation. Am J Cardiol 2003;92:1257–1262.
cluding detailed delineation of coronary calcification. 8. Morin RL, Gerber TC, McCollough CH. Radiation dose in computed tomogLong occlusions and severe calcifications on MSCT cor- raphy of the heart. Circulation 2003;107:917–922.
9. Trabold T, Buchgeister M, Kuttner A, Heuschmid M, Kopp AF, Schroder S,
onary angiograms were found to be important predictors Claussen
CD. Estimation of radiation exposure in 16-detector row computed tomogof procedural failure, whereas neither feature was iden- raphy of the heart with retrospective ECG-gating. Rofo Fortschr Geb Rontgenstr
tified as an independent predictor on conventional angio- Neuen Bildgeb Verfahr 2003;175:1051–1055.
10. Hunold P, Vogt FM, Schmermund A, Debatin JF, Kerkhoff G, Budde T,
grams.
Erbel R, Ewen K, Barkhausen J. Radiation exposure during cardiac CT:
The need to use contrast material may pose a effective doses at multi-detector row CT and electron-beam CT. Radiology
limitation to preprocedural MSCT coronary angiogra- 2003;226:145–152.
BRIEF REPORTS
243
Angiographic and Clinical Outcomes of
Polytetrafluoroethylene-Covered Stent Use in
Significant Coronary Perforations
Hung Ly,
MD,
Jean-Pierre S. Awaida, MSc, MD, Jacques Lespérance,
Luc Bilodeau, MD
Coronary perforations remain a rare but life-threatening complication of percutaneous coronary intervention. In the setting of complex coronary lesions,
14 of 25 consecutive perforations related to percutaneous coronary intervention were managed with
stents covered with polytetrafluoroethylene (PTFE),
and 11 were managed conventionally with reversal
of anticoagulation and prolonged balloon inflation.
Procedural success was achieved in 71.4% (10 of 14
perforations) of the PTFE arm compared with 27.3%
(3 of 11 perforations) in the standard management
arm (p ⴝ 0.047). Smaller final percent diameter stenosis and postprocedural cessation of dye extravasation were achieved with PTFE-covered stents. No delayed cardiac tamponade or in-hospital mortality was
reported among patients in the PTFE group. 䊚2005
by Excerpta Medica Inc.
(Am J Cardiol 2005;95:244 –246)
ignificant coronary perforations remain a rare but
life-threatening complication during percutaneous
S
coronary intervention (PCI) associated with urgent
bypass surgery in ⬎50% of cases and high mortality
rates.1– 4 Stents covered with polytetrafluoroethylene
(PTFE) have emerged as a new strategy to seal such
perforations.5 This study reports angiographic and
clinical outcomes of PCI-induced coronary perforations after using PTFE-covered stents.
•••
The JOSTENT GraftMaster coronary stent (Abbott
Vascular Devices, Redwood City, California) is a 6Fr
compatible, balloon-expandable, slotted-tube stent
with a layer of porous PTFE sandwiched between 2
coaxial 316-L stainless steel stents. Overall wall thickness reaches 0.30 mm, and the device crossing profile
is 1.6 mm. Diameters from 2.5 to 5.0 mm may be
achieved.
Data related to all consecutive coronary perforations from databases of the catheterization laboratory
of the Montreal Heart Institute (Montreal, Quebec,
Canada) were included in this study. All coronary
procedures were performed according to current standards of practice. All patients were pretreated with
From the Department of Medicine, Division of Interventional Cardiology and the Department of Radiology, Montreal Heart Institute, and the
University of Montreal, Montreal, Quebec, Canada. Dr. Bilodeau’s address is: Cardiac Catheterization Laboratory, Montreal Heart Institute,
5000 Belanger Street, Montreal, Quebec H1T 1C8, Canada. E-mail:
bilodeau@bellnet.ca. Manuscript received April 18, 2004; revised
manuscript received and accepted September 8, 2004.
244
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MD,
and
oral aspirin and intravenous unfractionated heparin.
An activated clotting time of 300 seconds was chosen
for standard PCI, whereas 250 seconds was preferred
if an intravenous glycoprotein IIb/IIIa inhibitor was
used as adjunctive therapy. In the event of a coronary
perforation, immediate reversal of anticoagulants was
performed. Choice of alternatives to seal perforations
was left to the operator’s discretion and included
deployment of a PTFE-covered stent as first-line strategy or prolonged balloon inflation with or without
bare metal stenting. If a PTFE-covered stent was deployed, aspirin and clopidogrel were given for ⱖ3
months. Six-month clinical follow-up was scheduled.
Procedural success was defined as successful delivery and deployment of the PTFE-covered or bare
metal stent to the perforation site, complete cessation
of dye extravasation at the site of rupture at the end of
the procedure (after stenting or after prolonged balloon inflation), and postprocedural Thrombolysis In
Myocardial Infarction grade 3 flow.
Angiographic analysis was performed with a validated automated edge-detection system (CMS 5.2,
MEDIS, Leiden, The Netherlands).6 Lesion morphology was classified according to the modified classification of the American College of Cardiology/American
Heart Association.7 The Thrombolysis In Myocardial
Infarction flow grade was evaluated as previously
described.8 The balloon-to-artery ratio was defined as
the ratio of maximal balloon diameter to the interpolated mean reference vessel diameter of the target
vessel. Coronary perforations were defined as extravasation of dye (blood) from the coronary artery, and
extent of perforation was reported according to the
classification of Ellis et al2: type 1, defined as an
extraluminal crater without extravasation (fully contained perforation); type 2, defined as a pericardial or
myocardial blush without contrast jet extravasation
(limited extravasation); and type 3, defined as brisk
extravasation through a frank perforation with a
⬎1-mm exit hole (Figure 1). Coronary perforations
were considered significant if they were type 2 or 3.
Statistical analysis was performed with STATA
8.0 (STATA Corp., College Station, Texas). Continuous variables are expressed as mean ⫾ SD and were
analyzed with Wilcoxon’s (Mann-Whitney) rank sum
test. Dichotomous variables are reported as frequencies (percentages) and were analyzed with Fisher’s
exact test for proportions.
Between April 1999 and June 2003, 9,382 patients
underwent PCI at the cardiac catheterization laboratory of the Montreal Heart Institute. During that period, 25 consecutive operator-diagnosed coronary per0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.010
FIGURE 1. Mid right coronary artery Ellis 3 type perforation. (A) Before and (B) after PTFE covered stent.
(age, gender, co-morbidities, smoking status, and reasons for PCI) was
noted across groups. Most perforaPTFE Stent Group
tions resulted from balloon or stent
Procedural
Procedural
Bare Metal
oversizing during deployment (11 of
Successes
Failures
Stent Group
14 patients in the PTFE group and 8
(n ⫽ 10)
(n ⫽ 4)
(n ⫽ 11)
of 11 patients in the bare metal
Age (yrs)
70.3 ⫾ 10.5
72.0 ⫾ 16.4
65.7 ⫾ 10.1
group). The mean balloon-to-artery
Men
2 (50)
6 (60)
7 (64)
ratio was 1.37 ⫾ 0.09. Other causes
Diagnosis
Stable angina
2 (20)
0 (0)
4 (36)
of perforations were guidewire damUnstable angina
4 (40)
2 (50)
4 (36)
age to the vessel (2 of 14 in the PTFE
Myocardial infarction
4 (40)
2 (50)
2 (18)
group and 3 of 11 in the bare metal
GP IIa/IIIb inhibitors
5 (50)
1 (25)
4 (36)
group) and rotational atherectomy (1
AHA lesion complexity
A1/B1
1 (10)
2 (50)
0 (0)
of 14 in the PTFE group and 0 of 11
B2
3 (30)
0 (0)
6 (55)
in the bare metal group).
C
6 (60)
2 (50)
5 (45)
Baseline angiographic characterisRVD (mm)
3.13 ⫾ 0.71
2.79 ⫾ 0.38
2.51 ⫾ 0.81
tics were similar across groups (target
MLD (mm)
0.93 ⫾ 0.18
0.88 ⫾ 0.30
0.66 ⫾ 0.17
vessel, bifurcation lesion, eccentricity,
BAR ⬎1.1
7 (70)
1 (25)
6 (55)
Perforation type*
tortuosity, presence of thrombus, le1
1 (10)
0 (0)
0 (0)
sion ulceration, or calcification), and
2
2 (20)
2 (50)
5 (45)
chronic total occlusions characterized
3
7 (70)
2 (50)
6 (55)
a minority of treated lesions (1 in the
Postprocedural TIMI grade 3 flow
10 (100)
3 (75)
8 (73)
Cessation of dye extravasation
0 (0)
1 (25)
6 (55)
PTFE group vs 3 in the bare metal
Postprocedural increase in cardiac
0 (0)
1 (25)
5 (45)
group). Smaller postprocedural mean
enzymes
percent diameter stenoses were
Values are mean ⫾ SD or numbers of patients or lesions (percentages).
achieved in the PTFE group than in the
*Classification according to Ellis et al.2
standard management group (69.1% to
AHA ⫽ American Heart Association; BAR ⫽ balloon-to-artery ratio; GP ⫽ glycoprotein; MLD ⫽
12.7% vs 74.8% to 35.4%, respecminimal luminal diameter; RVD ⫽ reference vessel diameter; TIMI ⫽ Thrombolysis In Myocardial
tively, p ⫽ 0.002).
Infarction.
Procedural success was achieved in
more patients who had been treated
with a PTFE-covered stent than in the
forations were observed (0.3%); all were included in standard treatment group (10 of 14 patients, 71.4%, vs 3
this analysis. Decision to use PTFE-covered stents to of 11 patients, 27.3%, respectively, p ⫽ 0.047; Table 1).
seal ruptures was left to the operator’s discretion. Use Reasons for failure of the PTFE-covered stent included
of the PTFE-covered stent was attempted in a sub- persistence of dye extravasation despite successful stent
group of 14 patients. Selected clinical and angio- deployment in 1 patient and lack of stent delivery to the
perforation site in 3 patients, despite aggressive guiding
graphic data are listed in Table 1.
No difference in baseline clinical characteristics support and use of a buddy-wire technique.
TABLE 1 Clinical and Angiographic Data
BRIEF REPORTS
245
Although comparable rates of final Thrombolysis
In Myocardial Infarction grade 3 flow were achieved,
most notable was the more common persistence of
postprocedural dye extravasation in the standard management arm (Table 1). Cardiac tamponade (immediate or delayed) occurred in 1 patient in the PTFE arm
compared with 4 patients in the standard management
group (p ⫽ 0.133). Emergency referral for cardiac
surgery was required in 1 patient in the PTFE arm and
2 patients in the standard management arm. After
excluding patients who had been referred for cardiac
surgery, postprocedural increases in cardiac enzyme
were documented more frequently in the standard
management arm (5 patients in the bare metal group
vs 1 patient in the PTFE group, p ⫽ 0.056). The only
in-hospital death from perforation (due to hemodynamic collapse) was in the standard management arm.
At 6-month follow-up, 13 of 14 patients in the PTFE
group remained asymptomatic or had mild angina on
exertion (Canadian Cardiovascular Society class 1); 1
patient underwent clinically driven target site revascularization for restenosis. In comparison, 3 of 10
survivors in the standard management group were
readmitted for acute coronary syndromes, whereas the
other patients remained asymptomatic or had mild to
moderate angina on exertion (Canadian Cardiovascular Society class 1 or 2). Of these patients, 2 received
medical therapy and 1 patient was referred for surgical
revascularization for persistent anginal symptoms.
•••
In this historical case-control study, we found that
PTFE-covered stents were significantly better than
bare metal stents at sealing coronary perforations.
They were associated with greater procedural success,
more frequent cessation of dye extravasation after
PCI, larger postprocedural minimal luminal diameter,
and favorable in-hospital and 6-month clinical
outcomes.
Previous management algorithms for coronary perforations have restricted the PTFE-covered stent to a
second-line strategy if conventional options were unsuccessful.9,10 Use of PTFE-covered stents for procedure-related coronary perforations has been shown to
be feasible. Briguori et al5 reported a 100% success
rate in their series of 11 patients who had been treated
with these stents compared with a 40% success rate in
246 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
17 patients who had been treated with bare metal
stents to seal coronary ruptures refractory to prolonged balloon inflation and anticoagulation reversal.
Clinical outcomes showed a significant benefit of covered stents over bare metal stents for all major adverse
cardiac events, except for mortality rate (18% vs 88%,
p ⬍0.001). In-stent restenosis was documented in
29% of patients whose treatment was successful during the 14-month follow-up period. The present study
reports a procedural success rate of 71% with 1 patient
who developed in-stent restenosis.
Although limited by a small sample, the results
from this study show that PTFE-covered stents are a
safe and feasible first-line strategy to seal significant
coronary perforations in the setting of complex coronary lesions PCI.
1. Stankovic G, Orlic D, Corvaja N, Airoldi F, Chieffo A, Spanos V, Montorfano
M, Carlino M, Finci L, Sangiorgi G, Colombo A. Incidence, predictors, inhospital, and late outcomes of coronary artery perforations. Am J Cardiol 2004;
93:213–216.
2. Ellis SG, Ajluni S, Arnold AZ, Popma JJ, Bittl JA, Eigler NL, Cowley MJ,
Raymond RE, Safian RD, Whitlow PL. Increased coronary perforation in the new
device era. Incidence, classification, management, and outcome. Circulation
1994;90:2725–3270.
3. Gunning MG, Williams IL, Jewitt DE, Shah AM, Wainwright RJ, Thomas
MR. Coronary artery perforation during percutaneous intervention: incidence and
outcome. Heart 2002;88:495– 498.
4. Fasseas P, Orford JL, Panetta CJ, Bell MR, Denktas AE, Lennon RJ, Holmes
DR, Berger PB. Incidence, correlates, management, and clinical outcome of
coronary perforation: analysis of 16,298 procedures. Am Heart J 2004;147:140 –
145.
5. Briguori C, Nishida T, Anzuini A, Di Mario C, Grube E, Colombo A.
Emergency polytetrafluoroethylene-covered stent implantation to treat coronary
ruptures. Circulation 2000;102:3028 –3031.
6. Reiber H, Van de Zwet PMJ, Von Land CD, Koning G, Meurs BV, Buis B,
Van Voorthuisen AE. Quantitative coronary arteriography: equipment and technical requirements. In: Reiber JHC, Serruys P, eds. Advances in Quantitative
Coronary Arteriography. Norwell, MA: Kluwer Academic Publishers, 1993:75–
111.
7. Ellis SG, Vandormael MG, Cowley MJ, DiSciascio G, Deligonul U, Topol EJ,
Bulle TM. Coronary morphologic and clinical determinants of procedural outcome with angioplasty for multivessel coronary disease. Implications for patient
selection. Multivessel Angioplasty Prognosis Study Group. Circulation 1990;82:
1193–1202.
8. TIMI Study Group. The Thrombolysis in Myocardial Infarction (TIMI) trial.
Phase I findings. N Engl J Med 1985;312:932–936.
9. Dippel EJ, Kereiakes DJ, Tramuta DA, Broderick TM, Shimshak TM, Roth
EM, Hattemer CR, Runyon JP, Whang DD, Schneider JF, Abbottsmith CW.
Coronary perforation during percutaneous coronary intervention in the era of
abciximab platelet glycoprotein IIb/IIIa blockade: an algorithm for percutaneous
management. Catheter Cardiovasc Interv 2001;52:279 –286.
10. Mulvihill NT, Boccalatte M, Sousa P, Farah B, Laborde JC, Fajadet J, Marco
J. Rapid sealing of coronary perforations using polytetrafluoroethylene-covered
stents. Am J Cardiol 2003;91:343–346.
JANUARY 15, 2005
Role of Postoperative Vasopressor Use in Occurrence
of Atrial Fibrillation After Coronary Artery
Bypass Grafting
Vikrant Salaria, MD, Nirav J. Mehta, MD, Syed Abdul-Aziz,
Syed M. Mohiuddin, MD, DSc, and Ijaz A. Khan, MD
Vasopressor use is common after coronary artery bypass grafting surgery. This study evaluated the role of
postoperative vasopressor use as a predictor of occurrence of atrial fibrillation after coronary artery bypass
grafting and demonstrates that vasopressor use is an
independent predictor of such an occurrence. 䊚2005
by Excerpta Medica Inc.
(Am J Cardiol 2005;95:247–249)
he most effective way of preventing complications
of postoperative atrial fibrillation (AF) is to preT
vent their occurrence. Age, atrial ischemia, intraoperative elements, postoperative fluctuations in autonomic tone, and postoperative sympathetic activation
have been suggested as causative factors of AF after
coronary artery bypass grafting (CABG) surgery.1–12
Sympathomimetic (adrenergic) agents are commonly
used as vasopressors after CABG. Use of these agents
sets a higher catecholamine stage, which, theoretically, may trigger postoperative AF. Practically, it is
not known whether use of vasopressor agents after
CABG is a predictor of an occurrence of postoperative
AF. This study evaluated the role of postoperative
vasopressor use as a predictor of AF after CABG.
•••
The study protocol was approved by the institutional
review board of Creighton University (Omaha, Nebraska). All patients who were ⬎18 years of age and
underwent CABG at Creighton University Medical Center between June 1999 and December 2000 were evaluated for inclusion into the study. Patients who had previous AF and those who were using antiarrhythmic
agents (except ␤ blockers) before or after CABG were
excluded. One hundred thirty-one patients comprised the
study population. Data collection was retrospective.
AF was defined as an irregular heart rate without P
waves on a 12-lead electrocardiogram or continuous
rhythm monitoring. Postoperative AF was defined as
development of AF any time after surgery during the
index hospitalization. The occurrence of postoperative
AF was determined. Data were collected and tabulated
for the following variables: postoperative vasopressor
use, age, gender, hypertension, preoperative echocardioFrom the Division of Cardiology, Creighton University School of
Medicine, Omaha, Nebraska; and the Division of Cardiology,
University of Maryland School of Medicine, Baltimore, Maryland. Dr.
Khan’s address is: Division of Cardiology, University of Maryland
School of Medicine, 22 South Greene Street, S3B06, Baltimore,
Maryland 21201. E-mail: ikhan@medicine.umaryland.edu. Manuscript received May 11, 2004; revised manuscript received and
accepted September 7, 2004.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MD,
graphic parameters (left atrial size, left ventricular ejection fraction, and left ventricular hypertrophy) that were
defined according to standard criteria, cardioplegia use,
aorta cross-clamp time, bypass time, perioperative
␤-blocker use, postoperative cardiac pacing (atrial), postoperative serum levels of magnesium, and length of stay
in an intensive care unit. These variables were examined
for the prediction of postoperative AF.
Continuous variables were expressed as means ⫾
SD and analyzed with Student’s unpaired t test. Categorical variables were expressed as numbers and
percentages and analyzed with chi-square statistics or
Fisher’s exact test, as appropriate. Stepwise logistic
regression analysis was used for multivariate analysis
to determine independent predictors for occurrence of
postoperative AF. A p value ⱕ0.05 was considered
statistically significant. SPSS 7.0 (SPSS Inc., Chicago,
Illinois) was used for statistical analyses.
Among 131 patients, postoperative AF occurred in
30, with an incidence of 23%. Vasopressor agents
were used in 74 patients (56%). The vasopressors used
were dopamine, dobutamine, and phenylephrine. Dopamine was used in 21 patients (16%), dobutamine in
40 (31%), and phenylephrine in 13 (10%). Postoperative AF occurred in 22 of 74 patients (30%) who used
vasopressors and 8 of 57 patients (14%) who did not
use vasopressors (p ⫽ 0.03). Baseline and clinical
characteristics with univariate analyses are presented
in Table 1. By univariate analyses, postoperative vasopressor use (p ⫽ 0.03) and age (p ⬍0.001) were
predictors of developing AF after CABG (Table 1). In
contrast, lack of vasopressor use (p ⫽ 0.02) and cardiac (atrial) pacing (p ⬍0.001) were predictors of not
developing AF after CABG. Among 131 patients,
CABG was performed off-pump in 43 patients. There
was no significant difference in the occurrence of AF
with off-pump versus on-pump CABG (10 of 43 patients, 23% vs 20 of 88 patients, 23%, p ⫽ NS).
According to stepwise logistic regression analysis
(Table 2), vasopressor use was an independent predictor of developing postoperative AF (odds ratio 2.783,
95% confidence interval 1.073 to 7.220, p ⫽ 0.03).
Age was the only other independent predictor (odds
ratio 1.085, 95% confidence interval 1.038 to 1.135, p
⬍0.001). Among various vasopressor drugs, use of
those with predominantly ␤-1–adrenergic affinity was
associated with a higher incidence of postoperative
AF (dopamine 43% and dobutamine 30% vs phenylephrine 8%, p ⫽ 0.05; Table 3).
•••
Our study demonstrates that use of vasopressors is
an independent predictor of development of AF after
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.011
247
TABLE 1 Atrial Fibrillation After Coronary Artery Bypass Grafting
Variables
Postoperative vasopressor use
Age (yrs)
Men/women
Hypertension
Left ventricular hypertrophy
Left ventricular ejection fraction (%)
Left atrial diameter (mm)
Off-pump surgery
On-pump surgery
Cardioplegia use
Aorta cross-clamp time (min)
Bypass time (min)
Perioperative ␤-blocker use
Postoperative cardiac (atrial) pacing
Postoperative cardiac index (L/m2)
Postoperative serum magnesium (mEq/L)
Hours in intensive care unit
AF Present
(n ⫽ 30)
AF Absent
(n ⫽ 101)
p Value
22 (73%)
74.6 ⫾ 9.3
19 (63%)/11 (37%)
26 (87%)
15 (50%)
47.0 ⫾ 16.7
42.7 ⫾ 5.8
10 (33%)
20 (67%)
29 (97%)
74.9 ⫾ 65.0
133.3 ⫾ 92.4
27 (90%)
21 (70%)
2.67 ⫾ 0.66
1.95 ⫾ 0.27
130.5 ⫾ 107.4
52 (52%)
64.8 ⫾ 12.2
74 (73%)/27 (27%)
87 (86%)
37 (37%)
46.3 ⫾ 13.5
41.8 ⫾ 5.4
33 (33%)
68 (67%)
97 (96%)
66.7 ⫾ 26.4
110.4 ⫾ 44.8
96 (95%)
100 (100%)
2.60 ⫾ 0.62
1.89 ⫾ 0.28
97.3 ⫾ 80.4
0.02
⬍0.001
0.29
0.94
0.19
0.82
0.55
0.88
0.88
0.87
0.59
0.29
0.31
⬍0.001
0.59
0.31
0.07
Values are numbers of patients (percentages) or mean ⫾ SD.
operative trauma and from being in
an intensive care unit, which are potent stimulators of the sympathetic
Variables
B
SE of B
OR (95% CI)
p Value*
nervous system. That the postoperative state is hyperadrenergic has been
Vasopressor use
1.024
0.486
2.783 (1.073–7.220)
0.03
proved by marked postoperative inAge (yrs)
0.082
0.023
1.085 (1.038–1.135)
⬍0.001
Constant
⫺7.578
1.721
⬍0.001
creases in norepinephrine levels in
mixed venous samples after sur*Overall model significance, p ⬍0.001.
gery,11,14 because the predominant
CI ⫽ confidence interval; OR ⫽ odds ratio.
source of circulating norepinephrine
in humans is from sympathetic nerve
endings.15 A significant association
has also been shown between postoperative mixed
TABLE 3 Incidence of Atrial Fibrillation by Vasopressor Agent
venous norepinephrine levels after cardiac surgery and
Incidence of
Agents Used
development of postoperative AF, suggesting that
Vasopressor
AF by Agent
Within AF Group
sympathetic activation may be an important factor in
most such patients in whom postoperative AF develNone
8/57 (14%)
8/30 (27%)
Dopamine
9/21 (43%)*
9/30 (30%)
ops.11 In another study that analyzed heart rate variDobutamine
12/40 (30%)*
12/30 (40%)
ability, a shift in autonomic balance that resulted in
Phenylepherine
1/13 (8%)*
1/30 (3%)
enhanced sympathetic tone and a loss of vagal tone
*p ⫽ 0.05 for dopamine or dobutamine versus phenylepherine.
was demonstrated before the onset of postoperative
AF.12 In brief, the postoperative hyperadrenergic state
is a strong factor responsible for the precipitation of
CABG. Moreover, the incidence of postoperative AF AF after surgery because AF usually resolves spontawas higher with the use of dopamine and dobutamine, neously once the provocative hyperadrenergic state is
which are predominantly ␤-1–adrenergic receptor no longer present. Use of adrenergic vasopressors
agonists, than with the use of phenylephrine, which in
would enhance the hyperadrenergic postoperative
therapeutic doses is predominantly an agonist of ␣-adrenergic receptors and lacks direct stimulatory effects state after CABG and result in a greater chance of
on atrial myocardium. This is the first study to eval- developing AF.
Age has been consistently shown as an indepenuate the role of vasopressor use in the development of
dent
predictor of postoperative AF,1–5 and the present
AF after CABG. These findings indicate that vasostudy
substantiates this finding. Age-related structural
pressor use after CABG should be limited to a real
and
hemodynamic
changes in atrial dilation and atrial
necessity.
substrate
and
higher
atrial pressures due to age-related
Excessive use of catecholamines enhances trigslow
ventricular
relaxation
might be plausible mechgered activity and automaticity, the triggers for initiation of AF.13 Further, catecholamines shorten atrial anisms of higher incidences of postoperative AF in
refractoriness, but in a nonhomogenous pattern, thus older patients.16,17 Further, advancing age has been
favoring continuation of AF.13 A postoperative state is associated with increased sympathetic outflow and
clearly a hyperadrenergic condition that results from increased levels of circulating norepinephrine.18
TABLE 2 Independent Predictors of Atrial Fibrillation After Coronary Artery Bypass
Grafting
248 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
1. Mathew JP, Parks R, Savino JS, Friedman AS, Koch C, Mangano DT, Browner
10. Abreu JE, Reilly J, Salzano RP, Khachane VB, Jekel JF, Clyne CA. Comparison
WS. Atrial fibrillation following coronary artery bypass graft surgery: predictors,
outcomes, and resource utilization. JAMA 1996;276:300 –306.
2. Funk M, Richards SB, Desjardins J, Bebon C, Wilcox H. Incidence, timing,
symptoms, and risk factors for atrial fibrillation after cardiac surgery. Am J Crit
Care 2003;12:424 – 433.
3. Aranki SF, Shaw DP, Adams DH, Rizzo RJ, Couper GS, VanderVliet M,
Collins JJ, Cohn LH, Burstin HR. Predictors of atrial fibrillation after coronary
artery surgery: current trends and impact on hospital resources. Circulation
1996;94:390 –397.
4. Mathew JP, Fontes ML, Tudor IC, Ramsay J, Duke P, Mazer CD, Barash PG,
Hsu PH, Mangano DT. A multicenter risk index for atrial fibrillation after cardiac
surgery. JAMA 2004;291:1720 –1729.
5. Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993;56:539 –549.
6. Taylor GJ, Malik SA, Colliver JA, Dove JT, Moses HW, Mikell FL, Batchelder JE, Schneider JA, Wellons HA. Usefulness of atrial fibrillation as a
predictor of stroke after isolated coronary bypass grafting. Am J Cardiol 1987;
60:905–907.
7. Villareal RP, Hariharan R, Liu BC, Kar B, Lee VV, Elayda M, Lopez JA,
Rasekh A, Wilson JM, Massumi A. Postoperative atrial fibrillation and mortality
after coronary artery bypass surgery. J Am Coll Cardiol 2004;43:742–748.
8. Crosby LH, Pifalo WB, Woll KR, Burkholder JA. Risk factors for atrial fibrillation
after coronary artery bypass grafting. Am J Cardiol 1990;66:1520 –1522.
9. Kolvekar S, D’Souza A, Akhatar P, Reek C, Garratt C, Spyt T. Role of atrial
ischemia in development of atrial fibrillation following coronary artery bypass
surgery. Eur J Cardiothorac Surg 1997;11:70 –75.
of frequencies of atrial fibrillation after coronary artery bypass grafting with and
without use of cardiopulmonary bypass. Am J Cardiol 1999;83:775–776.
11. Kalman JM, Munawar M, Howes LG, Louis WJ, Buxton BF, Gutteridge G,
Tonkin AM. Atrial fibrillation after coronary artery bypass grafting is associated
with sympathetic activation. Ann Thorac Surg 1995;60:1709 –1715.
12. Dimmer C, Tavernier R, Gjorgov N, Nooten GV, Clement DL, Jordaens L.
Variation of autonomic tone preceeding onset of atrial fibrillation after coronary
artery bypass grafting. Am J Cardiol 1998;82:22–25.
13. Waldo AL. Mechanisms of atrial fibrillation, atrial flutter, and ectopic atrial
tachycardia. Circulation 1987;75(suppl 3):37– 40.
14. de Leeuw PW, van der Starre PJ, Harinck-de Weerdt JE, de Bos R, Tchang
PT, Birkenhager WH. Humoral changes during and following coronary bypass surgery: relationship to postoperative blood pressure. J Hypertens 1983;
1:52–54.
15. Goldstein DS, McCarty R, Polinsky RJ, Kopin IJ. Relationship between
plasma norepinephrine and sympathetic neural activity. Hypertension 1983;5:
552–559.
16. Manyari DE, Patterson C, Johnson D, Melendez L, Kostuk WJ, Cape RD.
Atrial and ventricular arrhythmias in asymptomatic active elderly subjects: correlation with left atrial size and left ventricular mass. Am Heart J 1990;119:
1069 –1076.
17. Aronow WS. The older man’s heart and heart disease. Med Clin North Am
1999;83:1291–1303.
18. Hoeldtke RD, Cilmi KM. Effects of aging on catecholamine metabolism.
J Clin Endocrinol Metab 1985;60:479 – 484.
Reaching Recommended Lipid and Blood Pressure
Targets With Amlodipine/Atorvastatin Combination in
Patients With Coronary Heart Disease
Jean-Francois Dorval, MD, Todd Anderson, MD, Jean Buithieu, MD, Sammy Chan, MD,
Stuart Hutchison, MD, Thao Huynh, MD, Jean Jobin, PhD, Eva Lonn, MD,
Paul Poirier, MD, Lawrence Title, MD, Ann Walling, MD, Thang Tran, MSc,
Ghyslain Boudreau, PhD, Francois Charbonneau, MD, and Jacques Genest, MD
The effects of combined atorvastatin and amlodipine
on blood pressure (BP) and low-density lipoprotein
(LDL) cholesterol levels were investigated in 134 patients with documented coronary heart disease
treated for 1 year. BP at baseline was 128 ⴞ 15/79
ⴞ 9 mm Hg and was controlled by the treating physician; no calcium channel blockers were allowed.
Baseline means for plasma cholesterol were 6.4 ⴞ
1.1 mmol/L (147 ⴞ 39 mg/dl), triglycerides 2.0 ⴞ
0.9 mmol/L (177 ⴞ 88 mg/dl), LDL cholesterol 4.4 ⴞ
1.0 mmol/L (170 ⴞ 39 mg/dl), and high-density
lipoprotein cholesterol 1.2 ⴞ 0.3 mmol/L (46 ⴞ 12
From McGill University Health Center, Montreal, Québec; Department
of Cardiovascular Sciences and the Libin Cardiovascular Institute of
Alberta, Calgary, Alberta; Vancouver General Hospital, Vancouver,
British Columbia; Toronto Hospital Network, Toronto, Ontario; Université Laval, Institut de Cardiologie de Québec, Québec; Department of
Medicine, Division of Cardiology, McMaster University, Hamilton,
Ontario; Queen Elizabeth II Medical Center, Halifax, Nova Scotia;
Jewish General Hospital, Montreal, Quebec; Pfizer Canada,
Montreal, Quebec; and Foothills Medical Centre, Calgary, Alberta,
Canada. This study was supported by an unrestricted grant from Pfizer.
Dr. Genest’s address is: Division of Cardiology, McGill University Health
Center/Royal Victoria Hospital, 687 Pine Avenue West, Montréal,
Quebec, H3A 1A1, Canada. E-mail: Jacques.genest@muhc.mcgill.ca.
Manuscript received July 7, 2004; revised manuscript received and
accepted August 31, 2004.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
mg/dl). Patients were all given atorvastatin 10 mg,
then increased to 80 mg if the LDL cholesterol was
<2.5 mmol/L (100 mg/dl). At 3 months, patients
were randomized to amlodipine 10 mg or placebo.
Plasma LDL cholesterol was decreased by 50%, and
the LDL cholesterol target of <2.5 mmol/L was
achieved in 81% of the patients. BP targets were
achieved in 69% of the atorvastatin ⴙ placebo group,
versus 96% in the atorvastatin ⴙ amlodipine group (p
ⴝ 0.0002). With use of combination atorvastatin ⴙ
amlodipine at doses ranging from 10 to 80 mg and 5
to 10 mg, respectively, recommended therapeutic
goals were reached in most select subjects with coronary artery disease who were concomitantly receiving aspirin and antihypertensive therapy. 䊚2005
by Excerpta Medica Inc.
(Am J Cardiol 2005;95:249 –253)
n the present study, we investigated the effects of
atorvastatin and amlodipine in subjects with estabIlished
and stable coronary heart disease (CHD) requiring a statin to decrease low-density lipoprotein (LDL)
cholesterol and who had adequate blood pressure control at baseline, as judged by the treating physician.
We used a conditional titration with atorvastatin 10 or
80 mg to reach a LDL target value of ⱕ2.5 mmol/L,
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.012
249
atorvastatin was increased to 80 mg/
day. At 3 months, patients were ranAtorvastatin ⫹
Atorvastatin ⫹
domized to amlodipine 10 mg/day or
Amlodipine
Placebo
Overall Group
placebo, and atorvastatin and amloVariable
(n ⫽ 64)
(n ⫽ 70)
(n ⫽ 134)
dipine/placebo therapy was continAge (yrs)
58 ⫾ 8
56 ⫾ 9
57 ⫾ 9
ued for 9 months. The dose of amloMen/women
55/9
60/10
115/19
dipine could be downtitrated at any
28.1 ⫾ 3.1
28.9 ⫾ 3.8
28.5 ⫾ 3.5
Body mass index
time to 5 mg/day if patients had
2
(kg/m )
symptoms related to hypotension.
Diabetes mellitus
2 (3.1%)
10 (14.3%)
12 (9.0%)
Smoking (current
81%
84%
83%
After 12 months, all study-driven
⫹ ex)
treatment was discontinued and paBP (mm Hg)
127 ⫾ 15/79 ⫾ 9
128 ⫾ 14/79 ⫾ 8
128 ⫾ 15/79 ⫾ 9
tients were followed for an addiData are expressed as mean ⫾ SD or number (%).
tional 2 months.
Plasma lipid and lipoprotein lipid
measurements were determined at
each institution using standardized
in keeping with current practice guidelines.1– 4 Amlo- techniques. LDL cholesterol was calculated according
dipine 10 mg was used for blood pressure (BP) control to the Friedewald formula. Apolipoprotein B was deaccording to current recommendations5–7 and was termined by nephelometry. BP was taken in triplicate,
downtitrated to 5 mg if symptoms of hypotension with the subject sitting for ⱖ5 minutes, in accordance
developed. Previous reports have indicated a low with the Canadian Hypertension Education Program
prevalence of reaching recommended targets for CHD guidelines.6
prevention in high-risk subjects8 –12; recent trials point
All continuous variables were analyzed at basetoward more aggressive LDL cholesterol targets in line and at months 3 and 12 for the intent-to-treat
very high-risk subjects.13
and per-protocol populations. Analysis of covariance methods with effects for treatment, adjusted
•••
Study subjects were selected in 11 centers across for center and baseline values, were performed. To
Canada. The study was designed to assess the effects evaluate the effect before and after the treatment, a
of atorvastatin and the combination of atorvastatin and paired Student’s t test was used within treatment
amlodipine on vascular endothelial function measured comparisons. Continuous data are reported as mean
by flow-mediated vasodilation. Inclusion criteria were ⫾ SEM when using analysis of covariance and
men and women aged ⬎30 years with documented mean ⫾ SD for raw data. For categorical variables,
CHD (determined by coronary angiography ⱖ70% which included the target rate for BP and LDL, the
cross-sectional stenosis), including patients after cor- Cochran-Mantel-Haenzsel chi-square test, stratified
onary revascularization, by nuclear or stress echocar- according to center, was performed. All statistical
diography or by a history of documented myocardial tests were 2-sided, and a p value ⬍0.05 was coninfarction. For inclusion into the study, study subjects sidered statistically significant. No adjustments for
had to be in stable condition for ⱖ3 months, without multiple comparisons were made.
One hundred thirty-four subjects were randomized
changes in current therapy, except those allowed
within the protocol. The lipid entry criteria were a into the study (115 men and 19 women; mean age
LDL cholesterol ⬎3.2 mmol/L (123 mg/dl), triglyc- 57 ⫾ 9 years). Baseline characteristics of all study
erides ⬍4.5 mmol/L (400 mg/dl), and no lipid-lower- subjects are listed in Table 1. As previously stated, at
ing medications for 2 months. No patient was allowed study entry (baseline) all patients received atorvastatin
to take calcium blocking drugs, hormone replacement 10 mg, which was uptitrated to 80 mg/day if the LDL
therapy, or statins at entry. Antihypertensive medica- cholesterol was ⬎2.5 mmol/L (100 mg/dl). Randomtion, antiplatelet agents, and antianginal agents were ization to amlodipine 10 mg or placebo was done at
required to be stable for ⱖ3 months before entry into month 3 and BP was taken shortly thereafter. No
the study. Choice of BP therapy was left to the dis- significant differences in baseline characteristics were
cretion of the treating physician as long as protocol- observed in subjects subsequently randomized to amdriven restrictions were respected. Patients were ex- lodipine or placebo. The total duration of the study
cluded if they had a coronary event, revascularization was 12 months; 56 of 64 subjects in the atorvastatin ⫹
procedure, or a cerebrovascular accident in the pre- amlodipine group and 65 of 70 in the atorvastatin ⫹
ceding 3 months. Diabetic subjects had to be well placebo group completed the study.
At the baseline evaluation, 83% of patients reported a
controlled for ⱖ3 months, as reflected by a hemoglobin A1c level ⱕ10% of total hemoglobin. No patient history of cigarette smoking, although only 21% were
current smokers; the rate of smokers at 12 months was
had renal or hepatic disease.
All patients were asked to discontinue lipid-modi- 22%. Sixty-two percent of subjects were ex-smokers and
fying agents and calcium blockers for 2 months, after only 16% never smoked. Mean body mass index (weight
which time baseline lipid and lipoprotein cholesterol [kilograms]/height[square meters]) was 28.5 ⫾ 3.5 at
levels were obtained and all patients were started on baseline and 28.7 ⫾ 3.6 at 12 months. The number of
atorvastatin 10 mg/day. If the LDL cholesterol was subjects with a body mass index ⬍25 kg/m2 was 14%
⬎2.5 mmol/L (100 mg/dl) after 6 weeks, the dose of and at 12 months 14%, with no differences between
TABLE 1 Baseline Characteristics of Study Population
250 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
TABLE 2 Lipid, Lipoprotein Lipids, and Apolipoprotein B
Levels at Baseline, at Three and 12 Months on Study
Medication, and After a Two-month Washout
Lipids
Cholesterol
mmol/l
mg/dl
Triglycerides
mmol/l
mg/dl
LDL cholesterol
mmol/l
mg/dl
HDL cholesterol
mmol/l
mg/dl
Cholesterol HDL
cholesterol
mmol/l
mg/dl
Non-HDL cholesterol
mmol/l
mg/dl
Apolipoprotein B
mg/dl
Baseline
3 Mo
12 Mo
6.4 ⫾ 1.1
147 ⫾ 49
4.0 ⫾ 0.6*
155 ⫾ 23
3.9 ⫾ 0.6*
150 ⫾ 23
2.0 ⫾ 0.9
177 ⫾ 80
1.4 ⫾ 0.3*
124 ⫾ 27
1.4 ⫾ 0.7*
124 ⫾ 62
4.4 ⫾ 1.0
170 ⫾ 36
2.2 ⫾ 0.5*
85 ⫾ 19
2.2 ⫾ 0.5*
85 ⫾ 19
1.2 ⫾ 0.3
46 ⫾ 11
1.2 ⫾ 0.3
46 ⫾ 11
1.2 ⫾ 0.3
46 ⫾ 11
5.7 ⫾ 1.5
220 ⫾ 58
3.6 ⫾ 0.8
139 ⫾ 31
3.6 ⫾ 0.9*
139 ⫾ 35
5.2 ⫾ 1.0
202 ⫾ 39
1.4 ⫾ 0.3
2.9 ⫾ 0.6*
112 ⫾ 23
0.80 ⫾ 0.1*
2.8 ⫾ 0.5*
108 ⫾ 19
0.8 ⫾ 0.1*
*p ⬍0.0001 versus baseline. Overall, there were no significant differences
between amlodipine and placebo groups.
FIGURE 1. Effect of atorvastatin on plasma LDL cholesterol (C)
levels.
groups. Body mass index remained stable throughout the
study.
At 12 months, in the atorvastatin ⫹ placebo group,
26% of patients (17 of 65) were taking atorvastatin 10
mg and 74% (48 of 65) were taking 80 mg. In the
atorvastatin ⫹ amlodipine group, 21% (12 of 56) were
taking atorvastatin 10 mg and the remaining subjects
(79% [44 of 56]) were taking 80 mg. In the atorvastatin ⫹ amlodipine group, 91% of patients continued
to take 10 mg of amlodipine throughout the study,
whereas only 9% (5 of 56) required downtitration
to 5 mg.
Baseline total cholesterol was 6.4 ⫾ 1.0 mmol/L
(147 ⫾ 39 mg/dl) in the atorvastatin ⫹ amlodipine
and 6.4 ⫾ 1.2 mmol/L (148 ⫾ 45 mg/dl) in the
FIGURE 2. Effects of amlodipine (Amlo) and placebo (Plac) on BP.
Ator ⴝ atrovastatin.
atorvastatin ⫹ placebo groups. An overall decrease in
total cholesterol by 2.4 ⫾ 0.1 mmol/L was observed in
the combined groups (p ⬍0.0001). A similar overall
decrease in LDL cholesterol was seen, from 4.4 ⫾ 1.0
mmol/L (170 ⫾ 39 mg/dl) to 2.2 ⫾ 0.5 mmol/L (85 ⫾
19 mg/dl) (Table 2 and Figure 1). There was a trend
toward a slightly lower LDL cholesterol in the atorvastatin ⫹ placebo compared with the atorvastatin ⫹
amlodipine groups (2.0 ⫾ 0.1 vs 2.2 ⫾ 0.1 mmol/L, p
⫽ 0.005) at 12 months; this represented an absolute
decrease in LDL cholesterol of 2.2 ⫾ 0.1 and 2.0 ⫾
0.1 mmol/L in each group, respectively, representing a
50% decrease in LDL cholesterol. Plasma triglycerides decreased from 2.0 ⫾ 1.0 mmol/L (177 ⫾ 88
mg/dl) in both groups to 1.4 ⫾ 0.7 mmol/L (117 ⫾ 56
mg/dl), with no significant differences between both
groups. High-density lipoprotein (HDL) cholesterol
levels did not change with atorvastatin treatment (1.2
⫾ 0.3 at baseline vs 1.2 ⫾ 0.3 mmol/L at 12 months).
Data at 3 months was virtually identical to 12-month
values for cholesterol, triglycerides, and LDL/HDL
cholesterol. Overall apolipoprotein B levels decreased
from a baseline level of 1.35 ⫾ 0.25 to 0.76 ⫾ 0.14
mg/dl, a 43% decrease (p ⬍0.0001), with no significant difference between treatment groups. The overall
cholesterol/HDL cholesterol ratio decreased from 5.6
⫾ 1.4 to 3.6 ⫾ 1.0 mg/dl (p ⬍0.0001; Table 2)., with
no significant difference between treatment groups.
The major LDL cholesterol target of ⬍2.5 mmol/L
(100 mg/dl) was achieved overall in 81% of study
subjects (97 of 120). Interestingly, there were slightly
fewer patients reaching an LDL cholesterol of ⬍2.5
mmol/L in the atorvastatin ⫹ amlodipine group (39 of
56 [71%], 95% confidence intervals 57% to 82%) than
in the atorvastatin ⫹ placebo group (58 of 65 [89%],
95% confidence intervals 79% to 96%) (p ⫽ 0.01).
This may be related to the slight increase in patients
taking the 80-mg dose of atorvastatin in the atorvastatin ⫹ placebo group (78% vs 74%) and to the absolute
0.3 mmol (11 mg/dl) difference among LDL cholesterol levels between the atorvastatin ⫹ placebo group
and the atorvastatin ⫹ amlodipine group (2.3 ⫾ 0.5 vs
BRIEF REPORTS
251
TABLE 3 Blood Pressure at Baseline, Three Months, 12
Months, and After a Two-month Washout
Time
Baseline (mm Hg)
3 mo (mm Hg)
12 mo (mm Hg)
Atorvastatin ⫹
Amlodipine
Atorvastatin ⫹
Placebo
127 ⫾ 15/79 ⫾ 9
117 ⫾ 11/74 ⫾ 9*†
118 ⫾ 10/76 ⫾ 7*†
128 ⫾ 14/79 ⫾ 8
127 ⫾ 16/79 ⫾ 7
130 ⫾ 17/80 ⫾ 9
*p ⬍0.0001 versus baseline; †p ⬍0.0001 versus atorvastatin ⫹ placebo.
FIGURE 3. Percentage of patients reaching LDL cholesterol (C)
targets (<2.5 mmol/L) taking atorvastatin 10 or 80 mg, and
reaching BP targets taking amlodipine 5 or 10 mg or placebo.
Abbreviations as in Figure 2.
2.0 ⫾ 0.4 mmol/L, p ⫽ 0.005). The percentage of
subjects reaching a cholesterol/HDL cholesterol ratio
⬍4.0 was 11.3% in the atorvastatin ⫹ amlodipine
group versus 2.9% in the atorvastatin ⫹ placebo group
at baseline (p ⫽ 0.066). At month 12, these rates were
70.9% and 73.8%, respectively. Overall, 72.5% of all
study participants reached a HDL cholesterol target of
⬍4.0 compared with 6.8% at baseline. At baseline,
non-HDL cholesterol was 5.17 ⫾ 0.84 mmol/L
(200 ⫾ 32 mg/dl) in the atorvastatin ⫹ amlodipine
group and 5.30 ⫾ 1.13 mmol/L (205 ⫾ 44 mg/dl) in
the atorvastatin ⫹ placebo subjects (p ⫽ NS). At 12
months, non-HDL cholesterol was 2.9 ⫾ 0.1 mmol/L
(109 ⫾ 27 mg/dl) and 2.7 ⫾ 0.1 mmol/L (102 ⫾ 27
mg/dl), respectively (p ⫽ 0.048).
Mean BP was 128 ⫾ 15/79 ⫾ 9 mm Hg at baseline
and was similar in both groups: 127 ⫾ 15/79 ⫾ 9 mm
Hg in the atorvastatin ⫹ amlodipine group and 128 ⫾
14/79 ⫾ 8 mm Hg in the atorvastatin ⫹ placebo
group. Diuretics were administered to 12% of patients,
␤-adrenergic blockers were given to 75%, angiotensin-converting enzyme inhibitors to 54%, and angiotensin receptors blockers to 5%. Other antihypertensives (␣-adrenoreceptor blockers) were administered
to 1 patient and accounted for ⬍1% of prescriptions
(calcium blockers were not allowed at the time of
randomization). At the end of the study, BP was 118
⫾ 10/76 ⫾ 7 mm Hg in the atorvastatin ⫹ amlodipine
group (an absolute decrease of 8 ⫾ 2/3 ⫾ 1 mm Hg
compared with baseline values, p ⬍0.0001), and was
252 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
130 ⫾ 17/80 ⫾ 9 mm Hg in the atorvastatin ⫹
placebo group. The differences in systolic and diastolic pressures at 12 months between the atorvastatin
⫹ amlodipine and the atorvastatin ⫹ placebo groups
was significant at the p ⬍0.0001 and 0.0006 levels,
respectively (Figure 2). We used the treatment goals
of the Canadian Hypertension Education Program in
the analysis: at baseline, 75% of the patients were at
BP target in the atorvastatin ⫹ amlodipine group
compared with 70% in the atorvastatin ⫹ placebo
group (p ⫽ 0.503). At 12 months, 96% (54 of 56) of
subjects in the atorvastatin ⫹ amlodipine group were
at or below Canadian Hypertension Education Program specified BP targets, whereas 69% (45 of 65)
were at or below target in the atorvastatin ⫹ placebo
group (p ⫽ 0.0002) (Table 3 and Figure 3).
The medications were well tolerated. Myalgias
were reported in 6 of 70 patients (8.6%) in the atorvastatin ⫹ amlodipine group and in 1 of 74 patients
(1.4%) in the atorvastatin ⫹ placebo group. Elevations
in creatine kinase (⬎3 times the upper limit of normal)
occurred in 3 of 70 (4.3%) and in 2 of 74 (2.7%)
subjects. There were 4 of 70 cases (4.3%) of symptomatic hypotension in the atorvastatin ⫹ amlodipine
group, resulting in a decrease in the amlodipine dosage and in 1 subject (1.4%) in the atorvastatin ⫹
placebo group. Angina occurred in 5 of 70 subjects
(7.1%) in the atorvastatin ⫹ amlodipine group and in
9 of 74 (12.3%) in the atorvastatin ⫹ placebo group.
One patient in each group developed acute myocardial
infarction.
•••
Despite a relatively normal BP at entry (127 ⫾
15/79 ⫾ 9 mm Hg), 25% of patients in the atorvastatin
⫹ amlodipine group and 30% of patients in the atorvastatin ⫹ placebo groups were not at target at baseline (p ⫽ 0.503 between-group assignment). With the
addition of amlodipine, 96% of patients were at target
at 3 and 12 months compared with 69% in the placebo
group (p ⬍0.0001). Target LDL cholesterol levels
have remained a source of controversy; an arbitrary
level of ⬍2.5 mmol/L (100 mg/dl),1,3 a single dose of
statin,14 or a 45% to 45% reduction in LDL cholesterol15 are all evidenced based. The use of other cardioprotective medications are also part of evidencebased medicine in high-risk prevention strategies.16 –19
The mean LDL cholesterol achieved at 1 year is less
than the recommended level, yet is in keeping with
novel studies13,20 that address the issue of aiming for
a LDL cholesterol value lower than recommended. In
our study, 81% of all patients reached a LDL cholesterol of ⬍2.5 mmol/L. The rates of LDL cholesterol
target at 3 and 12 months were 79% and 71%, respectively, in the atorvastatin ⫹ amlodipine group and
87%, and 89% in the atorvastatin ⫹ placebo group.
The data presented here suggest that intensive
patient follow-up can lead to very high rates of
reaching multiple recommended targets for the prevention of cardiovascular disease in high-risk patients (Figure 3). We used a combination of atorvastatin 10 or 80 mg and amlodipine 5 or 10 mg.
Therapy with atorvastatin 10 mg ⫹ amlodipine 10
JANUARY 15, 2005
mg was administered to 26% of patients, atorvastatin 80 mg ⫹ amlodipine 5 mg to 9%, and atorvastatin 80 mg ⫹ amlodipine 10 mg to 70%. The
combination of the 2 medications was well tolerated. There were no unexpected adverse effects
reported in either treatment group. A hard cardiovascular end point (acute myocardial infarction)
occurred in 1 subject in each treatment arm (2 of
134 [1.5%]) during the 12-month treatment period.
Although not designed to assess cardiovascular end
points, the low event rate is a reflection of baseline
treatment according to current recommendations for
secondary prevention.4
1. Expert Panel on Detection, Evaluation, and Treatment of High Blood Choles-
terol in Adults. Executive Summary of The Third Report of The National
Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation,
and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III).
JAMA 2001;285:2486 –2497.
2. Genest J, Frohlich J, Fodor G, McPherson R, and the Working Group on
Hypercholesterolemia and Other Dyslipidemias. Recommendations for the management of dyslipidemia and the prevention of cardiovascular disease: summary
of the 2003 update. Can Med Assoc J 2003;169:921–924.
3. De Backer G, Ambrosioni E, Borch-Johnsen K, Brotons C, Cifkova R,
Dallongeville J, Ebrahim S, Faergeman O, Graham I, Mancia G, et al, and the
European Society of Cardiology Committee for Practice Guidelines. European
guidelines on cardiovascular disease prevention in clinical practice: third joint
task force of European and other societies on cardiovascular disease prevention
in clinical practice (constituted by representatives of eight societies and by invited
experts). Eur J Cardiovasc Prev Rehabil 2003;10(suppl):S1–S10.
4. Genest J, Pedersen TR. Prevention of cardiovascular ischemic events: high-risk
and secondary prevention. Circulation 2003;107:2059 –2065.
5. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr,
Jones DW, Materson BJ, Oparil S, Wright JT Jr, Roccella EJ, and the National
Heart, Lung, and Blood Institute Joint National Committee on Prevention,
Detection, Evaluation, and Treatment of High Blood Pressure, and the National High Blood Pressure Education Program Coordinating Committee. The
Seventh Report of the Joint National Committee on Prevention, Detection,
Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA
2003;289:2560 –2572.
6. Hemmelgarn BR, Zarnke KB, Campbell NR, Feldman RD, McKay DW,
McAlister FA, Khan N, Schiffrin EL, Myers MG, Bolli P, et al, and the Canadian
Hypertension Education Program, Evidence-Based Recommendations Task
Force. The 2004 Canadian Hypertension Education Program recommendations
for the management of hypertension: part I– blood pressure measurement, diagnosis and assessment of risk. Can J Cardiol 2004;20:31– 40.
7. European Society of Hypertension-European Society of Cardiology Guideline
Committee 2003. European Society of Hypertension—European Society of Car-
diology guidelines for the management of arterial hypertension. J Hypertens
2003;21:1011–1053.
8. Wolf-Maier K, Cooper RS, Kramer H, Banegas JR, Giampaoli S, Joffres MR,
Poulter N, Primatests P, Stegmayr B, Thamm M. Hypertension treatment and
control in five European Countries, Canada and the Unites States. Hypertension
2004;43:10 –17.
9. Touyz RM, Campbell N, Logan A, Gledhill N, Petrella R, Padwal R. Canadian
Hypertension Education Program. The 2004 Canadian recommendations for the
management of hypertension: part III—lifestyle modifications to prevent and
control hypertension. Can J Cardiol 2004;20:55–59.
10. EUROASPIRE I and II Group. European Action on Secondary Prevention by
Intervention to Reduce Events. Clinical reality of coronary prevention guidelines:
a comparison of EUROASPIRE I and II in nine countries. EUROASPIRE I and
II Group. European Action on Secondary Prevention by Intervention to Reduce
Events. Lancet 2001;357:972–973.
11. Vanuzzo D, Pilotto L, Ambrosio GB, Pyorala K, Lehto S, De Bacquer D,
De Backer G, Wood D, and the EUROASPIRE Study Group. Potential for
cholesterol lowering in secondary prevention of coronary heart disease in
Europe: findings from EUROASPIRE study. European Action on Secondary
Prevention through Intervention to Reduce Events. Atherosclerosis 2000;153:
505–517.
12. Pearson TA, Laurora I, Chu H, Kafonek S. The lipid treatment assessment
project (L-TAP): a multicenter survey to evaluate the percentages of dyslipidemic
patients receiving lipid-lowering therapy and achieving low-density lipoprotein
cholesterol goals. Arch Intern Med 2000;160:459 – 467.
13. Grundy SM, Cleeman JL, Merz CN, Brewer HB Jr., Clark T, Hunninghake
DB, Pastanak RC, Smith SC Jr., Stone NJ, for the Coordinating Committee of the
National Cholesterol Education Program. Implication of recent clinical trials for
the National Cholesterol Education Program adult Treatment Panel III Guidelines. Circulation 2004;110:227–239.
14. Genest J, Libby P, Gotto AM. Lipoprotein disorders and cardiovascular risk.
In: Braunwald E, Libby P, Zipes D, eds. Braunwald’s Heart Disease. Philadelphia: WB Saunders, 2004;1013–1033.
15. Bucher HC, Griffith LE, Guyatt GH. Systematic review on the risk and
benefit of different cholesterol-lowering interventions. Arterioscler Thromb Vasc
Biol 1999;19:187–195.
16. Freemantle N, Cleland J, Young P, Mason J, Harrison J. Beta blockade after
myocardial infarction: systematic review and meta regression analysis. BMJ
1999;26318:1730 –1737.
17. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. Effects of an
angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in
high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med 2000;342:145–153.
18. Fox KM, and the EURopean trial On reduction of cardiac events with
Perindopril in stable coronary Artery disease Investigators. Efficacy of perindopril in reduction of cardiovascular events among patients with stable coronary
artery disease: randomised, double-blind, placebo-controlled, multicentre trial
(the EUROPA study). Lancet 2003;362:782–788.
19. Pfeffer MA, McMurray JJ, Velazquez EJ, Rouleau JL, Kober L, Maggioni
AP, Solomon SD, Swedberg K, Van de Werf F, White H, et al, and the Valsartan
in Acute Myocardial Infarction Trial Investigators. Valsartan, captopril, or both in
myocardial infarction complicated by heart failure, left ventricular dysfunction,
or both. N Engl J Med 2003;349:1893–1906.
20. Nissen SE, Tuzcu EM, Schoenhagen P, Brown BG, Ganz P, Vogel RA,
Crowe T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria AN, and the
REVERSAL Investigators. Effect of intensive compared with moderate lipidlowering therapy on progression of coronary atherosclerosis: a randomized controlled trial. JAMA 2004;291:1071–1080.
BRIEF REPORTS
253
Benefits of Niacin by Glycemic Status in Patients With
Healed Myocardial Infarction (from the Coronary
Drug Project)
Paul L. Canner,
PhD,
Curt D. Furberg, MD, PhD, Michael L. Terrin,
Mark E. McGovern, MD
The Coronary Drug Project, conducted during 1966 to
1974, was a randomized, double-blind, placebo-controlled trial of 5 lipid-modifying agents in 8,341 men
with previous myocardial infarction.1 Among the 5
drug treatment regimens, only niacin significantly reduced the risk of (1) cardiovascular events during a
mean follow-up of 6.2 years and (2) total mortality
during 6.2 years with study treatment plus an additional 9 years of post-trial follow-up (Figure 1).2,3
Cardiovascular and total mortality outcomes in the
niacin and placebo groups are presented by baseline
glycemic status and by change in glycemic status
from baseline to year 1. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:254 –257)
iacin increases plasma glucose levels in some
patients through mechanisms that are not comN
pletely understood. As a result, caution has been
4
urged in the use of niacin among patients with elevated fasting glucose levels or overt diabetes.5,6
Whether the effect on glucose control results in any
disadvantage with respect to cardiovascular events or
mortality risk has not been studied in detail. A previously reported finding may cast some light on the
matter. In the Coronary Drug Project (CDP), the niacin–placebo difference in 15-year total mortality was
similar between patients with baseline fasting plasma
glucose (FPG) levels ⬍100 mg/dl (48.6% vs 53.5%)
and in patients with FPG levels ⱖ100 mg/dl (56.1%
vs 63.9%).3 Patients treated with insulin were not
eligible for enrollment in the CDP. The present investigation examines in detail the effects of niacin on
cardiovascular events and total mortality in patient
subgroups defined by glycemic status.
•••
Event rates were obtained for recurrent nonfatal
definite myocardial infarction (MI) and for coronary
heart disease death or MI at 6 years, and for total
mortality at 15 years by baseline fasting and 1-hour
plasma glucose levels and by changes in fasting and
1-hour plasma glucose levels from baseline to year 1.
Treatment group comparisons were made using the
From the Maryland Medical Research Institute, Baltimore, Maryland;
Wake Forest University, Winston-Salem, North Carolina; and Kos
Division of Medical Affairs, Weston, Florida. This report was supported by an unrestricted research grant from Kos Pharmaceuticals,
Inc. Dr. Canner’s address is: Maryland Medical Research Institute Inc.,
600 Wyndhurst Avenue, Baltimore, Maryland 21210. E-mail:
plcanner@juno.com. Manuscript received June 3, 2004; revised
manuscript received and accepted August 31, 2004.
254
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MD,
and
FIGURE 1. CDP: results for cardiovascular end points at 6 years
and mortality at 15 years. *p <0.005; †p <0.05. CHD ⴝ coronary heart disease; MI ⴝ nonfatal definite MI.
Cox proportional-hazards model (SAS software, version 8.2, SAS Inc., Cary, North Carolina),7 which
yielded a hazard ratio for the niacin–placebo comparison of event rates in each glucose subgroup. A hazard
ratio of ⬍1.0 indicates a lesser event rate with niacin
than with placebo. The hazard ratios for the glucose
subgroups were evaluated for homogeneity using a
Cox regression model with 6-year MI, 6-year coronary heart disease death or MI, or 15-year total mortality as the dependent variable and treatment group
(niacin coded 1 or placebo coded 2), glucose category
(e.g., ⬍95, 95 to 104, 105 to 125, ⱖ 126 mg/dl coded
1, 2, 3, or 4, respectively), and the product of treatment group and glucose category codes as the independent variables. If the z value for the estimated
regression coefficient for this product (or interaction)
term is reasonably close to zero and not close to
⫾1.96, it can be inferred that the subgroup hazard
ratios are homogeneous or do not differ significantly
from one another. A negative interaction z value indicates that the hazard ratios are getting smaller (i.e.,
favoring niacin) as the baseline glucose level or the
change in glucose levels at year 1 increases. A positive interaction z value indicates that the niacin effect
is becoming less advantageous with increasing glucose levels. In analyses based on change in fasting or
1-hour plasma glucose levels from baseline to year 1,
nonfatal events occurring before the year 1 visit are
excluded.
This analysis included all but 2 patients from the
original niacin (n ⫽ 1,119) and placebo (n ⫽ 2,789)
cohorts. When transferring the CDP data from 3 magnetic tapes to more modern electronic storage devices,
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.013
TABLE 1 Coronary Drug Project: Mean Glycemia in the Niacin and Placebo Groups for the Cohort of Patients Completing
Closeout Visit (CV) Two
Fasting Plasma Glucose (mg/dl)
Time
Baseline
Yr 1
Yr 2
Yr 4
CV1*
CV2†
1-h Plasma Glucose (mg/dl)
Niacin
(n ⫽ 567)
Placebo
(n ⫽ 1,447)
p Value
Niacin
(n ⫽ 542)
Placebo
(n ⫽ 1,382)
p Value
101
107
107
108
108
105
100
101
102
104
106
107
NS
⬍0.001
⬍0.001
⬍0.01
NS
NS
168
179
179
183
188
171
169
164
165
170
177
177
NS
⬍0.001
⬍0.001
⬍0.001
⬍0.001
⬍0.05
*CV1 ⫽ first closeout visit (i.e., the visit at which study treatment was ended).
†
CV2 ⫽ ⬃4 months after CV1.
TABLE 2 Percentages of Events by Baseline Glucose Levels in the Niacin and Placebo Groups in the Coronary Drug Project
Fasting Plasma Glucose (mg/dl)
Event*
MI at 6 yrs
CHD death or MI at 6 yrs
Total mortality at 15 yrs
⬍95
(n ⫽ 424, 1,026)‡
95–104
(n ⫽ 409, 1,028)
105–125
(n ⫽ 216, 552)
ⱖ126
(n ⫽ 70, 181)
z Value for
Interaction†
11.1 vs 15.4
(0.70)
24.8 vs 32.4
(0.74)
50.5 vs 53.9
(0.91)
9.8 vs 12.9
(0.76)
27.9 vs 29.6
(0.94)
47.7 vs 55.5
(0.82)
10.6 vs 14.3
(0.75)
26.4 vs 31.7
(0.84)
56.5 vs 63.9
(0.86)
7.1 vs 15.5
(0.43)
37.1 vs 44.8
(0.76)
71.4 vs 77.9
(0.83)
⫺0.44
0.47
⫺0.63
1-h Plasma Glucose (mg/dl)
MI at 6 yrs
CHD death or MI at 6 yrs
Total mortality at 15 yrs
⬍140
(n ⫽ 299, 765)‡
140–179
(n ⫽ 386, 919)
180–219
(n ⫽ 246, 654)
ⱖ220
(n ⫽ 182, 443)
9.7 vs 13.1
(0.73)
25.4 vs 27.8
(0.89)
46.5 vs 49.8
(0.91)
11.1 vs 14.6
(0.73)
25.1 vs 32.3
(0.75)
50.8 vs 56.4
(0.86)
9.8 vs 13.6
(0.74)
28.0 vs 31.5
(0.92)
49.2 vs 61.2
(0.76)
10.4 vs 16.7
(0.61)
31.9 vs 39.5
(0.79)
65.9 vs 70.9
(0.86)
⫺0.47
⫺0.23
⫺0.66
*Event rates given as percent in niacin group versus percent in placebo group (hazard ratio).
†
z value between ⫾ 1.96 indicates homogeneity or no statistical difference (p ⱖ0.05) in hazard ratios among the various glucose levels.
‡
Sample sizes for the niacin and placebo groups, respectively.
CHD ⫽ coronary heart disease; MI ⫽ nonfatal definite MI.
data for the last patient on each of the first 2 tapes
were lost. Therefore, the actual placebo cohort for the
present analysis numbers 2,787. One-hour plasma glucose determinations were not obtained at baseline for
6 patients taking niacin and 6 taking placebo due to
problems with the glucose challenge. Analyses of
glucose change from baseline to year 1 excluded patients who either died, withdrew during the first year,
or missed the first annual follow-up visit. Thus, 965
and 2,471 patients in the niacin and placebo groups,
respectively, contributed to the analyses of change in
FPG levels, and 959 and 2,460 correspondingly contributed to the analyses of change in 1-hour plasma
glucose levels.
At baseline, about 42% of patients in the CDP had
elevated FPG levels (ⱖ100 mg/dl), and approximately
39% had impaired glucose tolerance as indicated by
elevated 1-hour plasma glucose levels (post-glucose
load) ⱖ180 mg/dl.2 Approximately 6% used oral hypoglycemic agents or had urine positive for glucose.
During the trial, niacin increased fasting and
1-hour plasma glucose levels (Table 1); however,
there was little difference at study end between the
niacin and placebo groups in the percentage of patients who were newly prescribed insulin (1.3% vs
0.9% placebo; p ⫽ NS) or oral hypoglycemic agents
(4.3% vs 4.6%; p ⫽ NS).2
Compared with placebo, niacin reduced the risk of
6-year recurrent MI and coronary heart disease death
or MI and 15-year total mortality similarly in patients
at all levels of baseline FPG, including those with
FPG levels ⱖ126 mg/dl (the current definition of
diabetes) or 1-hour plasma glucose levels of ⱖ180 or
ⱖ220 mg/dl (Table 2 and Figures 2 and 3). The
beneficial effect of niacin on cardiovascular events
and total mortality was not diminished, even among
those with the highest baseline fasting or 1-hour
plasma glucose levels.
Compared with placebo, niacin reduced the risk of
6-year recurrent MI and coronary heart disease death
or MI and 15-year total mortality similarly in patients
at all levels of change in fasting or 1-hour plasma
BRIEF REPORTS
255
FIGURE 2. CDP at 6 years: MI by baseline FPG. z Value for interaction ⴝ ⴚ0.44 (p ⴝ 0.66), indicating homogeneity of treatment
effects across all levels of FPG. Abbreviation as in Figure 1.
FIGURE 3. CDP at 15 years: total mortality by baseline FPG. z
Value for interaction ⴝ ⴚ0.63 (p ⴝ 0.53), indicating homogeneity.
TABLE 3 Percentages of Events by Changes in Baseline Glucose Levels at Year One in the Niacin and Placebo Groups in the
Coronary Drug Project
Change in Fasting Plasma Glucose (mg/dl)
Event*
MI at 6 yrs
CHD death or MI at 6 yrs
Total mortality at 15 yrs
⬍0
(n ⫽ 318, 1,165)‡
0 –9
(n ⫽ 346, 941)
ⱖ10
(n ⫽ 301, 365)
z Value for
Interaction†
7.9 vs 12.7
(0.62)
23.0 vs 26.2
(0.87)
49.4 vs 55.0
(0.88)
9.0 vs 12.9
(0.68)
20.2 vs 26.2
(0.75)
45.4 vs 51.9
(0.82)
9.0 vs 15.1
(0.55)
20.6 vs 34.5
(0.55)
49.8 vs 64.1
(0.69)
⫺0.21
⫺1.93
⫺1.14
Change in 1-h Plasma Glucose (mg/dl)
MI at 6 yrs
CHD death or MI at 6 yrs
Total mortality at 15 yrs
⬍0
(n ⫽ 395, 1,357)‡
0–29
(n ⫽ 270, 670)
ⱖ30
(n ⫽ 294, 433)
7.1 vs 13.5
(0.51)
20.8 vs 27.3
(0.74)
48.4 vs 53.5
(0.87)
8.9 vs 12.4
(0.71)
21.5 vs 26.9
(0.78)
47.4 vs 53.7
(0.86)
10.5 vs 13.2
(0.78)
21.4 vs 29.6
(0.70)
47.3 vs 62.6
(0.66)
1.49
⫺0.18
⫺1.80
*Event rates given as percentage in niacin group versus percentage in placebo group (hazard ratio).
†
z value between ⫾ 1.96 indicates homogeneity or no statistical difference (p ⱖ0.05) in hazard ratios among the various glucose levels.
‡
Sample sizes for the niacin and placebo groups, respectively.
Abbreviations as in Table 2.
glucose levels from baseline to year 1 (Table 3 and
Figures 4 and 5). As with baseline fasting and 1-hour
plasma glucose levels, the effect of niacin on cardiovascular events and mortality was not diminished,
even among patients with the largest increases in
fasting or 1-hour plasma glucose levels from baseline
to year 1. Recurrent MI by change in 1-hour plasma
glucose levels was an exception to this trend (Table 3).
Some caution must be exercised in interpreting the
results based on glucose change, because these are
variables defined at follow-up rather than at baseline.8
Glucose change in the CDP is positively correlated
with level of adherence to study protocol prescription
in patients taking niacin, but not in those taking placebos. Also, as reported elsewhere,8 the level of adherence was inversely correlated with total mortality
in the CDP. Nevertheless, analyses of events by glucose change for patients at full adherence for the first
256 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
year of follow-up gave results very similar to those in
Table 3 without regard to adherence.
•••
The results of this study confirm and extend the
evidence from the CDP relating to the effects of niacin
in reducing the risk of cardiovascular events and mortality.2,3 Among patients receiving placebo, elevated
baseline fasting and 1-hour plasma glucose levels
were both positively associated with total mortality.9
This can also be seen in the present article by comparing the placebo group data across each line of
Table 2 for 15-year mortality. At the same time, it has
been reported from both the CDP and other studies
that niacin treatment, especially at high doses, may
elevate plasma glucose levels in some patients.2,5,10
The present results show that the increase in glucose
levels with niacin did not translate into any disadvantage
with respect to cardiovascular events or mortality risk.
JANUARY 15, 2005
FIGURE 4. CDP at 6 years: MI by change in year 1 FPG. z Value
for interaction ⴝ ⴚ0.21 (p ⴝ 0.83), indicating homogeneity. Abbreviation as in Figure 1.
in this subgroup had a baseline FPG ⱖ126 mg/dl;
thus, the numbers are too small to perform any meaningful analyses.
In the aggregate, the findings detailed in this
report demonstrate that niacin has favorable effects
on clinical outcome in patients with evidence of
abnormal glucose metabolism or overt diabetes, although patients treated with insulin were excluded
from enrollment in the CDP. Recently, there has
been a resurgence of interest in using niacin for the
treatment of dyslipidemia, especially for raising
low levels of high-density lipoprotein cholesterol.11
Although niacin is the most effective agent for
raising low levels of high-density lipoprotein cholesterol, it can cause hyperglycemia in some patients, and its use in patients with abnormal glucose
metabolism or overt diabetes mellitus has previously been cautioned.6 Such caution is not supported by these analyses of the CDP.
1. The Coronary Drug Project Research Group. The Coronary Drug Project.
Design, methods, and baseline results. Circulation 1973;47(suppl I):I-1–I-50.
2. The Coronary Drug Project Research Group. Clofibrate and niacin in coronary
heart disease. JAMA 1975;231:360 –381.
3. Canner PL, Berge KG, Wenger NK, Stamler J, Friedman L, Prineas RJ,
FIGURE 5. CDP at 15 years: total mortality by change in year 1
FPG. z Value for interaction ⴝ ⴚ1.14 (p ⴝ 0.25), indicating
homogeneity.
In the CDP, niacin produced average reductions of
10% in serum cholesterol and 26% in serum triglycerides over the first 5 years of follow-up despite modest compliance with the study medication.2 Niacin is
known to raise low levels of high-density lipoprotein
cholesterol.6 In the CDP, only 492 patients (138 taking niacin and 354 taking placebos) at baseline and
none at follow-up had high-density lipoprotein cholesterol determinations. Only 6 patients taking niacin
Friedewald W, for the Coronary Drug Project Research Group. Fifteen year
mortality in Coronary Drug Project patients: long-term benefit with niacin. J Am
Coll Cardiol 1986;8:1245–1255.
4. McGovern ME. Use of nicotinic acid in patients with elevated fasting glucose,
diabetes, or metabolic syndrome. Br J Diabetes Vasc Dis 2004;4:78 – 85.
5. Garg A, Grundy SM. Nicotinic acid as therapy for dyslipidemia in non–insulindependent diabetes mellitus. JAMA 1990;264:723–726.
6. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and
Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final
report. Circulation 2002;106:3143–3421.
7. Cox DR. Regression models and life-tables. J R Stat Soc B 1972;34:187–202.
8. The Coronary Drug Project Research Group. Influence of adherence to treatment and response of cholesterol on mortality in the Coronary Drug Project.
N Engl J Med 1980;303:1038 –1041.
9. The Coronary Drug Project Research Group. The prognostic importance of
plasma glucose levels and of the use of oral hypoglycemic drugs after myocardial
infarction in men. Diabetes 1977;26:453– 465.
10. Berge KG, Molnar GD. Effects of nicotinic acid on clinical aspects of
carbohydrate metabolism. In: Altschul R, ed. Niacin in Vascular Disorders and
Hyperlipidemia. Springfield, IL: Charles C Thomas, 1964:136 –155.
11. Miller M. Niacin as a component of combination therapy for dyslipidemia.
Mayo Clin Proc 2003;78:735–742.
BRIEF REPORTS
257
Occurrence of Acute Myocardial Infarction in
Worcester, Massachusetts, Before, During, and After
the Terrorists Attacks in New York City and
Washington, DC, on 11 September 2001
Robert J. Goldberg, PhD, Frederick Spencer, MD, Darleen Lessard, MS,
Jorge Yarzebski, MD, MPH, Craig Lareau, BS, and Joel M. Gore, MD
The Worcester Heart Attack Study is an ongoing population-based investigation examining changes over time
in the incidence and case-fatality rates of acute myocardial infarction (AMI) in residents of metropolitan
Worcester, Massachusetts.1–3 Using this large population-based database of patients with independently validated AMI, we examined the impact of the September
11th tragedy on the occurrence rates of AMI in residents
of greater Worcester presenting to all area hospitals
during 2001 compared with rates in the previous 10
years. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:258 –260)
esidents of the Worcester metropolitan area hospitalized with a primary or secondary discharge
R
diagnosis of acute myocardial infarction (AMI)
(International Classification of Disease-9th Revision
[ICD-9] code 410) from all Worcester standard metropolitan statistical area hospitals during 1991, 1993,
1995, 1997, 1999, and 2001 formed the study population of this report. The present study is part of an
ongoing, population-based investigation examining
changes over time in the incidence, hospital, and longterm case-fatality rates of residents of the Worcester
metropolitan area hospitalized with confirmed
AMI.1–3
•••
Although additional earlier study years were included in this community-wide investigation (1975 to
1990), we restricted the study sample to patients hospitalized with validated AMI during the decade before
and during 2001. This time period was chosen to
examine relatively contemporary and decade-long
trends in the occurrence of AMI. This period was also
chosen to provide a more systematic and extensive
contrast of the usual, or endemic, magnitude of AMI
between 1991 and 1999 and evaluation of the possible
“epidemic” effects of the World Trade Center attacks
on September 11, 2001. These 6 annual periods (1991,
1993, 1995, 1997, 1999, and 2001) were originally
From the Department of Medicine, Division of Cardiovascular Medicine, University of Massachusetts Medical School, Worcester,
Massachusetts. This report was supported by Grant RO1 HL35434
from the National Heart, Lung, and Blood Institute, Bethesda,
Maryland. Dr. Goldberg’s address is: Division of Cardiovascular
Medicine, Department of Medicine, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, Massachusetts
01655. E-mail: goldberr@ummhc.org. Manuscript received May 11,
2004; revised manuscript received and accepted August 31, 2004.
258
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
selected because of the availability of grant funding
and to examine changes over time in our principal
study end points over an alternating yearly basis. All
11 hospitals in the Worcester standard metropolitan
statistical area (2000 census population, 478,000) participated in this study.
The medical records of greater Worcester residents
with a discharge diagnosis of AMI from these hospitals were individually reviewed and validated according to pre-established criteria that have been previously described.1–3 The present report is based on the
6,072 residents of the Worcester metropolitan area
who satisfied our predefined diagnostic criteria for
AMI. Of these, 4,834 patients were hospitalized at all
greater Worcester medical centers during the study
years between 1991 and 1999, whereas 1,238 patients
were hospitalized in 2001.
The hospital records of patients with validated
AMI were abstracted for information about patient’s
age, gender, medical history, occurrence of several
clinical complications during hospitalization, and
AMI order (initial vs prior) and type (Q-wave or
non–Q wave). Detailed descriptions of these factors
have been provided in previous publications.1–3 Information on possible precipitating factors associated
with onset of AMI was not collected.
We examined the average daily occurrence of all
AMIs to determine whether the frequency of AMI
events increased in relation to the terrorist attacks on
September 11 (Figure 1). In 2001, the average daily
number of cases of AMI was 3.4 in the 9-month
period ending on Sepember 10. On the day of September 11, there were 7 cases of AMI, 9 cases of AMI
on September 12, and 13 additional cases during the
remainder of the week. In the month after September
11 there were 122 confirmed cases of AMI (daily
mean 4.1), and 322 (mean 3.5) in the 3-month period
thereafter. In contrast, among patients hospitalized
during the “unexposed” endemic or control period
between 1991 and 1999, the average number of
greater Worcester residents hospitalized with confirmed AMI on a daily basis was 2.7 over the period
January 1 to September 10. On September 11, there
were an average of 1.8 cases of AMI and an average
of 2.6 cases on September 12 during this period. The
average daily number of cases of AMI was 2.7 in the
1-week, 1-month, and 3-month period after September
11 in the control or comparison study years. Essentially similar trends were observed in patients who
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.094
FIGURE 1. Occurrence of acute myocardial infarction during selected time periods
(Worcester Heart Attack Study).
TABLE 1 Descriptive Characteristics of Patients With Acute Myocardial Infarction
(AMI)
Pre 9/11
(n ⫽ 3,436)
Characteristic
Age (median yrs)
Age (yrs)
⬍55
55–64
65–74
ⱖ75
Men
Angina pectoris
Diabetes mellitus
Heart failure
Hypertension
Stroke
AMI characteristics
Initial
Q wave
Duration of prehospital
delay (median h)
9/11 and
9/12
(n ⫽ 22)
Month after
9/11 (9/12–
10/11)
(n ⫽ 406)
3 Months After
9/11 (9/12–
12/11)
(n ⫽ 1,208)
According to Time Period of Hospitalization: 1991 to 1999
71
69
72
72
16.4%
17.2%
26.6%
39.8%
57.4%
26.4%
29.7%
20.3%
58.9%
10.4%
18.2%
18.2%
31.8%
31.8%
72.7%
27.3%
40.9%
27.3%
45.5%
0%
18.5%
16.8%
23.7%
41.1%
57.9%
21.4%
24.4%
19.2%
58.6%
9.9%
16.8%
17.0%
24.8%
41.5%
56.4%
22.4%
28.5%
18.6%
59.7%
10.4%
64.1%
37.5%
2.2
59.1%
27.3%
3.1
65.8%
38.2%
2.2
65.0%
35.9%
2.1
According to Time Period of Hospitalization: 2001
(n ⫽ 856)
Age (median yrs)
Age (yrs)
⬍55
55–64
65–74
ⱖ75
Men
Angina pectoris
Diabetes mellitus
Heart failure
Hypertension
Stroke
AMI characteristics
Initial
Q wave
Duration of prehospital
delay (median h)
(n ⫽ 16)
(n ⫽ 122)
(n ⫽ 322)
74
71
73
73
14.1%
13.6%
23.0%
49.3%
57.7%
23.8%
33.2%
24.0%
69.4%
12.1%
25.0%
18.8%
12.5%
43.8%
62.5%
18.8%
25.0%
12.5%
50.0%
0%
12.3%
18.9%
23.8%
45.1%
54.6%
24.6%
34.4%
24.6%
70.5%
11.5%
14.3%
18.9%
22.7%
44.1%
57.6%
20.5%
30.8%
24.2%
66.8%
11.5%
64.8%
22.4%
1.8
75.0%
18.8%
2.3
64.8%
18.0%
2.0
65.8%
19.3%
2.0
presented to all greater Worcester
hospitals with an initial AMI.
The risk of AMI was significantly increased on the day of September 11, 2001 (odds ratio [OR]
3.04, 95% confidence interval [CI]
1.13 to 8.17), and on September 11
and 12, 2001 (OR 2.85, 95% CI 1.49
to 5.45) compared with similar periods during the control or unexposed
period. There was no significant increase in the risk of AMI during the
week after September 11, 2001 (OR
1.23, 95% CI 0.81 to 1.87), the
month after September 11 (OR 1.20,
95% CI 0.97 to 1.48), and the
3-month period thereafter (OR 1.05,
95% CI 0.91 to 1.22) compared with
similar periods during the unexposed or control years.
To examine the potential role of
chance in the increased occurrence of
AMIs on the day of September 11,
2001, and the day thereafter, we performed a simple count of the number
of days that there were ⱖ7 confirmed
cases of AMI on consecutive days in
the unexposed cohorts. There were no
times during the control years between
1991 and 1999 when the number of
patients hospitalized with AMI on consecutive days was ⱖ16 as was seen on
September 11 and 12, 2001.
We examined the demographic and
clinical characteristics of patients with
AMI during different study periods to
determine whether there were any differences in the distribution of these
characteristics according to the time
period of hospitalization (Table 1). Patients experiencing AMI on September
11 and 12, 2001, included a greater
proportion of younger patients,
women, and those with an initial AMI
compared with patients hospitalized on
the same days during the control years.
Appropriate caution needs to be exercised in the interpretation of these data,
however, given the small number of
cases of AMI occurring on September
11 and 12.
•••
The results of this multihospital,
population-based study suggest that
the impact of the tragic events in
New York City and the Pentagon on
September 11, 2001 may have resulted in a transient increase in the
occurrence of AMIs in the greater
Worcester population on September
11 and on the day thereafter. The
Worcester metropolitan area is ⬍200
BRIEF REPORTS
259
miles from New York City and both of the planes that
crashed into the twin towers on September 11 originated from Boston, Massachusetts. The frequency of
AMIs in the weeks to months after September 11,
2001, was similar to earlier periods of this year and
similar to the frequency of AMIs observed during the
control years of 1991 to 1999. Because of the relatively small number of hospitalized cases of AMI
occurring in the metropolitan Worcester population on
a daily basis, it is difficult and tenuous to extrapolate
this pattern of events to a cause– effect association.
There are limited published reports on the effects of
natural disasters, such as earthquakes, on the occurrence
or death rates associated with atherosclerotic cardiovascular disease.4 – 8 There is an even sparser literature describing the effects of war or battlefield-type situations
on the risk of acute cardiac disease in the civilian or
military population.9,10 We were able to identify a small
number of descriptive studies that examined the impact
of the events of September 11 on hospital admissions for
cardiac disease, or on cardiovascular-associated mortality, in New York City.11,12 A terrorist attack may be
viewed in a relatively similar context to that of an earthquake or natural disaster, in that there is generally little
warning that the event will occur.
Our results suggest that the terrorist attacks on
September 11 may have resulted in a transitory increase in the number of heart attacks occurring in
greater Worcester residents. Although unknown, and
if real, this may have resulted in the occurrence of a
considerable number of additional heart attacks nationally in the few days after the events of September
11. Conversely, the deleterious health effects associated with the World Trade Center and Pentagon attacks may have been limited to persons residing on the
East Coast of the United States. Although the play of
chance cannot be completely ruled out given the small
number of AMIs occurring in the greater Worcester
population on a daily basis, these attacks may have
placed undue psychological and emotional strain and
stress in at-risk patients, resulting in an increased
triggering of acute coronary events.
High-risk cardiac patients may be at particularly increased risk for adverse coronary events after an acute
260 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
life stressor.13 Although suggestive evidence for the role
of different triggering agents in the precipitation of AMI
has been obtained from previous studies,14 it remains
unclear what may have caused this increase in the number of AMIs in the greater Worcester population on
September 11 and the day thereafter. The shock of the
initial event, uncertainty with regard to one’s future, and
a number of additional factors may have contributed to
the observed increase in acute coronary events. Our
findings reinforce the need for the exploration of different mechanisms through which acute stress, or other
potential triggering agents, may result in the onset of
AMI.
1. Goldberg RJ, Gore JM, Alpert JS, Dalen JE. Recent changes in the attack rates
and survival rates of acute myocardial infarction (1975–1981): The Worcester
Heart Attack Study. JAMA 1986;255:2774 –2779.
2. Goldberg RJ, Gore JM, Alpert JS, Dalen JE. Incidence and case fatality rates
of acute myocardial infarction (1975–1984): The Worcester Heart Attack Study.
Am Heart J 1988;115:761–767.
3. Goldberg RJ, Yarzebski J, Lessard D, Gore JM. A two-decades (1975–1995)
long experience in the incidence, in-hospital and long-term case-fatality rates of
acute myocardial infarction: a community-wide perspective. J Am Coll Cardiol
1999;33:1533–1539.
4. Kloner RA, Leor J, Poole WK, Perritt R. Population-based analysis of the
effect of the Northridge earthquake on cardiac death in Los Angeles County,
California. J Am Coll Cardiol 1997;30:1174 –1180.
5. Suzuki S, Sakamoto S, Miki T, Matsuo T. Hanshin-Awaji earthquake and acute
myocardial infarction. Lancet 1995;345:981–982.
6. Leor J, Kloner RA. The Northridge earthquake as a trigger for acute myocardial infarction. Am J Cardiol 1996;77:1230 –1232.
7. Trichopoulos D, Katsouyanni K, Zavitsanos X, Tzonou A, Dalla-Vorgia P.
Psychological stress and fatal heart attack: the Athens (1981) earthquake natural
experiment. Lancet 1983;1:441– 444.
8. Dobson AJ, Alexander HM, Malcolm JA, Steele PL, Miles TA. Heart attacks
and the Newcastle earthquake. Med J Aust 1991;155:757–761.
9. Meisel SR, Kutz I, Dayan KI, Pauzner H, Chetboun I, Arbel Y, David D. Effect
of Iraqi missile war on incidence of acute myocardial infarction and sudden death
in Israeli civilians. Lancet 1991;338:660 – 661.
10. Rumboldt Z, Guinio L, Miric D, Polic S, Bozic I, Tonkic A. War-stress
induced medical emergencies in South Croatia. Lancet 1993;341:965–966.
11. Chi JS, Speakman MT, Poole WK, Kandefer SC, Kloner RA. Hospital
admissions for cardiac events in New York City after September 11, 2001. Am J
Cardiol 2003;92:61– 63.
12. Chi JS, Poole K, Kandefer SC, Kloner RA. Cardiovascular mortality in New
York City after September 11, 2001. Am J Cardiol 2003;92:857– 861.
13. Qureshi EA, Merla V, Steinberg J, Rozanski A. Terrorism and the heart:
implications for arrhythmogenesis and coronary artery disease. Cardiac Electrophys Rev 2003;7:80 – 84.
14. Muller JE, Abela GS, Nesto RW, Tofler GH. Triggers, acute risk factors, and
vulnerable plaques: the lexicon of a new frontier. J Am Coll Cardiol 1994;23:
809 – 813.
JANUARY 15, 2005
Characteristics and Outcomes of Patients With Acute
Myocardial Infarction and Angiographically Normal
Coronary Arteries
Alf Inge Larsen, MD, PhD, P. Diane Galbraith, BN, William A. Ghali, MD, MPH,
Colleen M. Norris, PhD, Michelle M. Graham, MD, and Merril L. Knudtson, MD, for the
APPROACH Investigators*
This study on patients undergoing coronary angiography for acute myocardial infarction demonstrated that
2.8% of patients had angiographically normal coronary
arteries and that these patients have a better prognosis
than patients with angiographically verified coronary
artery disease. The trend toward a higher prevalence of
malignancy in this unique patient group raises the possibility of malignancy-induced hypercoagulability or inflammation as an underlying etiologic factor. 䊚2005
by Excerpta Medica Inc.
(Am J Cardiol 2005;95:261–263)
he aims of the present study were (1) to evaluate a
cohort of patients with acute myocardial infarction
T
(AMI) who underwent subsequent coronary angiography to determine the proportion of patients with normal coronary arteries and (2) to compare the clinical
characteristics and outcomes of AMI patients with
normal coronary arteries with those of patients with
varying degrees of coronary artery disease (CAD).
•••
The Alberta Provincial Project for Outcome Assessment in Coronary Heart disease (APPROACH) is a
From the Cardiology Division, University of Bergen, Central Hospital,
Stavanger, Norway; the Department of Community Health Sciences
and Centre for Health and Policy Studies, Calgary, Alberta; University
of Alberta, Edmonton, Alberta; Division of Cardiology, University of
Alberta Hospital, Edmonton, Alberta; and Foothills Medical Centre,
Calgary, Alberta, Canada. APPROACH was initially funded with a
grant from the W. Garfield Weston Foundation. The ongoing operation
has been made possible by contributions from Merck Frosst Canada Inc.,
Pointe-Claire, Dorval, Quebec; Monsanto Canada Inc.– Searle, Mississauga, Ontario; Eli Lilly Canada Inc., Toronto, Ontario; Guidant Corporation, Santa Clara, California; Boston Scientific Ltd., Mississauga, Ontario; Hoffmann-La Roche Ltd., Mississauga, Ontario; and Johnson &
Johnson Inc.–Cordis, Markham, Ontario, Canada. Dr. Knudtson received
partial support from the Libin Cardiovascular Trust Fund, University of
Calgary, Calgary, Alberta, Canada. Dr. Ghali was supported by a
Government of Canada Research Chair in Health Services Research and
by a Health Scholar Award from the Alberta Heritage Foundation for
Medical Research, Edmonton, Alberta. Miss Galbraith and Dr. Norris
were supported by Canadian Cardiovascular Outcomes Research Team
(www.torchalberta.ca) student fellowship, funded by an operating grant
from the Canadian Institutes of Health Research and the Heart and Stroke
Foundation, and Tomorrows Research Cardiovascular Health Professionals (www.torchalberta.ca) student fellowship. Dr. Knudtson’s address is:
University of Calgary/Foothills Hospital, Division of Cardiology,
1403-29 Street NW, Calgary, Alberta, Canada, T2N 2T9. E-mail:
knudtson@shaw.ca. Manuscript received April 4, 2004; revised
manuscript received and accepted August 31, 2004.
*See Appendix for members of the APPROACH Clinical Steering
Committee.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
clinical data collection initiative in which data are prospectively compiled for all adult patients undergoing
cardiac catheterization in the province of Alberta, Canada.1 The database and its contents are described in detail
elsewhere.2– 4 The database also tracks subsequent revascularization procedures. The extent of CAD is documented in a computerized data template (HeartView,
Heartware, Durham, North Carolina). For this analysis,
patients were classified into 5 subgroups by severity of
CAD: (1) normal coronary arteries (no coronary artery
stenoses), (2) minor disease (⬍50% diameter stenoses),
(3) low-risk disease (1- or 2-vessel disease with ⬎50%
diameter stenoses), (4) high-risk disease (2-vessel disease with proximal left anterior descending artery or
3-vessel disease), and (5) left main disease.
We focused specifically on the subset of patients who
underwent cardiac catheterization with an indication of
AMI. This indication was defined in the registry by the
clinician who performed the procedure, and then data
were further enhanced by a post hoc data-merging procedure in which registry data were merged to International Classification of Disease-9th Revision (ICD-9)
administrative data.5 This process provides a 2-level
verification of the indication for catheterization.
Mortality was ascertained through semi-annual
merging with data from the Alberta Bureau of Vital
Statistics Records.
The emphasis of our analysis was a description of
clinical characteristics of patients with no angiographically obvious coronary artery lesions, referred
to herein as the “normal” group, but we also documented the characteristics of patients with increasing
degrees of CAD. We used chi-square tests, Fisher’s
exact tests, or t tests for statistical comparisons of
clinical characteristics for the normal coronary arteries
group versus all other groups combined. Survival
analysis after AMI was estimated using the KaplanMeier method with log-rank tests to compare survival
for the normal coronary arteries group versus all other
groups combined. All statistical analyses were performed using SPSS software, version 11.0 (SPSS Inc.,
Chicago Illinois).
From January 1995 to December 2000, 9,796 patients underwent coronary angiography with AMI
listed as the indication for the procedure. Of these, 273
(2.8%, 95% confidence interval 2.5 to 3.1) had angiographically normal coronary arteries.
Baseline characteristics of the patient groups according to increasing degrees of CAD are listed in
Table 1. Compared with patients with identifiable
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.014
261
TABLE 1 Demographics and Risk Factors for Patients With Acute Myocardial Infarction (AMI) and Normal Coronary Arteries
Compared With Patients With AMI and Identifiable Coronary Artery Disease
Variable
Mean age (yrs)
Patients ⬎75 yrs
Patients ⱕ45 yrs
Men
Hypertension
Hyperlipidemia
Diabetes mellitus
Smokers
Current smoker
Previous thrombolytic
Previous PCI
Cerebrovascular disease
Congestive heart failure
Peripheral vascular disease
Creatine ⬎200 mmol/L
Dialysis
Chronic lung disease
Liver/gastroIntestinal
Malignancy
Ejection fraction
⬍30
30–50
⬎50
Not done (unstable)
Missing
Normal
(n ⫽ 273)
⬍50
(n ⫽ 452)
Low Risk
(n ⫽ 4,942)
High Risk
(n ⫽ 3,430)
Left Main
(n ⫽ 699)
p
Value*
49.4
4.4%
39.6%
62.3%
34.1%
23.1%
3.7%
38.5%
31.1%
10.6%
0.0%
4.8%
11.0%
1.8%
2.2%
0.4%
16.5%
4.4%
5.1%
58.1
12.2%
17.5%
60.2%
45.6%
40.9%
13.1%
44.7%
38.3%
20.6%
2.0%
5.1%
15.5%
6.0%
1.5%
1.1%
15.7%
7.3%
3.3%
59.7
10.7%
11.7%
73.9%
43.7%
45.1%
14.9%
42.8%
42.8%
25.5%
5.7%
4.2%
14.3%
5.0%
2.0%
0.8%
9.3%
3.7%
3.0%
65.1
21.4%
4.5%
76.6%
55.9%
45.8%
26.4%
42.8%
31.7%
21.1%
7.8%
7.8%
25.5%
8.9%
4.7%
1.7%
11.8%
5.1%
3.3%
68.5
30.5%
2.0%
78.0%
56.4%
43.9%
25.8%
43.9%
26.6%
15.3%
5.6%
14.2%
36.1%
12.6%
7.0%
2.3%
16.7%
5.4%
5.3%
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.154
0.036
0.000
0.000
0.373
0.000
0.000
0.389
0.267
0.008
1.0
0.090
5.5%
10.9%
70.3%
4.0%
9.2%
2.4%
16.6%
67.9%
5.8%
7.3%
3.0%
24.7%
54.9%
8.5%
8.9%
7.7%
31.1%
40.4%
9.6%
11.2%
10.7%
34.9%
35.8%
11.0%
7.6%
*The p values represent comparisons between normal versus all other groups combined.
PCI ⫽ percutaneous coronary intervention.
TABLE 2 Treatment Strategy and Mortality in Acute Myocardial Infarction (AMI) Patients With Normal Coronary Arteries
Compared With AMI Patients With Identifiable Coronary Artery Disease
Variable
Treatment
Medical
Coronary bypass
PCI
Coronary bypass 30 d
Coronary bypass 1 yr*
PCI 30 d
PCI 1 yr*
Death 30 d
Death 1 yr*
Normal
(n ⫽ 274)
⬍50%
(n ⫽ 452)
Low Risk
(n ⫽ 4,942)
High Risk
(n ⫽ 3,430)
Left Main
(n ⫽ 699)
Total
(n ⫽ 9,796)
100%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
1.8%
4.0%
90.7%
0.4%
8.9%
0.2%
0.4%
6.7%
8.2%
0.7%
2.9%
27.3%
4.2%
68.5%
3.2%
5.5%
64.3%
68.0%
1.7%
3.3%
25.5%
31.7%
42.9%
22.2%
34.2%
40.1%
42.9%
5.1%
9.4%
22.7%
67.4%
9.9%
59.4%
69.4%
9.7%
10.7%
11.6%
19.2%
31.3%
18.0%
50.7%
13.6%
19.8%
47.5%
50.5%
3.5%
6.6%
*One-year data includes 30-day data.
Abbreviation as in Table 1.
CAD, the mean age was considerably lower in the
normal coronary group (p ⬍0.001). Compared with
patients with identifiable CAD, prevalence of hypertension, hyperlipidemia, diabetes mellitus, previous
thrombolytic therapy, and history of congestive heart
failure and peripheral vascular disease occurred less
often in the normal group. There was also an intriguing trend toward a difference in the prevalence of
malignancy between normal patients and patients in
all other groups. Patients in the latter groups were
older and therefore may have been expected to have a
higher prevalence of malignancy. The specific types
of malignancy in the normal group were prostatic
cancer (n ⫽ 2), skin cancer (n ⫽ 1), cervical cancer
(n ⫽ 2), breast cancer (n ⫽ 3), malignant myeloma
262 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
(n ⫽ 1), Hodgkins’ disease (n ⫽ 1), and bladder
cancer (n ⫽ 1). In 3 of the patients, the type of
malignancy was not available.
Table 2 demonstrates treatment strategy and outcomes at 30 days and 1 year for each of the patient
groups. No patient with normal angiograms after AMI
underwent a revascularization procedure within 1
year. Of the total group of patients with AMI who
underwent angiography, 68.7% underwent revascularization.
Mortality in normal patients was 1.8% at 30 days and
4.0% at 1 year. In all, 11 patients with normal coronary
arteries died within the first year. These mortality rates
were generally similar to those for patients with stenosis
⬍50% and the low-risk anatomy groups, but signifiJANUARY 15, 2005
the probability that this is due to chance alone is
relatively low. At a minimum, this finding warrants
further exploration.
Our other notable finding, diagnostic and etiologic
considerations aside, is that patients presenting with
the clinical syndrome of AMI with normal coranary
arteries have a significantly better prognosis than do
those with identified CAD.
FIGURE 1. Kaplan-Meier survival curves for patients with AMI
with normal coronary arteries compared with patients with AMI
who have coronary disease.
cantly less than the high-risk and left main groups (Table
2). Figure 1 shows survival curves for the group with
normal coronary arteries versus all other groups combined extending to up to 7 years and confirms the general
survival picture seen in the first year (log-rank test for
difference in survival across groups, p ⫽ 0.017). The
median time of follow-up to assess survival was 3.3
years (range 0 to 7.0).
•••
These findings indicate that approximately 2.8% of
patients undergoing coronary angiography for indication of an AMI have normal coronary arteries. Patients
in this group are younger and less likely to have
traditional risk factors for atherosclerosis, such as
hypertension, hyperlipidemia, and diabetes mellitus,
than are patients with documented coronary disease.
Interestingly, there was a trend toward a higher prevalence of malignancy in the normal group than in
other groups, although normal patients were younger
and therefore would be expected to be at lower risk for
malignancy.
Other investigators6 –11 speculated on potential etiologies and/or mechanisms underlying the clinical
syndrome of AMI with normal coronary arteries. Possible explanations include cocaine ingestion, coronary
spasm, misdiagnosed myocarditis, and hypercoagulability states (both inherited and acquired). Our finding
of a trend toward a higher prevalence of malignancy in
the group with normal coronary arteries raises the
possibility that malignancy, regardless of the type of
cancer, with associated coagulation disorders and/or
inflammation, may place these patients at higher
risk of thromboembolic events. Another possibility is
that aggressive cancer treatments, such as chemotherapy, radiation therapy, and major surgery, may predispose to physiologically stressed states and induce
AMI.12–14 Our finding does not meet traditional statistical significance (p ⬍0.05). However, at p ⫽ 0.09,
Acknowledgment: We appreciate the assistance of
the Calgary Health Region and the Capital Health
Authority for supporting on-line data entry. We gratefully acknowledge personnel in the cardiac catheterization laboratories for their diligence in data collection and entry: Holy Cross Hospital, Foothills Medical
Center, Royal Alexandra Hospital, and the University
of Alberta Hospital. We thank Jason Badry, BSc, our
systems analyst, who is responsible for software development and maintenance.
APPENDIX
Members of the APPROACH Clinical Steering Committee: William Hui,
MD (chair); Stephen Archer, MD; Micheal Curtis, MD; William A. Ghali, MD;
Michelle Graham, MD; Merril L. Knudtson, MD; Arvind Koshal, MD; Andrew
Maitland, MD; L. Brent Mitchell, MD; and Ross Tsuyuki, MD.
1. Ghali WA, Knudtson ML, for the APPROACH Investigators. Overview of
“APPROACH”—the Alberta Provincial Program for outcome assessment in
coronary heart disease. Can J Cardiol 2000;16:1225–1230.
2. Ghali WA, Quan H, Brant R, van Melle G, Norris CM, Faris PD, Knudtson
ML, for the APPROACH Investigators. A comparison of two methods for
calculating adjusted survival curves from proportional hazards models. JAMA
2001;286:1494 –1497.
3. Ghali WA, Faris PD, Galbraith PD, Norris CM, Curtis MJ, Saunders LD,
Dzavik V, Mitchell LB, Knudtson ML. Sex differences in access to coronary
revascularization after cardiac catheterization: importance of detailed clinical
data. Ann Intern Med 2002;136:723–732.
4. Graham MM, Ghali WA, Faris PD, Galbraith PD, Norris CM, Merril Knudtson, for the APPROACH Investigators. Survival after coronary revascularization
in the elderly. Circulation 2002;105:2378 –2384.
5. Norris CM, Ghali WA, Knudtson ML, Saunders LD, Naylor CD. Dealing with
missing data in observational health care outcome analysis. J Clin Epidemiol
2000;53:377–383.
6. Minor RL Jr, Scott BD, Brown DD, Winniford MD. Cocaine-induced myocardial infarction in patients with normal coronary arteries. Ann Intern Med
1991;115:797– 806.
7. Zurbano MJ, Heras M, Rigol M, Roig E, Epelde F, Miranda F, Sanz G, Escolar
G, Ordinas A. Cocaine administration enhances platelet reactivity to subendothelial components: studies in a pig model. Eur J Clin Invest 1997;27:116 –120.
8. Pierron F, Panagides D, Bonnet JL, Yvorra S, Desfossez L, Bory M. Spasm of
normal or irregular coronary arteries. Long-term outcome of 277 patients. Arch
Mal Coeur Vaiss 1995;88:1819 –1825.
9. Da Costa A, Tardy-Poncet B, Isaaz K, Cerisier A, Mismetti P, Simitsidis S,
Reynaud J, Tardy B, Piot M, Decousus H, Guyotat D. Prevalence of factor V
leiden and other inherited thrombophilias in young patients with myocardial
infarction and normal coronary arteries. Heart 1998;80:338 –340.
10. Mansourati J, Da Costa A, Munier S, Mercier B, Tardy B, Ferec C, Isaaz K,
Blanc JJ. Prevalence of factor V Leiden in patients with myocardial infarction and
normal coronary angiography. Thromb Haemost 2000;83:822– 825.
11. Caussin C, Ohanessian A, Lancelin B, Rahal S, Hennequin R, Dambrin G,
Brenot P, Angel CY, Paul JF. Coronary plaque burden detected by multislice
computed tomography after acute myocardial infarction with near-normal coronary arteries by angiography. Am J Cardiol 2003;92:849 – 852.
12. Zeymer U, Neuhaus KL. Acute coronary thrombosis and myocardial ischemia
following chemotherapy of Hodgkin’s disease. Onkologie 1990;13:221–224.
13. Kruit WH, Punt KJ, Goey SH, de Mulder PH, van Hoogenhuyze DC,
Henzen-Logmans SC, Stoter G. Cardiotoxicity as a dose-limiting factor in a
schedule of high dose bolus therapy with interleukin-2 and alpha-interferon: an
unexpectedly frequent complication. Cancer 1994;74:2850 –2856.
14. Scholz KH, Herrmann C, Tebbe U, Chemnitius JM, Helmchen U, Kreuzer H.
Myocardial infarction in young patients with Hodgkin’s disease—potential
pathogenic role of radiotherapy, chemotherapy, and splenectomy. Clin Invest
1993;71:57– 64.
BRIEF REPORTS
263
Long-Term Safety and Efficacy of High-Dose
Atorvastatin Treatment in Patients With
Familial Hypercholesterolemia
Sanne van Wissen, MD, PhD, Tineke J. Smilde, MD, PhD, Mieke D. Trip, MD,
Anton F.H. Stalenhoef, MD, PhD, and John J.P. Kastelein, MD, PhD
In the 2-year Atorvastatin versus Simvastatin on Atherosclerosis Progression extension study, patients
with familial hypercholesterolemia who continued to
take atorvastatin 80 mg for an additional 2 years had
complete arrest of the progression of mean carotid
intima-media thickness (0.89 mm at the start vs 0.90
mm at the end of the study, p ⴝ 0.58). In contrast,
patients previously taking simvastatin 40 mg had
significant regression of intima-media thickness (0.95
mm at the start vs 0.92 mm at the end of the study,
p ⴝ 0.01). Therefore, both placebo- and statintreated patients with familial hypercholesterolemia
are best treated with high-dose atorvastatin, a therapeutic regimen that induces atherosclerosis regression and is safe and well tolerated over a 4-year
period. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:264 –266)
reviously, we demonstrated that high-dose atorvastatin therapy was effective in atherosclerosis
P
regression in the 2-year Atorvastatin versus Simvastatin on Atherosclerosis Progression (ASAP)
study.1 In this study, patients with familial hypercholesterolemia (FH) were treated with either atorvastatin 80 mg/day or simvastatin 40 mg/day. Patients with FH have sharply increased low-density
lipoprotein (LDL) cholesterol levels, resulting from
defects in the LDL receptor, which leads to an
increased risk for cardiovascular disease. Patients
with FH are therefore considered excellent models
for atherogenesis. The effect of statin intervention
on the atherosclerotic process was monitored in this
trial by measuring intima-media thickness (IMT), a
validated surrogate marker for future cardiovascular
end points.2 After completing the first 2-year period
of the ASAP trial, patients either continued with or
were switched to atorvastatin 80 mg/day in a 2-year
extension study. This provided us with the opportunity (1) to study change in IMT in patients previously randomized to simvastatin and (2) to obtain
long-term follow-up IMT data for patients origiFrom the Department of Vascular Medicine, Academic Medical Center, University of Amsterdam, Amsterdam, and Department of Medicine, Division of General Internal Medicine, University Medical Center
Nijmegen, Nijmegen, The Netherlands. This study was supported by
Pfizer, Capelle aan de Yssel, The Netherlands. Dr. van Wissen’s
address is: Academic Medical Center, Department of Vascular Medicine, Room F4 –159.2, Meibergdreef 9, 1105 AZ Amsterdam, The
Netherlands. E-mail: s.vanwissen@amc.uva.nl. Manuscript received
May 6, 2004; revised manuscript received and accepted September
7, 2004.
264
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
PhD,
nally randomized to high-dose atorvastatin. Finally,
this study provides the first report on 4-year safety
data of the highest registered dose of atorvastatin.
•••
Patients in the ASAP extension study were previously randomized to either atorvastatin 80 mg/day or
simvastatin 40 mg/day as previously described.1 In
short, participating patients (aged 30 to 70 years) were
diagnosed with FH and had LDL cholesterol levels
⬎4.5 mmol/L (174 mg/dl). After completion of the
controlled and randomized period of the ASAP study,
patients were invited to continue in a 2-year extension
study with atorvastatin 80 mg/day. The ethics committees approved the protocol and written informed
consent was obtained from every patient. The last visit
of the ASAP study was the first of the extension study.
Laboratory parameters were measured at 12 weeks
after the start of the study and every 24 weeks thereafter. At each visit, a brief physical examination and a
laboratory analysis was performed, compliance was
checked, and standard dietary instructions were given.
The ultrasound scanning protocol and its reproducibility have been described.1 Ultrasound examinations
were performed every year with a Biosound Phase-2
real-time scanner (BiosoundEsaote, Indianapolis, Indiana) equipped with a 10-MHz transducer. Three segments of 10 mm of the intima-media were scanned
bilaterally, ie., the distal portion of the common carotid
artery, the carotid bifurcation, and the proximal portion
of the internal carotid artery. Images were analyzed with
a semiautomatic program (Eurequa, TSA Company,
Meudon, France). The IMT readers were blinded to all
clinical information, including the original treatment to
which the patients were randomized.
Differences between treatment groups in IMT
change from baseline after 24 months and treatment
differences in the percent changes from baseline of
lipoprotein parameters were analyzed with analysis of
covariance. Analyses were done on an intent-to-treat
basis. The last observation carried forward was used
for patients who did not complete the study or who
had missing values. Statistical analyses were done
with SAS software (version 6.12, SAS Inc., Cary,
North Carolina).
Of the 280 patients who completed the ASAP study,
255 entered the extension study. Reasons for not entering
the extension study were desire to become pregnant (n ⫽
1), travel distance too great (n ⫽ 2), physical or mental
discomfort (n ⫽ 6), personal reasons (n ⫽ 5), alcohol
abuse (n ⫽ 7), dislike of atorvastatin (n ⫽ 1), too little
compliance (n ⫽ 2) and visual complaints (n ⫽ 1). Of
the 255 patients in the extension study, 30 patients dis0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.015
plained of muscle ache. In 7 patients
(3 in the atorvastatin group and 4 in
the group previously treated with
Atorvastatin
Simvastatin
simvastatin), aspertate aminotrans(n ⫽ 132)
(n ⫽ 123)
ferase and alanine transferase inStart
End
Start
End
creased to ⬎3 times the upper limit
of normal (⬎120 and 135 U/L, reTC (mg/dl)
215 ⫾ 42
223 ⫾ 46*
251 ⫾ 46
220 ⫾ 45*
spectively).
TG (mg/dl)
45 ⫾ 59
53 ⫾ 95*
54 ⫾ 69
45 ⫾ 56
HDL (mg/dl)
52 ⫾ 16
50 ⫾ 14*
51 ⫾ 13
51 ⫾ 12
Percent withdrawals due to adverse
LDL (mg/dl)
143 ⫾ 40
152 ⫾ 42*
177 ⫾ 45
150 ⫾ 43*
events was small for groups and not
significant: 5 patients (3.8%) in the
Values are expressed as mean ⫾ SD.
*p ⬍0.05 for percent change within 1 treatment arm.
atorvastatin group versus 3 patients
HDL ⫽ high-density lipoprotein cholesterol; TC ⫽ total cholesterol; TG ⫽ triglycerides.
(2.4%) in the simvastatin group.
The primary results of ASAP
have been previously published.1 In
short, in the atorvastatin 80-mg group, mean carotid
IMT decreased by 0.031 mm after 2 years (p ⫽
0.002); in contrast, in the simvastatin 40 mg group,
IMT increased by 0.036 mm (p ⬍0.001).1
In the extension study, patients taking atorvastatin
80 mg for an additional 2 years (4 years total) exhibited a complete arrest of the progression of mean
carotid IMT (0.89 mm at the start vs 0.90 mm at the
end of the extension period; p ⫽ 0.58), whereas, in
contrast, patients who were previously taking simvastatin 40 mg had significant regression of IMT after 2
years of treatment with atorvastatin 80 mg (0.95 mm
at the start vs 0.92 mm at the end of the extension
period; p ⫽ 0.01) (Figure 1). Mean carotid IMT between the 2 groups (0.90 vs 0.92 mm) no longer
differed at 4 years (p ⫽ 0.06). In a stepwise regression
FIGURE 1. Changes (mean ⴞ SE) in mean carotid IMT in the
analysis, the change in IMT over 4 years was only
ASAP study. Gray squares, atorvastatin; black triangles, simvarelated to baseline IMT, was negatively associated
statin.
with age (implying a greater increase in older patients)
and gender (implying a greater increase in men).
•••
continued (reasons were travel distance too great [n ⫽
We conclude that patients with FH are best treated
16] personal reasons [n ⫽ 8], increased transaminases
[n ⫽ 4], muscle ache [n ⫽ 1], and skin rash [n ⫽ 1]). with high-dose atorvastatin therapy, a therapeutic regPatient characteristics between both study groups were imen that induces regression of atherosclerosis and is
comparable with regard to demographic variables and safe and well tolerated over a period of ⱖ4 years.
The results of the ASAP are supported by the recent
vital signs.
Changes in lipid and lipoprotein levels are listed in outcome of the Arterial Biology for the Investigation of
Table 1. In patients previously treated with simvasta- the Treatment Effects of Reducing Cholesterol trial,
tin 40 mg, total cholesterol and LDL cholesterol de- which assessed the effects of atorvastatin 80 mg versus
creased significantly after the change to atorvastatin pravastatin 40 mg on IMT in a mixed hyperlipidemic
80 mg. In patients treated with atorvastatin 80 mg, population and also reported regression of carotid IMT in
there was a small but significant increase in total the atorvastatin group.3 In another recent study, the Reversal of Atherosclerosis with Aggressive Lipid Lowercholesterol, LDL cholesterol, and triglycerides.
The overall observed adverse event rates were ing trial, using intracoronary intravascular ultrasound in
comparable between groups, although they showed a patients with cardiovascular disease, atorvastatin 80 mg
trend toward being higher in the group previously also arrested progression of atherosclerosis in the corotreated with simvastatin. Muscle-related complaints nary arteries, whereas pravastatin 40 mg led to progreswere 9% versus 14% in the atorvastatin group and the sion of atherosclerotic vascular disease.4 Furthermore,
group previously treated with simvastatin, respec- and more importantly, the use of high dosages of atortively. The percentages for abdominal complaints vastatin decreased the number of major cardiovascular
were 18% and 29% for the atorvastatin and simvasta- events compared with moderate cholesterol reductions
tin groups, respectively. All these differences were not with pravastatin 40 mg in patients with a recent acute
coronary syndrome in the Pravastatin or Atorvastatin
statistically significant.
No significant differences with regard to laboratory Evaluation and Infection Therapy trial.5 However, folparameters were noticed. In 8 patients (4 in each low-up of these 3 last trials ran from 12 to approximately
group), creatine kinase increased to ⬎3 times the 24 months, whereas follow-up in the ASAP was 48
upper limit of normal (⬎540 U/L). No patient com- months.3–5
TABLE 1 Plasma Lipid Levels in Patients With Familial Hypercholesterolemia at
Start and End of Extension Study
BRIEF REPORTS
265
These results make it likely that the initial regression
and subsequent arrest of progression due to 80 mg of
atorvastatin will be maintained over protracted periods of
time. This is compatible with current guidelines.
We could not show further IMT regression in
patients who were previously randomized to atorvastatin. It could be hypothesized that the maximum
potential of statin treatment is reached in these FH
heterozygotes. If this is the case, additional interventions may be required to optimize the treatment
of dyslipidemia in this disorder. Such therapeutic
modalities would include even more potent LDL
lowering, increasing high-density lipoprotein cholesterol levels, or inhibition of very low-density
lipoprotein production. Cholesterol absorption inhibitors (e.g., ezetimibe) may lower LDL cholesterol levels by an additional 20%,6 cholesteryl ester
transfer protein inhibitors may increase high-density lipoprotein cholesterol by 30% to 40%,7 and
peroxisome proliferator-activated receptor agonists
decrease very low-density lipoprotein production
and increase apolipoprotein A-I synthesis.8
1. Smilde TJ, van Wissen S, Wollersheim H, Trip MD, Kastelein JJP, Stalenhoef
AFH. Effect of aggressive versus conventional lipid lowering on atherosclerosis
progression in familial hypercholesterolaemia (ASAP): a prospective, randomised, double-blind trial. Lancet 2001;357:577–581.
2. Lekakis JP, Papamichael CM, Cimponeriu AT, Stamatelopoulos KS, Papaioannou
TG, Kanakakis J, Alevizaki MK, Papapanagiotou A, Kalofoutis AT, Stamatelopoulos
SF. Atherosclerotic changes of extracoronary arteries are associated with the extent of
coronary atherosclerosis. Am J Cardiol 2000;85:949 –952.
3. Taylor AJ, Kent SM, Flaherty PJ, Coyle LC, Markwood TT, Vernalis MN.
ARBITER: Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol: a randomized trial comparing the effects of atorvastatin and
pravastatin on carotid intima medial thickness. Circulation 2002;106:2055–2060.
4. Nissen SE, Tuzcu EM, Schoenhagen P, Brown BG, Ganz P, Vogel RA, Crowe
T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria AN. Effect of intensive
compared with moderate lipid-lowering therapy on progression of coronary
atherosclerosis: a randomized controlled trial. JAMA 2004;291:1071–1080.
5. Cannon CP, Braunwald E, McCabe CH, Rader DJ, Rouleau JL, Belder R,
Joyal SV, Hill KA, Pfeffer MA, Skene AM. Intensive versus moderate lipid
lowering with statins after acute coronary syndromes. N Engl J Med 2004;
350:1495–1504.
6. Evans M, Roberts A, Rees A. The future direction of cholesterol-lowering
therapy. Curr Opin Lipidol 2002;13:663– 669.
7. de Grooth GJ, Kuivenhoven JA, Stalenhoef AF, de Graaf J, Zwinderman AH,
Posma JL, van Tol A, Kastelein JJ. Efficacy and safety of a novel cholesteryl ester
transfer protein inhibitor, JTT-705, in humans: a randomized phase II doseresponse study. Circulation 2002;105:2159 –2165.
8. Fruchart JC, Duriez P, Staels B. Peroxisome proliferator-activated receptoralpha activators regulate genes governing lipoprotein metabolism, vascular inflammation and atherosclerosis. Curr Opin Lipidol 1999;10:245–257.
Effect of Medical and Surgical Weight Loss on
Endothelial Vasomotor Function in Obese Patients
Noyan Gokce, MD, Joseph A. Vita, MD, Marie McDonnell, MD, Armour R. Forse,
Nawfal Istfan, MD, PhD, Maria Stoeckl, BA, Izabella Lipinska, PhD,
John F. Keaney, Jr., MD, and Caroline M. Apovian, MD
We prospectively examined brachial artery endothelial function using vascular ultrasound in 41 obese
subjects treated with medical or surgical (gastric bypass) weight loss interventions. Surgical intervention
produced greater weight loss and more pronounced
improvement in endothelial function than medical
treatment alone. Improved endothelial function with
weight loss correlated strongly with fasting glucose
but not with alteration in blood pressure, lipids, degree of weight loss, or plasma resistin concentrations.
These data demonstrate that weight loss in markedly
obese patients improves endothelial function and glycemic control that may represent important mecha-
From the Evans Department of Medicine, Cardiology and Endocrinology Sections, the Department of Surgery, Gastrointestinal Section, and
the Whitaker Cardiovascular Institute, Boston University School of
Medicine, Boston, Massachusetts. This work was supported by Grant
HL 74097 to Drs. Gokce, Forse, and Apovian from the National
Institutes of Health, Bethesda, Maryland. Dr. Gokce is the recipient of
a mentored patient-oriented research career transition Award
HL04425 from the National Institutes of Health, Bethesda, Maryland.
Dr. Gokce’s address is: Section of Cardiology, Boston Medical Center, C-818, 88 East Newton Street, Boston, Massachusetts 02118.
E-mail: Noyan.Gokce@bmc.org. Manuscript received June 2, 2004;
revised manuscript received and accepted September 8, 2004.
266
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MD,
nisms of the cardiovascular benefit associated with
weight reduction. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:266 –268)
everal recent studies have suggested that weight
reduction improves endothelial function in obese
S
subjects, although the importance of mode and extent
of weight loss remain largely unexplored.1–3 The purpose of this study was to determine the relative benefit
of marked weight loss produced by bariatric surgery
compared with more modest weight loss induced with
medical therapy and to investigate the mechanisms of
benefit.
•••
We enrolled consecutive obese patients (body mass
index [BMI] ⱖ30 kg/m2) receiving care at the Boston
Medical Center Nutrition and Weight Management
Center, which provides comprehensive dietary, medical, behavioral, and surgical treatments to promote
weight loss. All subjects provided written informed
consent and the study was approved by the Boston
Medical Center Institutional Review Board. Each subject made a visit at baseline and ⱖ3 months after
initiation of medical or surgical weight loss treatment.
During each visit, endothelium-dependent flow-mediated dilation (FMD) and endothelium-independent,
nitroglycerin (NTG)-mediated dilation of the brachial
artery were examined in the upper arm using an es0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.016
loss were determined by the Boston
Medical Center clinical laboratory.
Medical Therapy
Surgical Therapy
Serum insulin levels were measured
(n ⫽ 17)
(n ⫽ 24)
p Value
using chemiluminometric methods
(Quest Diagnostics, Cambridge,
Age (yrs)
44 ⫾ 9
44 ⫾ 10
0.89
Body weight (kg)
112 ⫾ 32
142 ⫾ 26
0.002*
Massachusetts), and insulin resis42 ⫾ 10
50 ⫾ 8
0.01*
Body mass index (kg/m2)
tance was estimated using the previHemoglobin A1C (%)
5.8 ⫾ 0.8
6.1 ⫾ 1.6
0.60
ously validated homeostasis model
Women
82%
63%
0.30
assessment (HOMA).8 Plasma resisDiabetes mellitus*
24%
29%
0.74
Hypertension†
41%
50%
0.81
tin concentrations were measured usHypercholesterolemia
35%
50%
0.54
ing a commercially available enTobacco use
24%
8%
0.21
zyme-linked immunosorbent assay
Coronary artery disease
6%
4%
0.99
kit (ALPCO Diagnostics, Windham,
Lipid lowering therapy
24%
21%
0.99
New Hampshire).9
ACE inhibitor or ARB therapy
18%
29%
0.48
All data are presented as mean ⫾
*Diabetes mellitus was defined by history or fasting glucose ⬎126 mg/dl.
†
SD,
unless otherwise indicated. BaseHypertension was defined by history or blood pressure ⬎140/90 mm Hg.
line clinical characteristics were comACE ⫽ angiotensin-converting enzyme; ARB ⫽ angiotensin-receptor blocker.
pared using unpaired t tests, and the
chi-square or Fisher’s exact test as appropriate. The effect of weight loss on
vascular, hemodynamic, and metabolic parameters
within each group was examined using paired t tests.
Treatment effects between different groups were examined using 2-way repeated measures analysis of variance
(ANOVA) with post hoc Tukey’s comparison. Correlations were examined using linear regression analysis.
We prospectively followed a total of 41 subjects who
achieved ⱖ5% weight reduction with dietary, medical,
and behavioral intervention (medical therapy group, n ⫽
17) or gastric bypass surgery (surgical therapy group, n
⫽ 24) during the investigation period. Baseline clinical
characteristics are listed in Table 1. As expected, the
surgical treatment group had higher weight and BMI;
otherwise, there were no significant differences between
the 2 groups. Baseline brachial artery FMD was similar
in the medical (7.3 ⫾ 5%) and surgical (6.8 ⫾ 4.0%)
FIGURE 1. Effect of medical and surgical (gastric bypass)
weight loss groups (p ⫽ 0.75) (Figure 1). Patients in the
weight loss on brachial artery FMD in obese patients (n ⴝ
medical group sustained an average 10% (range 5% to
41). Examinations were performed at baseline (black bars)
18%) decrease in mean body weight and BMI, from 112
and after medical and surgical weight loss interventions
⫾ 32 to 101 ⫾ 29 kg (p ⬍0.001) and from 42 ⫾ 10 to
(white bars). Surgical gastric bypass was more effective in
improving brachial dilation than medical therapy (p ⴝ 0.004).
38 ⫾ 10 kg/m2 (p ⬍0.001), respectively. This effect was
Data are presented as mean ⴞ SEM.
associated with an increase in FMD to 9.6 ⫾ 4.3% (p ⫽
0.08). As expected, surgery was more effective in protablished noninvasive, standardized, blinded method moting weight loss than medical or dietary interventions
of vascular ultrasound as previously described.4 After alone. Bariatric surgery produced an average 25% (range
the baseline visit, all subjects initiated a clinically 9% to 49%) decrease in mean body weight (142 ⫾ 26 to
23 kg, p ⬍0.001) and BMI (50 ⫾ 8 to 37 ⫾ 7
prescribed weight loss program that included thera- 107 ⫾
2
kg/m
,
p
⬍0.001). Gastric bypass surgery was associated
peutic medical and/or surgical intervention. Medical
with
a
significant
improvement in brachial artery FMD,
therapy consisted of a comprehensive dietary, behavioral, and educational treatment regimen based on to 10.2% ⫾ 4.0% (p ⫽ 0.006 compared with baseline),
National Heart, Lung, and Blood Institute guide- and was more effective than medical therapy in improvlines.5,6 Increasing physical activity was encouraged, ing endothelial function (p ⫽ 0.004 by ANOVA) (Figure
but specifically tailored exercise regimens were not 1). In a subset of patients who underwent examination of
prescribed. We also followed patients with clinically non– endothelium-dependent dilator function, weight
severe obesity defined as BMI ⱖ40 or ⱖ35 kg/m2 modification produced no significant effect on NTGwith co-morbid conditions who underwent gastric by- mediated dilation (12.6 ⫾ 5.6% vs 11 ⫾ 6.8%, p ⫽
pass (bariatric) surgery as weight reduction treat- 0.62). In 22 separate subjects with similar clinical demoment.7 The standard operation involved Roux-en-Y graphics (data not shown) who failed to lose weight,
gastric bypass surgery with an isolated gastric pouch endothelial function remained unchanged (FMD 7.5 ⫾
4.3% at baseline vs 7.3 ⫾ 4.9% at follow-up, p ⫽ 0.87).
of ⬍2 ounces and a 100-cm distal bowel limb.
Biochemical analyses at baseline and after weight Correlation between percent weight reduction and FMD
TABLE 1 Baseline Clinical Characteristics
BRIEF REPORTS
267
was ⫺0.17 (p ⫽ 0.30); percent change in BMI with
FMD was ⫺0.15 (p ⫽ 0.35).
To investigate the potential mechanisms that could
account for the beneficial effects of weight loss on
endothelial function, we examined blood pressure,
lipid profiles, fasting glucose and insulin, and HOMA
before and after weight loss. Blood pressure did not
change significantly with weight loss in this group of
patients. Gastric bypass surgery resulted in significant
decreases in total, low-density cholesterol, and highdensity lipoprotein cholesterol levels that were not
observed with more modest medical weight loss intervention. In contrast, both weight reduction strategies significantly improved glycemic indexes. In the
medical treatment group, fasting glucose and insulin
decreased from 104 ⫾ 24 mg/dl and 16 ⫾ 10 ␮IU/ml
at baseline, to 93 ⫾ 16 mg/dl and 11 ⫾ 5 ␮IU/ml after
weight loss, respectively (p ⬍0.05). Similarly, HOMA
decreased from 4.5 ⫾ 3.7 to 2.6 ⫾ 1.5 (p ⬍0.05) with
medical weight loss. Surgical treatment was more
effective in lowering glucose (119 ⫾ 39 to 90 ⫾ 18
mg/dl), insulin (22 ⫾ 14 to 7 ⫾ 5 ␮IU/ml) and
HOMA (6.5 ⫾ 4.7 to 1.6 ⫾ 1.5) than medical or
dietary interventions alone (p ⬍0.001 for all variables
by ANOVA). The only significant correlation for improved endothelial function was fasting glucose (r ⫽
⫺0.46, p ⫽ 0.004). We observed no significant differences in resistin levels at baseline or after weight
loss (4.3 ⫾ 3.6 vs 4.8 ⫾ 4.1 ng/ml, p ⫽ 0.29, n ⫽ 18).
•••
This prospective study demonstrated that endothelium-dependent flow-mediated vasodilation is impaired in obese patients and improves significantly
with weight loss intervention. Gastric bypass surgery
was associated with more dramatic weight loss, more
pronounced improvement in endothelial function, and
greater metabolic changes than medical therapy alone.
Improved vascular function correlated significantly
with reduced glucose concentration but was not linked
to magnitude of weight reduction per se. Thus, the
study suggests that altered glycemic status associated
with therapeutic weight loss may be an important
determinant of vascular homeostasis in obese patients.
Several previous studies have examined endothelial
function in the setting of dietary and behavioral weight
reduction strategies, but to our knowledge, the present
investigation is the first to incorporate surgically treated
patients. In 2 previous studies, the combination of medical weight loss and regular exercise significantly improved vascular reactivity1,2; however, the benefit may
have been partly due to a therapeutic vascular effect of
standardized exercise regimens.10 The mechanisms for
the effect of weight reduction on endothelial function are
incompletely understood and likely multifactorial, relating to changes in glucose tolerance, lipid profiles, and
hypertension.5,11 In agreement with 2 previous studies,1,3
our study suggests that an alteration in glycemic status
after weight loss may primarily account for the benefits
on endothelial function and extends these findings to
gastric bypass patients who sustained dramatic reductions in body weight. Our present investigation failed to
provide evidence for a role of resistin, an adipocyte268 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
derived peptide hormone implicated in mechanisms of
obesity-related insulin resistance12 as a key modulator of
endothelial function. Our finding that surgical treatment
produced greater weight loss and improved metabolic
parameters and endothelial function to a greater extent
than dietary interventions alone may be clinically relevant. It remains unknown whether improved endothelial
function is sustained in the long term in these patients
who are otherwise still considered obese and whether
further weight reduction will translate into vascular benefit.
Our present study has several limitations. The
study was relatively small and not randomized. However, it would not be ethical to randomize patients to
medical therapy when surgical treatment is clinically
indicated. Only a subset of patients received NTG;
thus, we cannot definitively exclude a change in endothelium-independent dilation, although no such effect was seen in previous similar studies.2,3 Last, our
data suggest that surgical therapy was predominantly
responsible for the improvement in endothelial function. We stress that dietary and medical interventions
are critical components of effective weight loss strategies and that the strong trend for improved vascular
function with medical therapy may have been significant in a larger study population.
In summary, this prospective study demonstrated
that weight loss significantly improves endothelial
function in obese patients. Reversal of endothelial
dysfunction may reduce the association between obesity and cardiovascular risk and serve as an important
therapeutic target for weight loss strategies.
1. Sciacqua A, Candigliota M, Ceravolo R, Scozzafava A, Sinopoli F, Corsonello
A, Sesti G, Perticone F. Weight loss in combination with physical activity
improves endothelial dysfunction in human obesity. Diabetes Care 2003;26:
1673–1678.
2. Hamdy O, Ledbury S, Mullooly C, Jarema C, Porter S, Ovalle K, Moussa A,
Caselli A, Caballero AE, Economides PA, Veves A, Horton ES. Lifestyle
modification improves endothelial function in obese subjects with the insulin
resistance syndrome. Diabetes Care 2003;26:2119 –2125.
3. Raitakari M, Ilvonen T, Ahotupa M, Lehtimaki T, Harmoinen A, Suominen P,
Elo J, Hartiala J, Raitakari OT. Weight reduction with very-low-caloric diet and
endothelial function in overweight adults: role of plasma glucose. Arterioscler
Thromb Vasc Biol 2004;24:124 –128.
4. Vita JA. Nitric oxide-dependent vasodilation in human subjects. Methods
Enzymol 2002;359:186 –200.
5. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults, NHLBI, 1998. Available at: http://www.nhlbi.nih.
gov/guidelines/obesity/ob_gdlns/pdf. Accessed May 5, 2004.
6. Kiernan M, Winkleby MA. Identifying patients for weight-loss treatment: an
empirical evaluation of the NHLBI obesity education initiative expert panel
treatment recommendations. Arch Intern Med 2000;160:2169 –2176.
7. Brolin RE. Bariatric surgery and long-term control of morbid obesity. JAMA
2002;288:2793–2796.
8. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC.
Homeostasis model assessment: insulin resistance and beta-cell function from fasting
plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412– 419.
9. Lee JH, Chan JL, Yiannakouris N, Kontogianni M, Estrada E, Seip R, Orlova
C, Mantzoros CS. Circulating resistin levels are not associated with obesity or
insulin resistance in humans and are not regulated by fasting or leptin administration: cross-sectional and interventional studies in normal, insulin-resistant, and
diabetic subjects. J Clin Endocrinol Metab 2003;88:4848 – 4856.
10. Vita JA, Keaney JF Jr. Exercise—toning up the endothelium? N Engl J Med
2000;342:503–505.
11. Schauer PR, Burguera B, Ikramuddin S, Cottam D, Gourash W, Hamad G,
Eid GM, Mattar S, Ramanathan R, Barinas-Mitchel E, et al. Effect of laparoscopic Roux-en Y gastric bypass on type 2 diabetes mellitus. Ann Surg 2003;
238:467– 484.
12. Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright CM, Patel
HR, Ahima RS, Lazar MA. The hormone resistin links obesity to diabetes. Nature
2001;409:307–312.
JANUARY 15, 2005
Prevalence of Adequate Control of Increased Serum LowDensity Lipoprotein Cholesterol in Self-Pay or Medicare
Patients Versus Medicaid or Private Insurance Patients
Followed in a University General Medicine Clinic
Raja Varma,
MD,
Wilbert S. Aronow, MD, Glenn Gandelman,
Christopher Zammit, BS
In a study of 514 patients with increased serum lowdensity lipoprotein (LDL) cholesterol followed in a general medicine clinic at a university hospital, the serum
LDL cholesterol in patients with coronary heart disease
(CHD), other atherosclerotic vascular disease, or diabetes mellitus was <100 mg/dl in 219 of 276 patients
(79%) with Medicaid or private insurance and in 28 of
67 self-pay or Medicare patients (42%) without pharmaceutical coverage (p <0.001). The serum LDL cholesterol was <130 mg/dl in patients with 2ⴙ risk factor
and a 10-year risk for CHD of <20% or <160 mg/dl in
patients with a 0 to 1 risk factor and a 10-year risk for
CHD of <10% in 54 of 141 patients (38%) with Medicaid or private insurance and in 5 of 30 self-pay or
Medicare patients (17%) (p <0.025). 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:269 –270)
umerous studies have demonstrated that treatment
of hypercholesterolemia in high-risk persons with
N
statins reduces cardiovascular morbidity and mortality.
1
The National Cholesterol Education Program Expert
Panel on Detection, Evaluation, and Treatment of High
Blood Cholesterol in Adults (NCEP III) guidelines recommend that the serum low-density lipoprotein (LDL)
cholesterol be reduced to ⬍100 mg/dl in persons with
coronary heart disease (CHD), other clinical forms of
atherosclerotic vascular disease, diabetes mellitus, and
with 2⫹ risk factors that confer a 10-year risk for CHD
of ⬎20%.2 These guidelines also recommend that the
serum LDL cholesterol be reduced to ⬍130 mg/dl in
persons with 2⫹ risk factors and a 10-year risk for CHD
of ⱕ20%, and to ⬍160 mg/dl in persons with a 0 to 1
risk factor and a 10-year risk for CHD of ⬍10%.2 This
study reports on the prevalence of adequate serum LDL
cholesterol control in 514 unselected patients with an
increased serum LDL cholesterol followed in the general
medicine clinic at a university hospital between July
2003 and June 2004.
•••
Charts of 514 unselected patients with increased
serum LDL cholesterol followed in the general medicine clinic at Westchester Medical Center/New York
From the Department of Medicine, Cardiology Division, Westchester
Medical Center/New York Medical College, Valhalla, New York.
Dr. Aronow’s address is: Cardiology Division, New York Medical
College, Macy Pavilion, Room 138, Valhalla, New York 10595.
E-mail: WSAronow@aol.com. Manuscript received July 27, 2004;
revised manuscript received and accepted September 7, 2004.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MD, MPH,
and
Medical College between July 2003 and June 2004
were reviewed according to a protocol designed by
WSA. The 514 patients included 281 women and 233
men (mean age 57 ⫾ 13 years, range 18 to 90). Of the
514 patients, 172 (34%) were white, 152 (30%) were
Hispanic, 133 (26%) were African-American, and 57
(11%) were of other races. Of the 514 patients, 343
(67%) had CHD, other clinical forms of atherosclerotic vascular disease, or diabetes mellitus (target
LDL cholesterol goal of ⬍100 mg/dl), 87 (17%) had
2⫹ coronary risk factors with a 10-year risk for CHD
of ⱕ20% (target LDL cholesterol goal of ⬍130 mg/
dl), and 84 (16%) had a 0 to 1 coronary risk factor
with a 10-year risk for CHD of ⬍10% (target LDL
cholesterol goal of ⬍160 mg/dl).
NCEP III guidelines were used to assess the adequacy of reaching target serum LDL cholesterol treatment goals using the last serum LDL cholesterol level
recorded in the chart. Student’s t tests were used to
analyze continuous variables and chi-square tests to
analyze dichotomous variables.
Table 1 shows the prevalence of adequate serum
LDL cholesterol control in patients who had Medicaid
or private insurance versus patients who did not have
pharmaceutical coverage because they were self-pay
or Medicare patients. Table 1 also lists the levels of
statistical significance. Table 2 shows the prevalence
of adequate control of serum LDL cholesterol in patients aged ⱖ65 versus ⬍65 years, in men versus
women, and in whites, Hispanics, African-Americans,
and patients of other races. No significant differences
were observed between these groups.
Of the 417 patients who had Medicaid or private
insurance, 335 (80%) were receiving statins versus 46
of 97 patients (47%) who were self-pay or Medicare
patients (p ⬍0.001). Table 3 shows the prevalence of
adequate control of serum LDL cholesterol in patients
treated with different statins. No significant differences were observed.
Two hundred ninety-five of 361 patients (82%)
treated with atorvastatin, simvastatin, or rosuvastatin
and 11 of 20 patients (55%) treated with pravastatin,
lovastatin, or fluvastatin achieved adequate control of
serum LDL cholesterol (p ⬍0.005). One of 82 Medicaid or private insurance patients (1%) not treated
with a statin was treated with ezetimide and none of
51 self-pay or Medicare patients (0%) not treated with
a statin was treated with other lipid-lowering medication. The 1 patient treated with ezetimide did not have
adequate serum LDL cholesterol control.
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.095
269
population involving 591 centers, 76
of 2,856 patients (3%) with proved or
probable CHD had a serum LDL choSelf-pay or
lesterol ⬍100 mg/dl.7 In the Lipid
Medicare
Treatment Assessment Project, 37% of
28/67 (42%)*
2,285 high-risk patients and 18% of
1,460 patients with CHD achieved adequate serum LDL cholesterol con†
trol.8 In the Heart and Estrogen/Pro5/30 (17%)
gestin Replacement Study, 262 of
2,752 postmenopausal women (10%)
with CHD had a serum LDL cholesterol of ⬍100 mg/dl.9
In the present study, the prevalence of a serum LDL cholesterol
⬍100 mg/dl in patients with CHD,
other clinical forms of atherosclerotic vascular disease, or diabetes mellitus was 79% in patients with
Medicaid or private insurance versus 42% in self-pay
or Medicare patients (p ⬍0.001). In patients without
CHD, other clinical forms of atherosclerotic vascular
disease or diabetes mellitus, adequate serum LDL
cholesterol control according to NCEP III guidelines
was achieved in 38% of Medicaid or private insurance
patients versus 17% of self-pay or Medicare patients
(p ⬍0.025). Adequate serum LDL cholesterol control
was not significantly different in elderly versus
younger patients, in men versus women, and in
whites, Hispanics, African-Americans, and patients of
other races.
Data from the present study show a significantly
lower prevalence of adequate serum LDL cholesterol
control in patients who have no pharmaceutical coverage
for medications prescribed by their physician. This problem needs to be addressed if we are to decrease the great
amount of cardiovascular morbidity and mortality due to
inadequate control of hypercholesterolemia.
TABLE 1 Prevalence of Adequate Serum Low-density Lipoprotein (LDL) Cholesterol
Control in 514 Patients With Hypercholesterolemia
Medicaid or Private
Insurance
Variable
LDL cholesterol ⬍100 mg/dl in
patients with CHD, other
atherosclerotic vascular disease, or
diabetes mellitus
LDL cholesterol ⬍130 mg/dl in
patients with 2⫹ risk factors and
10-year risk for CHD ⱕ20% or
⬍160 mg/dl in patients with 0–1
risk factor and 10-year risk for CHD ⬍10%
219/276 (79%)
54/141 (38%)
*p ⬍0.001;†p ⬍0.025.
TABLE 2 Prevalence of Adequate Serum Low-density
Lipoprotein (LDL) Cholesterol Control in Persons ⱖ65 Versus
⬍65 Years, in Men Versus Women, and in Whites,
Hispanics, African-Americans, and Persons of Other Races
Variable
No.
Age ⱖ65 yrs
Age ⬍65 yrs
Women
Men
Whites
Hispanics
African-Americans
Other races
190
324
281
233
172
152
133
57
Adequacy of Serum LDL
Cholesterol Control
115
191
170
136
105
88
80
33
(61%)
(59%)
(60%)
(58%)
(61%)
(58%)
(60%)
(58%)
TABLE 3 Prevalence of Adequate Serum Low-density
Lipoprotein (LDL) Cholesterol Control in Persons Taking
Different Statins
Statin
Atorvastatin
Simvastatin
Pravastatin
Lovastatin
Fluvastatin
Rosuvastatin
No. of
Patients
Adequacy of Serum LDL
Cholesterol Control
201
159
11
7
2
1
171
123
6
4
1
1
(85%)
(77%)
(55%)
(57%)
(50%)
(100%)
•••
Harnick et al3 reported data from 270 patients with
CHD seen at the Mount Sinai Medical Center (New
York, New York), either in the cardiology clinic (n ⫽
131) or in the cardiology private practice suite (n ⫽ 139).
Of the 270 patients, 26 (22%) had a serum LDL cholesterol ⱕ100 mg/dl.3 In a primary care outpatient clinic in
Southern Florida, 16 of 118 patients (14%) needing
lipid-lowering therapy ever had a cholesterol-lowering
drug prescribed during a 5-year study period.4
In 27 managed care plans, 1,072 of 3,018 patients
(36%) with CHD had a serum LDL cholesterol of ⱕ100
mg/dl.5 In the Duke University Medical Center Outpatient Clinic, 348 of 518 patients with CHD (68%) had
measurements of their serum LDL cholesterol.6 Of the
268 patients found to be at risk, 71 (26%) had a serum
LDL cholesterol of ⱕ100 mg/dl.6 In a German outpatient
270 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
1. Aronow WS. Should the National Cholesterol Education Program guidelines
be changed for persons at high risk for cardiovascular events? Preventive Cardiol
2004;7:71–72.
2. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and
Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III).
JAMA 2001;285:2486 –2497.
3. Harnick DJ, Cohen JL, Schechter CB, Fuster V, Smith DA. Effects of practice
setting on quality of lipid-lowering management in patients with coronary artery
disease. Am J Cardiol 1998;81:1416 –1420.
4. Lai LL, Poblet M, Bello C. Are patients with hyperlipidemia being treated?
Investigation of cholesterol treatment practices in an HMO primate care setting.
Southern Med J 2000;93:283–286.
5. Latts LM. Assessing the results: phase 1 hyperlipidemia outcomes in 27 health
plans. Am J Med 2001;110(suppl 6A):17S–23S.
6. Higgins PDR, Russo C, Scheurer M, Duvall WL. How well do we treat
elevated LDL cholesterol? Results from a university residents’ clinic. NC Med J
2002;63:247–252.
7. Ruof J, Klein G, Marz W, Wollschlager H, Neiss A, Wehling M. Lipidlowering medication for secondary prevention of coronary heart disease in a
German outpatient population: the gap between treatment guidelines and real life
treatment patterns. Preventive Med 2002;35:48 –53.
8. Pearson TA, Laurora I, Chu H, Kafonek S. The Lipid Treatment Assessment
Project (L-TAP). A multicenter survey to evaluate the percentages of dyslipidemic patients receiving lipid-lowering therapy and achieving low-density lipoprotein cholesterol goals. Arch Intern Med 2000;160:459 – 467.
9. Schrott HG, Bittner V, Vittinghoff E, Herrington DM, Hulley S, for the HERS
Research Group. Adherence to National Cholesterol Education Program Treatment goals in postmenopausal women with heart disease. The Heart and Estrogen/Progestin Replacement Study. JAMA 1997;277:1281–1286.
JANUARY 15, 2005
Clinical Variables Predicting Inappropriate Use of
Implantable Cardioverter-Defibrillator in Patients
With Coronary Heart Disease or Nonischemic
Dilated Cardiomyopathy
Dominic A.M.J. Theuns, MSc, A. Peter J. Klootwijk, MD, PhD,
Maarten L. Simoons, MD, PhD, and Luc J. Jordaens, MD, PhD
Inappropriate therapy is a common clinical problem
in recipients of implantable cardioverter-defibrillators
(ICDs). The present study evaluated whether clinical
characteristics could predict inappropriate ICD therapy due to atrial tachyarrhythmias in a series of 260
patients. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:271–274)
he implantable cardioverter-defibrillator (ICD) has
become the standard therapy for life-threatening
T
ventricular tachyarrhythmias. Despite the accuracy
1–3
and effectiveness in the diagnosis and treatment of
ventricular tachyarrhythmias, a substantial proportion
of patients with ICDs experience inappropriate interventions. The reported incidence ranges from 8% to
40%.4 –7 Several studies have investigated clinical risk
predictors for ventricular arrhythmia recurrence.8 –10
In contrast, clinical risk predictors for inappropriate
ICD use have not been investigated. This study examined the variables that may predict which patients are
more likely to receive inappropriate therapy.
•••
The study population consisted of 326 consecutive
patients who underwent first transvenous implantation at the Erasmus Medical Center (Rotterdam, The
Netherlands). Of these, 57 patients were excluded
because of participation in a prospective, randomized
study of single- and dual-chamber detection algorithms. Another 9 were excluded because of the presence of hypertrophic cardiomyopathy. Thus, 260 patients were eligible for analysis. The patients were
assigned to ICD therapy because of a history of cardiac arrest, spontaneous sustained ventricular tachycardia (VT), or nonsustained VT with subsequent inducible sustained VT. Data in the ICD registry are
updated prospectively after each clinic visit.
The prospectively collected clinical and functional
variables for each patient include age, gender, the
presence of coronary artery disease (including myocardial infarction and cardiomyopathy), the left ventricular ejection fraction (LVEF) as determined by
nuclear isotopes, index arrhythmia, history of atrial
From the Department of Cardiology, Thoraxcenter, Erasmus Medical
Center, Rotterdam, The Netherlands. Dr. Theuns’s address is: Erasmus
Medical Center, Department of Electrophysiology, Bd402, Dr. Molewaterplein 40, 3015 GD Rotterdam, The Netherlands. E-mail:
d.theuns@erasmusmc.nl. Manuscript received June 15, 2004; revised
manuscript received and accepted September 20, 2004.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
tachyarrhythmias documented in the clinical file, and
pharmacologic treatment.
The implanted devices were manufactured by
Biotronik (Phylax AV, Tachos DR, and Belos VR-T;
Biotronik GmbH & Company, Berlin, Germany),
ELA Medical (Defender IV and Alto DR; ELA Medical, Paris, France), Guidant (Mini IV, Contak CD,
Renewal I, and Renewal II; Guidant Corporation, St.
Paul, Minnesota), and Medtronic (7227, 7250, 7271,
and 7272; Medtronic, Inc., Minneapolis, Minnesota).
The tachycardia detection rate was programmed according to the clinical presentation of each patient. For
all patients, the detection enhancements were activated immediately after ICD implantation. In singleand dual-chamber devices, the stability criterion was
programmed at 40 to 50 ms, and the onset criterion
was programmed at 15% to 20%. In all dual-chamber
devices, the respective dual-chamber detection algorithms were activated.
Follow-up began at ICD implantation. At every
follow-up visit (at 3-month intervals) or every visit
prompted by ICD therapy, all stored data of tachyarrhythmia episodes were collected. Two independent
researchers reviewed the stored electrocardiograms. In
case of disagreement between the 2 reviewers about
the stored electrocardiograms, a third reviewer was
consulted and made a decision. For each episode, the
date, type, morphology (monomorphic or polymorphic), and mean cycle length (CL) of the tachyarrhythmia and the type and outcome of delivered ICD therapy were recorded. A ventricular tachyarrhythmia was
defined as an event with a sudden increase in rate
combined with a change in electrocardiographic morphology from the baseline rhythm. If an atrial electrocardiogram was present, the presence of atrioventricular dissociation was used to classify a ventricular
tachyarrhythmia. Therapy delivered for atrial arrhythmias (including atrial fibrillation, atrial flutter, atrial
tachycardia, and sinus tachycardia) was defined as
inappropriate.
Continuous variables were evaluated using Student’s t test or analysis of variance. The chi-square
test was used for the analysis of categorical variables.
The actuarial event-free rates from ventricular and
atrial tachyarrhythmias triggering ICD therapy were
calculated according to the Kaplan-Meier method.
Differences between pairs of actuarial curves were
tested by the log-rank test. Relative risks expressed as
hazard ratios with 95% confidence intervals (CIs)
were based on a Cox proportional-hazards model.
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.017
271
TABLE 1 Patients’ Clinical Characteristics (n ⫽ 260)
Characteristic
Men
Age (yrs)
LVEF (%)
Underlying cardiac disease
Coronary artery disease
Dilated cardiomyopathy
History of atrial tachyarrhythmias
Index arrhythmia
Ventricular fibrillation
VT
Nonsustained VT
Pharmacologic treatment at discharge
Amiodarone
␤ blockers
Digoxin
ACE inhibitor
Diuretic
Lipid-lowering drug
Value
216 (83%)
60 ⫾ 13
31 ⫾ 14
184 (71%)
61 (24%)
79 (29%)
78 (30%)
125 (48%)
57 (22%)
97
120
57
189
158
129
(37%)
(46%)
(23%)
(73%)
(61%)
(50%)
ACE ⫽ angiotensin-converting enzyme.
Covariates previously identified to be independently
associated with the occurrence of inappropriate ICD
therapy were used in the multivariate model. A
2-tailed p value ⬍0.05 was considered significant.
The clinical characteristics are listed in Table 1.
Coronary artery disease was present in 70% of the
patients, nonischemic cardiomyopathy (excluding hypertrophic cardiomyopathy) in 30%, with dilation in
21%. Twenty-four percent received single-chamber
devices, 53% dual-chamber devices, and 23% dualchamber devices with cardiac resynchronization capability. The programmed mean detection CL of the VT
zone was 377 ⫾ 52 ms.
During a mean follow-up of 22 ⫾ 16 months (range
1 to 60), 107 patients (41%) experienced ⱖ1 episode of
sustained ventricular tachyarrhythmia, triggering ICD
therapy. The actuarial event-free rates for ventricular
tachyarrhythmias were 66.0%, 55.8%, and 45.0% at 1, 2,
and 4 years, respectively. The mean CL of monomorphic
VT was 333 ⫾ 62 ms; for polymorphic VT or ventricular
fibrillation, it was 223 ⫾ 26 ms.
A total of 37 patients (14%) experienced inappropriate ICD therapy due to atrial tachyarrhythmias. The
actuarial event-free rates for inappropriate therapy
were 87.0%, 83.6, and 80.8%, at 1, 2, and 4 years,
respectively (Figure 1). Nineteen patients experienced
inappropriate therapy for atrial fibrillation at least
once, and 18 patients received inappropriate therapy
for sinus or atrial tachycardia. There was no significant difference between actuarial event-free rates for
inappropriate therapy triggered by atrial fibrillation
(94.4% and 88.8%) or atrial or sinus tachycardia
(92.7% and 91.3%) at 1 and 4 years, respectively. The
mean ventricular CL during atrial fibrillation was 319
⫾ 44 ms (range 260 to 400); for atrial or sinus tachycardia, the mean ventricular CL was 374 ⫾ 48 ms
(range 300 to 480).
Clinical variables for patients with and without
inappropriate device therapy are presented in Table 2.
Age, gender, LVEF, underlying cardiac disease, and
272 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
FIGURE 1. Actuarial event rates for inappropriate device therapy.
pharmacologic treatment did not differ between those
with and without inappropriate device therapy. The
incidence of inappropriate device therapy was greater
after a history of atrial tachyarrhythmias (p ⫽ 0.003).
Additionally, inappropriate ICD therapy was noted
more frequently in patients who received appropriate
device therapy (p ⫽ 0.001).
To evaluate clinical predictors of inappropriate device therapy, variables were entered in a Cox proportional-hazards model (age, pharmacologic treatment,
type of ICD, LVEF, coronary artery disease, cardiomyopathy, a history of atrial tachyarrhythmias, and
recurrent VT). This analysis revealed a history of
atrial tachyarrhythmias and recurrent VT with a CL
ⱖ350 ms triggering device therapy as independent
clinical predictors of inappropriate ICD therapy. The
relative risk was 2.4 (95% CI 1.2 to 4.8, p ⫽ 0.01) for
a history of atrial tachyarrhythmias. This was supported by lower actuarial event-free rates for inappropriate device therapy for patients with a history of
atrial tachyarrhythmias compared with patients without such a history (62.5% vs 88.2% at 4 years, p ⫽
0.002; Figure 2).
The relative risk increased to 3.1 if patients had
recurrent VT with a CL ⱖ350 ms that triggered device
therapy (95% CI 1.5 to 6.3, p ⫽ 0.002). To address the
question of whether long interval programming explained this greater risk, we analyzed a detection interval of ⱖ350 ms as a tachycardia criterion. In multivariate analysis, this detection interval was not
identified as an independent predictor for inappropriate therapy, with a relative risk of 2.2 (95% CI 0.6 to
7.5, p ⫽ 0.21). Although not significant, the programmed detection CL of the VT zone tended to be
shorter in patients with inappropriate therapy compared with those without inappropriate therapy (386 ⫾
46 vs 371 ⫾ 44 ms, p ⫽ 0.06). Proportionally, patients
with dual-chamber devices experienced more inappropriate therapy compared with those with single-chamber devices (6% vs 17%, p ⬍0.05).
•••
The present study evaluated whether clinical characteristics could predict inappropriate ICD therapy
JANUARY 15, 2005
tion for atrial tachyarrhythmias. Given
the epidemiology of atrial fibrillation,
No Inappropriate ICD
atrial tachyarrhythmias are common in
Therapy (n ⫽ 223)
p Value
ICD recipients, of whom most have
structural heart disease.13 More re186 (83%)
NS
60 ⫾ 13
NS
cently, it has been reported that a his31 ⫾ 14
NS
tory of paroxysmal atrial fibrillation
predicted a higher recurrence rate of
159 (71%)
NS
atrial tachyarrhythmias.14
52 (23%)
NS
57 (26%)
0.003
More notable is the association of
recurrent VT with a CL ⱖ350 ms with
66 (30%)
NS
an increased risk for inappropriate
106 (47%)
NS
therapy. The association between ven51 (23%)
NS
tricular tachyarrhythmias and paroxys85 (38%)
NS
mal atrial tachyarrhythmias in ICD re101 (45%)
NS
cipients has been established in
46 (21%)
NS
previous studies.15–17 Slow ventricular
159 (74%)
NS
tachyarrhythmias were associated with
135 (63%)
NS
112 (52%)
NS
left ventricular dysfunction (LVEF
⬍40%) and class III antiarrhythmic
drug therapy.18 Another aspect to be
addressed is the programmed detection
interval. In case of slow ventricular
tachyarrhythmias, an overlap with the ventricular rate of
atrial tachyarrhythmias is present. The programmed detection interval was not identified in multivariate analysis
as an independent predictor.
Whether device selection should depend on the
knowledge of a history of atrial tachyarrhythmias is an
open question, because inappropriate therapy occurs
equally in patients with single- and dual-chamber devices. The addition of an atrial lead might improve the
specificity of arrhythmia discrimination but introduces
potential surgical and technical problems. This should be
balanced against the potential minor advantages, such as
the presence of slow ventricular tachyarrhythmias.
TABLE 2 Comparison of Patients With and Without Inappropriate Device Therapy
Characteristic
Men
Age (yrs)
LVEF (%)
Underlying cardiac disease
Coronary artery disease
Dilated cardiomyopathy
History of atrial tachyarrhythmias
Index arrhythmia
Ventricular fibrillation
VT
Nonsustained VT
Pharmacologic treatment
Amiodarone
␤ blockade
Digoxin
ACE inhibitor
Diuretic
Lipid-lowering drug
Abbreviation as in Table 1.
Inappropriate ICD
Therapy (n ⫽ 37)
30 (81%)
61 ⫾ 12
31 ⫾ 14
25 (68%)
9 (24%)
19 (51%)
12 (32%)
19 (52%)
6 (16%)
12 (32%)
19 (51%)
11 (30%)
30 (82%)
23 (62%)
17 (46%)
FIGURE 2. Actuarial event rates for inappropriate device therapy
for patients with and without a history of atrial tachyarrhythmias.
due to atrial tachyarrhythmias. The observations noted
in this study are: (1) a history of atrial tachyarrhythmias is an independent predictor of inappropriate therapy and (2) recurrent VT with a CL ⱖ350 ms that
triggered device therapy is associated with an increased risk for first inappropriate therapy.
In this study, the incidence of inappropriate ICD
therapy was 14%. Most patients experienced inappropriate ICD interventions in the first year after device
implantation, regardless of primary and secondary
prevention. These findings agree with studies reporting on inappropriate ICD therapy.11,12 In the Antiarrhythmics Versus Implantable Defibrillators Trial,
atrial tachyarrhythmias were responsible for inappropriate therapy in 22% of patients and 16% of all
treated episodes.12
It is no surprise that patients with a history of atrial
tachyarrhythmias are at risk for inappropriate interven-
1. Moss AJ, Hall WJ, Cannom DS, Daubert JP, Higgins SL, Klein H, Levine JH,
Saksena S, Waldo AL, Wilber D, et al. Improved survival with an implanted
defibrillator in patients with coronary disease at high risk for ventricular arrhythmia. Multicenter Automatic Defibrillator Implantation Trial Investigators. N Engl
J Med 1996;335:1933–1940.
2. The Antiarrhythmics Versus Implantable Defibrillators (AVID) Investigators.
A comparison of antiarrhythmic-drug therapy with implantable defibrillators in
patients resuscitated from near-fatal ventricular arrhythmias. N Engl J Med
1997;337:1576 –1583.
3. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS, Daubert JP,
Higgins SL, Brown MW, Andrews ML. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl
J Med 2002;346:877– 883.
4. Grimm W, Flores BF, Marchlinski FE. Electrocardiographically documented
unnecessary, spontaneous shocks in 241 patients with implantable cardioverter
defibrillators. Pacing Clin Electrophysiol 1992;15:1667–1673.
5. Nunain SO, Roelke M, Trouton T, Osswald S, Kim YH, Sosa-Suarez G,
Brooks DR, McGovern B, Guy M, Torchiana DF, et al. Limitations and late
complications of third-generation automatic cardioverter-defibrillators. Circulation 1995;91:2204 –2213.
6. Kuhlkamp V, Dornberger V, Mewis C, Suchalla R, Bosch RF, Seipel L.
Clinical experience with the new detection algorithms for atrial fibrillation of a
defibrillator with dual chamber sensing and pacing. J Cardiovasc Electrophysiol
1999;10:905–915.
7. Sticherling C, Schaumann A, Klingenheben T, Hohnloser SH. First worldwide
clinical experience with a new dual chamber implantable cardioverter defibrillator: advantages and complications. Europace 1999;1:96 –102.
8. Costeas XF, Link MS, Foote CB, Homoud MK, Wang PJ, Estes NA. Predictors
of ventricular tachycardia recurrence in 100 patients receiving tiered therapy
defibrillators. Clin Cardiol 2000;23:852– 856.
9. Raitt MH, Klein RC, Wyse DG, Wilkoff BL, Beckman K, Epstein AE,
Coromilas J, Friedman PL, Martins J, Ledingham RB, et al. Comparison of
BRIEF REPORTS
273
arrhythmia recurrence in patients presenting with ventricular fibrillation versus
ventricular tachycardia in the Antiarrhythmics Versus Implantable Defibrillators
(AVID) trial. Am J Cardiol 2003;91:812– 816.
10. Raitt MH, Dolack GL, Kudenchuk PJ, Poole JE, Bardy GH. Ventricular
arrhythmias detected after transvenous defibrillator implantation in patients with
a clinical history of only ventricular fibrillation. Implications for use of implantable defibrillator. Circulation 1995;91:1996 –2001.
11. Gradaus R, Block M, Brachmann J, Breithardt G, Huber HG, Jung W, Kranig
W, Mletzko RU, Schoels W, Seidl K, et al. Mortality, morbidity, and complications in 3344 patients with implantable cardioverter defibrillators: results from the
German ICD Registry EURID. Pacing Clin Electrophysiol 2003;26:1511–1518.
12. Klein RC, Raitt MH, Wilkoff BL, Beckman KJ, Coromilas J, Wyse DG,
Friedman PL, Martins JB, Epstein AE, Hallstrom AP, et al. Analysis of implantable cardioverter defibrillator therapy in the Antiarrhythmics Versus Implantable
Defibrillators (AVID) Trial. J Cardiovasc Electrophysiol 2003;14:940 –948.
13. Schmitt C, Montero M, Melichercik J. Significance of supraventricular
tachyarrhythmias in patients with implanted pacing cardioverter defibrillators.
Pacing Clin Electrophysiol 1994;17:295–302.
14. Wolpert C, Jung W, Spehl S, Schimpf R, Omran H, Schumacher B, Esmailzadeh B, Tenzer D, Mehra R, Luderitz B. Incidence and rate characteristics of
atrial tachyarrhythmias in patients with a dual chamber defibrillator. Pacing Clin
Electrophysiol 2003;26:1691–1698.
15. Tavernier RH, Kimman GP, Theuns DAMJ, Klootwijk AP, Kazmierczak J,
Jordaens LJLM. Comparison of events and survival in dual- versus single
chamber ICDs. Neth Heart J 1999;6:604 – 608.
16. Gronefeld GC, Mauss O, Li YG, Klingenheben T, Hohnloser SH. Association
between atrial fibrillation and appropriate implantable cardioverter defibrillator
therapy: results from a prospective study. J Cardiovasc Electrophysiol 2000;11:
1208 –1214.
17. Stein KM, Euler DE, Mehra R, Seidl K, Slotwiner DJ, Mittal S, Markowitz
SM, Lerman BB. Do atrial tachyarrhythmias beget ventricular tachyarrhythmias
in defibrillator recipients? J Am Coll Cardiol 2002;40:335–340.
18. Bansch D, Castrucci M, Bocker D, Breithardt G, Block M. Ventricular
tachycardias above the initially programmed tachycardia detection interval in
patients with implantable cardioverter-defibrillators. Incidence, prediction and
significance. J Am Coll Cardiol 2000;36:557–565.
Design of the SHock Inhibition Evaluation with
Azimilide (SHIELD) Study: A Novel Method to Assess
Antiarrhythmic Drug Effect in Patients With an
Implantable Cardioverter-Defibrillator
Craig M. Pratt, MD, Paul Dorian, MD, Hussein R. Al-Khalidi, PhD, Jose M. Brum, MD,
Martin Borggrefe, MD, Daljit S. Tatla, PhD, Johannes Brachmann, MD,
Robert J. Myerburg, MD, David S. Cannom, MD, Michael J. Holroyde, PhD,
Michael van der Laan, MD, and Stefan H. Hohnloser, MD, on behalf of the SHIELD
Investigators
This report presents the rationale and study design details of the SHock Inhibition Evaluation with Azimilide
study, which is recruiting 624 patients with implantable
cardioverter-defibrillators (ICDs) who are at risk for lifethreatening ventricular arrhythmia, randomized to azimilide 75 mg, azimilide 125 mg, or placebo and followed for 1 year. The objective of this study is to
determine the effect of azimilide versus placebo on the
symptomatic ventricular arrhythmia burden using a
unique statistical analysis based on the unusual temporal distribution of symptomatic ICD therapies. The primary efficacy end points are time to all-cause shocks
and time to all-cause shocks plus symptomatic ventricular arrhythmic events triggering antitachycardia pacFrom The Methodist DeBakey Heart Center and the Section of Cardiology, Department of Medicine, Baylor College of Medicine,
Houston, Texas; the Division of Cardiology, St. Michael’s Hospital,
Toronto, Ontario, Canada; Health Care Research Center, Procter &
Gamble Pharmaceuticals, Cincinnati, Ohio; Klinikum Mannheim, Universtatsklinikum, Mannheim, Germany; Landkrankenhaus Coburg, II
Medizinische Klinik, Coburg, Germany; University of Miami School of
Medicine, Miami, Florida; Los Angeles Cardiology Associates, Los
Angeles, California; Procter & Gamble Pharmaceuticals, Egham,
United Kingdom; and J. W. Goethe University, Medizinische Klinik IV,
Kardiologie, Frankfurt, Germany. This study was sponsored by Procter
& Gamble Pharmaceuticals, Cincinnati, Ohio. Dr. Pratt’s address is:
The Methodist DeBakey Heart Center, 6565 Fannin Street, F1001,
Houston, Texas 77030. E-mail: cpratt@bcm.tmc.edu. Manuscript received July 14, 2004; revised manuscript received and accepted
August 31, 2004.
274
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
ing measured from randomization. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:274 –276)
he SHock Inhibition Evaluation with Azimilide
(SHIELD) study is designed to expand our knowlT
edge of the efficacy of azimilide in reducing symptomatic ventricular arrhythmias and implantable cardioverter-defibrillator (ICD) therapies, total ventricular
arrhythmias, and “electrical storms” in patients with
ICDs.
•••
The SHIELD study is a randomized, double-blind,
placebo-controlled, parallel-group study of 624 patients assigned to placebo or oral azimilide 75 or 125
mg and followed for 365 days. The scientific organization of the SHIELD study includes the Executive
Steering Committee, the Event Committee, and the
Data and Safety Monitoring Board. A list of investigators is presented in the Appendix.
The major inclusion criterion is to include patients
with ICDs capable of arrhythmia discrimination algorithms (e.g., high rate, sudden onset, or rate stability).
Patients who have received their first ICDs will be
randomized ⱕ72 days after documented episodes of
sustained ventricular tachycardia (VT), cardiac arrest,
or ventricular fibrillation that occurred within a
6-week time period (42 days) before the implantation
of the ICDs. Patients with ICDs fitted for previous
cardiac arrest or ventricular fibrillation are required to
have left ventricular ejection fractions ⱕ40% to qual0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.096
ify for the study. Patients who have an ICD implanted
for ⬎90 days and had an ICD shock triggered by
spontaneous VT or ventricular fibrillation must be
randomized ⱕ180 days after this shock. Major cardiac
exclusion criteria include New York Heart Association class IV congestive heart failure, heart transplantation, unstable angina pectoris, recent myocardial infarction (within 30 days), a history of Torsades de
Pointes VT, QTc values ⬎440 ms (with a QRS interval ⱕ120 ms), or JTc values ⬎320 ms (if QRS ⬎120
ms). Noncardiac exclusions are standard in nature.
Patients with QTc values ⬎525 ms (with QRS intervals ⱕ120 ms) or JTc values ⬎400 ms (with QRS
intervals ⬎120 ms) will be withdrawn from the study.
Patients with absolute neutrophil counts ⱕ1,000
cells/␮l at any time will be withdrawn.
In an attempt to minimize heterogeneity regarding
ICD programming and to minimize the effect of different devices, investigators are required to program
each ICD in a standardized fashion: antitachycardia
pacing (ATP) must be programmed “on” for all patients at the time of randomization and set for a
minimum of 2 attempts of ATP in the lowest detection
zone; the “floor” and “ceiling” for this zone are prescribed according to the slowest documented or induced VT. The first shock strength is to be set no less
than the smallest successful defibrillation energy at
implantation or the VT before randomization. At least
1 discriminator to reduce the likelihood of inappropriate therapy must be enabled. All ICD therapy events
will be adjudicated by the independent Event Committee (Appendix) blinded to treatment assignment.
ICD interrogation is mandatory at all scheduled visits
and unscheduled visits (caused by symptomatic arrhythmias and/or ICD therapies). At each of these
visits, patients are questioned regarding the occurrence of arrhythmia symptoms before the interrogation of the ICD. The date of occurrence and the nature
of these symptoms are documented at each visit.
The primary efficacy end points are (1) time to
all-cause shocks and (2) time to all-cause shocks plus
symptomatic ventricular arrhythmic events terminated
by ATP measured from randomization. The secondary
efficacy end point is time to all ventricular arrhythmias terminated by ICD shocks or ATP. An important
tertiary end point is time to all-cause electrical storms,
where an “electrical storm” is defined as ⱖ3 separate
episodes of ventricular arrhythmia terminated by ICD
shocks or ATP within a time period of 24 hours.
Assuming a placebo incidence of 4 events per patientyear for all-cause shocks,1 a total of 624 patients
followed for 365 days is required to detect a 30%
reduction (hazard ratio 0.7) in the frequency of allcause shocks with 90% power at a significance level
of 0.25%.2 The Andersen-Gill mean intensity model is
the primary statistical technique to evaluate the observed differences in the primary end point between
randomized treatment groups.3 All events for each
patient are thus counted, not only the first event. This
model compares the distribution of all the interevent
intervals in the placebo and drug treatment groups and
adjusts for the correlated events over time. Patients are
FIGURE 1. The cumulative relative frequency of interevent intervals is plotted against the values of interevent intervals in days
in a patient with 21 appropriate shocks during 380 days of follow-up. The abscissa represents the time interval between successive shocks, not a representation of follow-up time. For this
patient, 70% (cumulative relative frequency 0.7) of all interevent
intervals were <1 day (24 hours). The dashed (“fitted”) line represents the expected cumulative frequency if interevent intervals
were randomly distributed, that is, followed an exponential distribution. The observed and fitted distributions are significantly
different (p <0.01).
stratified by ejection fraction (ⱕ40% vs ⬎40%),
␤-blocker use, and ICD type (new vs existing).
The unique features of the SHIELD study design
include the exploration of a dose range of azimilide
on symptomatic ventricular arrhythmias and ICDdelivered therapies, the total ventricular arrhythmia
burden, and electrical storms in a patient population
with uniform programming of ICD settings to reduce device variability. This latter feature is of
crucial importance, because end points related to
device therapy are determined by the device settings as well as the detected arrhythmia and can be
confounded by variability in programming arrhythmia detection and therapies.
An additional innovative aspect of the SHIELD
study relates to the inclusion of all symptomatic ventricular arrhythmic events, not just the first event in
the primary end point, to capture the total symptomatic arrhythmia burden (ICD shocks and symptomatic
ventricular arrhythmias terminated by ATP) and, similarly, all ventricular arrhythmic events, as the secondary end point. In contrast, previous ICD trials assessing sotalol and metoprolol focused only on time to
first ICD shock.4,5 This is not an optimal end point to
assess the effect of an antiarrhythmic therapy, because
it does not consider either all symptomatic events
(shocks and symptomatic ventricular arrhythmias terminated by ATP) or all ventricular arrhythmias. In
addition, the “time to first event” approach is a valid
end point only when the events are distributed according to a Poisson process, that is, randomly, with the
same distribution as the interevent intervals.6 The
Poisson process provides a mathematic model for
completely random events for which the cumulative
BRIEF REPORTS
275
Andersen-Gill mean intensity model is a semiparametric regression model that applies to recurrent events
and generates robust standard error estimators, using
all accumulated interevent intervals.
APPENDIX
FIGURE 2. The cumulative relative frequency of interevent intervals is plotted against the values of the interevent intervals in
days. There are 86 appropriate shocks (events) in 18 patients.
For this population, half (cumulative relative frequency 0.5) of all
interevent intervals were <1 day (24 hours). See also legend in
Figure 1.
interevent intervals (the time intervals between each
consecutive pair of ICD therapies) follow an exponential distribution.
As illustrated in Figure 1, the observed cumulative
relative frequency of appropriate ICD shocks (data
were adapted from the placebo patients in the azimilide pilot study1 to remove any treatment effect on the
distribution of interevent intervals) are clustered and
nonrandomly distributed and do not follow an exponential distribution. Under a Poisson process (assuming that shocks occur at random), this patient would be
expected to have, on average, an appropriate shock
every 20 days. However, Figure 1 shows that about
70% of this patient’s shocks occurred ⱕ24 hours after
the initial shock. Similarly, in our pilot study, the
cumulative placebo population interevent intervals
(Figure 2) are also clustered, illustrating a greater risk
for arrhythmia recurrence shortly after each event. For
events with this type of distribution, the time to the
first event is an inadequate depiction of the overall
antiarrhythmic drug effect on ICD therapies. The total
symptomatic arrhythmia burden is the best representation of the disease severity and the efficacy of therapy. However, the simple calculation of total number
of events per patient or total events per patient per
year, does not adjust for the correlation between
events within patients, because most of the events
occur within a short time after the previous event. The
276 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
Executive Steering Committee: Paul Dorian, Craig Pratt, Martin Borggrefe, Johannes Brachmann, Stefan Hohnloser, David Cannom, and Robert
Myerburg.
Data and Safety Monitoring Board: John Cairns, Leone Greene, William
E. Wilkinson, and Carina Blomström Lundqvist.
Event Committee: Igor Singer, Jesus Almendral, Tony Simmons, Patrick
Tchou, Vilma Torres, Kenneth Stein, Imran Niazi, Francois Philippon, Walid
Saliba, and David Martin.
Investigators: Etienne Aliot, Barry Alpert, Dietrich Andresen, Fernando
Arribas, John Bailey, Malcolm Bersohn, Ulrika Birgersdotter-Green, Tristram
Bahnson, Madgy Basta, John T. Beard, David Benditt, Joseph Bissett, Louis
Blier, Martin Borggrefe, Johannes Brachmann, Christopher Brown, Joseph Brugada, Jeffrey Buetikofer, David Cannom, Mark Carlson, Anthony Chang, Don
Chilson, Ted Chow, Lai Chow-Kok, Jacques Clementy, James Coman, Stuart
Connolly, Sean Connors, James Cook, Steven Kalbfleisch, Richard Corbelli,
Pierre Cosnay, George Crossley, Ken Curry, Joel Cutler, James Daubert, Luc De
Roy, Rainer Dietz, John DiMarco, Paul Dorian, Kenneth Ellenbogen, Martin
Emert, Roger Freedman, Eli Gang, J. Anthony Gomes, Martin Gottwik, Ricardo
Ruiz Granell, Martin Green, Arnold Greenspon, Thomas Guarnieri, Hartmut
Gulker, Moshe Gunsburg, Charles Haffajee, Mark Hamer, David Henderson,
Steven Higgins, Ellen Hoffmann, Stefan Hohnloser, Andrew Hordes, Alberto
Interian, Mohammed Jazayeri, Jose Joglar, Roy John, Luc Jordaens, Salem Kacet,
Steven Kalbfleisch, Mohammed Yousuf Kanjwal, Stephen Keim, Shane Kimber,
Helmut Klein, Zdzislawa Kornacewicz-Jach, Volker Kühlkamp, Herve Le Marec,
Antoine Leenhardt, Robert Leman, Bruce Lerman, Samuel Levy, Albert Lin,
Ignacio Lozano, Andrzej Lubinski, Richard Luceri, Berndt Luderitz, Jean-Claude
Mabo, Paul Maccaro, Michael Markel, Frank McGrew, Craig McPherson, A.
Meijer, Thomas Meinertz, William M. Miles, Ralph Mletzko, Foad Moazez,
Michael Mollerus, George Monir, David Nabert, Imran Niazi, Brendan
O’Cochlain, Gearaoid O’Neill, Eraldo Occhetta, Michael Oeff, Brian Olshansky,
Sudha Pai, John Payne, Christian Perings, Heinz Pitschner, Edward Platia, Scott
Pollak, AR Ramdat Misier, Eric Rashba, Pratap Reddy, Paolo Rizzon, Francisko
Alzueta Rodriguez, Marc Roelke, Lawrence Rosenthal, Andrea Russo, Sanjeev
Saksena, Robert Sangrigoli, Massimo Santini, Martin Jan Schalij, Alexander
Schirdewan, Claus Schmitt, Wolfgang Schols, Peter Schwartz, Richard Sheahan,
Hue-Teh Shih, Karlheinz Seidl, Tony Simmons, Igor Singer, Nicholas Stamato,
Jonathan Steinberg, Laurence Sterns, Hanna Szwed, Mario Talajic, Patrick
Tchou, Paul Touboul, Maria Trusz-Gluza, Isabelle Van Gelder, Yves Vandekerckhove, Ernst Vester, Juergen Vogt, Abdul Wase, Peter Wells, Eric Wever,
Teresa Widomska-Czekajska, Arthur Wilde, Stephen Winters, Deborah Wolbrette, Kevin Wolfe, Seth Worley, and Manfred Zehender.
1. Singer I, Al-Khalidi H, Niazi I, Tchou P, Simmons T, Henthorn R, Holroyde
M, Brum J. Azimilide decreases recurrent ventricular tachyarrhythmias in patients with implantable cardioverter defibrillators. J Am Coll Cardiol 2004;
43:39 – 43.
2. Signorini D. Sample size for Poisson regression. Biometrika 1991;78:
446 – 450.
3. Lin DY, Wei LJ, Yang I, Ying Z. Semiparametric regression for the mean and
rate functions of recurrent events. R Stat Soc Ser B 2000;62:711–730.
4. Pacifico A, Hohnloser SH, Williams JH, Tao B, Saksena S, Henry PD,
Prystowsky EN. Prevention of implantable-defibrillator shocks by treatment with
sotalol. d,l-Sotalol Implantable Cardioverter-Defibrillator Study Group. N Engl
J Med 1999;340:1855–1862.
5. Seidl K, Hauer B, Schwick NG, Zahn R, Senges J. Comparison of metoprolol
and sotalol in preventing ventricular tachyarrhythmias after the implantation of a
cardioverter/defibrillator. Am J Cardiol 1998;82:744 –748.
6. Rose MS, Gillis AM, Sheldon RS. Evaluation of the bias in using time to the
first event when the inter-event intervals have a Weibull distribution. Stat Med
1999;18:139 –154.
JANUARY 15, 2005
Refeeding Normalizes the QT Rate Dependence of
Female Anorexic Patients
Frédéric Roche, MD, PhD, Jean-Claude Barthélémy, MD, PhD, Norbert Mayaud, MD,
Vincent Pichot, PhD, David Duverney, PhD, Natacha Germain, MD, François Lang, MD,
and Bruno Estour, MD
We postulated that dynamic alterations in QT interval
adaptation could characterize patients with anorexia
nervosa (AN) and could be restored after weight
gain. To assess ventricular repolarization features,
we evaluated the QT dynamicity along RR intervals
from 24-hour electrocardiographic data of patients
with AN before and after refeeding. Ten young
women with AN (19 ⴞ 3 years) were included in the
study. The QT/RR slope was found significantly enhanced compared with normals (ⴚ1.82 ⴞ 0.62 vs
ⴚ1.40 ⴞ 0.30; p <0.05). This slope returned to
normal range values after refeeding. The QT/RR
slope was significantly correlated with the body mass
index (r ⴝ 0.59; p <0.007) in the patient
group. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:277–280)
he clinical manifestations of anorexia nervosa
(AN) have been well described for several deT
cades. Its mortality rate is the highest of any major
psychiatric disorder, with many deaths occurring suddenly.1 Although reports have suggested an association between sudden death in anorexic patients and
QT prolongation on routine electrocardiography
(ECG), the incidence and the clinical consequences of
such abnormalities remain controversial.2– 4 That a
primary cardiac abnormality may be present at least in
some cases is suggested by findings of repolarization
abnormalities and QT-interval prolongation under a
variety of circumstances associated with severe
weight loss.5 QT-rate dependence is 1 of the major
properties of ventricular repolarization, with its circadian and autonomic modulations.6 Susceptibility to
severe ventricular arrhythmias has been associated
with the alteration of the rate-dependence adaptation
of repolarization after myocardial infarction as well as
in acquired or congenital long-QT syndrome. We thus
confirmed that dynamic alterations in QT-interval adaptation7 and its link with abnormalities in autonomic
control could characterize patients with AN.
To assess ventricular repolarization features, we
compared QT parameters and their dynamicity along
RR intervals from the 24-hour electrocardiographic
From Service de Physiologie Clinique et de l’Exercice – Groupe PPEH,
CHU Nord, Service d’Endocrinologie, CHU Bellevue, and Service de
Psychiatrie, CHU Bellevue, Faculté de Médecine Jacques Lisfranc,
Université Jean Monnet, Saint-Etienne, France. Dr. Roche’s address is:
EFCR, CHU Nord, Niveau 6, F-42055 Saint-Etienne Cedex 2,
France. E-mail: frederic.roche@univ-st-etienne.fr. Manuscript received
May 18, 2004; revised manuscript received and accepted September 7, 2004.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
data of patients with and without this serious medical
eating disorder. Furthermore, the improvement of
these parameters was tested after medically controlled
refeeding.
•••
Ten young women (mean age 19 ⫾ 3 years) with
diagnosed AN fulfilling the Diagnostic and Statistical
Manual of Mental Disorders (fourth edition) criteria
for AN were included in the study. All patients were
hospitalized because of clinical signs of malnutrition,
with a mean body mass index (BMI) of 14.7 ⫾ 2.3
kg/m2. Their body weights averaged ⫺34.2 ⫾ 6.9% of
their ideal body weights. The study procedure was
fully explained to the subjects, who gave their informed consent to participate. All patients were reevaluated after 5 ⫾ 2 months of medically controlled
refeeding, which allowed a significant increase in
BMI to 17.6 ⫾ 4.2 kg/m2 (p ⬍0.003 vs baseline) and
a reduction of body weight deficit reaching ⫺21.6 ⫾
9.2% (p ⬍0.003 vs baseline). Reference values for
24-hour electrocardiographic recordings and QT analysis were obtained from a control group of 10 healthy
female medical students aged 20 ⫾ 3 years (mean
BMI 23.1 ⫾ 1.9 kg/m2).
Standard 3-channel Holter tape recorders were
used to acquire the data (Del Mar Reynolds Medical,
Inc., Irvine, California). The Holter electrocardiographic system allowed the extraction of the list of RR
intervals with a precision of 1/256 second.
Heart rate variability (HRV) was first evaluated
using time-domain analysis. To perform the analysis,
only intervals between normal beats were considered.
Several classic parameters were calculated8: the rootmean-square of successive differences (the square root
of the mean of the sum of the square differences
between adjacent normal RR intervals), the SD of all
normal RR intervals in the entire recording, and the
proportion of consecutive beats with RR intervals
showing a ⬎50-ms difference. Spectral analysis of
HRV was thus performed by the fast-Fourier transform method using sliding 256-point Hanning windows.8 The interpolation mode was linear. Frequencies in the band from 0.04 to 0.14 Hz were considered
low frequencies (LFs), and frequencies from 0.15 to
0.40 Hz were considered high frequencies (HFs). The
LF/HF ratio was calculated. The LF and HF powers of
HRV were further computed in normalized units.
A RR interval was considered to be a pause when
it reached or exceeded 2.5 seconds. Bradycardia was
defined as a heart rate (HR) ⬍40 beats/min. Isolated
and grouped ventricular ectopic activities were analyzed. Nonsustained ventricular tachycardia was de0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.018
277
the time of hospital admission for the
AN group are listed in Table 1. Periods
of bradycardia were noted in 6 of 10
patients and pauses were observed in
AN Before
AN After
Controls
10% of patients with AN. Frequent
Parameter
(n ⫽ 10)
(n ⫽ 10)
(n ⫽ 10)
supraventricular premature beats
Minimal HR (beats/min)
46 ⫾ 9
61 ⫾ 16†
55 ⫾ 4
(ⱖ10/hour) were observed in 10% and
Maximal HR (beats/min)
130 ⫾ 14
141 ⫾ 10
145 ⫾ 20
frequent
ventricular premature beats
Day, mean HR (beats/min)
73 ⫾ 12*
92 ⫾ 16
82 ⫾ 10
were also noted in 1 patient. Neither
Night, mean HR (beats/min)
56 ⫾ 11*
73 ⫾ 19
67 ⫾ 11
VPCs ⱖ10/h (%)
1 (10)
0
1 (10)
high-grade nor frequent (ⱖ10/hour)
Nonsustained VT (%)
0
0
0
ventricular ectopy was evident in the
Sinoatrial block (%)
1 (10)
0
0
patients, but 1 presented with idioven†
Bradycardia (%)
6 (60)*
0
0
tricular accelerated rhythm (1 salve, 4
*p ⬍0.05, AN before versus controls; †p ⬍0.05, AN before versus AN after refeeding.
beats). After refeeding, none of the paVPCs ⫽ ventricular premature complexes; VT ⫽ ventricular tachycardia.
tients with AN presented with bradycardia, pauses, or significant ventricular arrhythmias. Patients with AN
fined as ⱖ6 consecutive ventricular ectopic complexes presented with lower HRs than controls during the night
at a rate ⬎120 beats/min and lasting ⬍30 seconds. and the day. The maximal and the minimal HRs were
The severity of ventricular arrhythmias was evaluated affected in anorexic patients, but the differences did not
by the mean number of ventricular premature com- reach statistical significance. All parameters returned
plexes per hour (threshold values ⱖ10 per hour). close to the control group values after refeeding. The
Supraventricular tachycardia was defined as ⬎3 con- relative bradycardia observed in patients with AN before
secutive supraventricular ectopic complexes at a rate refeeding was associated with an enhanced autonomic
tone, as the SD of all normal RR intervals in the entire
⬎120 beats/min.
The Holter tapes were analyzed for QT intervals by recording was found to be significantly greater and the
a computer-assisted method (QT Analysis System; HFs as well as the HFs in normalized units tended to be
Del Mar Reynolds Medical, Inc.). Data for QT anal- greater but without significant difference in this small
ysis was derived from a good-quality Holter recording population. The autonomic tone indexes tended to go
with a 256-Hz sample rate and containing no more back to a normal range value after refeeding. Mean QT
than 20 minutes of total artifact. T-wave amplitude duration and corrected QT values did not differ statistimust be ⱖ0.2 V to ensure consistent discrimination cally during the 24-hour period in subjects with AN
between QT peak and QT end. The operator (FR) first (Table 2). No abnormal 24-hour mean QTc duration was
performed a careful Holter analysis before attempting observed in the control group, and only 1 anorexic paQT analysis to be sure to detect and correctly label all tient showed a mean QTc reaching 440 ms after the
cardiac beats. The Holter channel corresponding to refeeding procedure. The QT/RR slope was found to be
lead V5 on the surface ECG was used. The operator the most significant discriminating parameter of the cardefined a template QT interval by the beginning of the diac repolarization analysis observed between patients
QRS and the end of the T wave on 1 beat. The end of with AN at baseline and controls. QT length related to
the T wave was defined as the point at which the HR was found to be longer for lower HRs in patients
down-slope of the T wave crossed the isoelectric line. with AN compared with normal controls. This QT/RR
The waveform analyzer was programmed to sample relation was significantly correlated in the AN group
QRS complexes at a specific time, and the RR and QT (before and after refeeding) with the BMI (r ⫽ 0.59,
intervals were measured automatically. Mean ⫾ SD p ⬍0.007), body weight deficit (r ⫽ 0.58, p ⬍0.008),
QT was calculated from all the beats measured. Cor- minimal HR (r ⫽ 0.75, p ⬍0.001), mean HR (r ⫽ 0.76,
rected QT values at 60 beats/min estimated from Ba- p ⬍0.001), SD of all normal RR intervals in the entire
zett’s quadratic correction formula were calculated recording (r ⫽ ⫺0.60, p ⬍0.03), LF/HF ratio (r ⫽ 0.38,
from all QT values for each hour of analysis. QT-rate p ⬍0.05), and HF in normalized units (r ⫽ ⫺0.49,
dependence graph plots for each HR range against QT p ⬍0.03).
were individually constructed and showed the least•••
squares fit regression lines and formulas for these
AN leads to abnormal beat-to-beat changes in venpoints.
tricular repolarization, characterized by an enhanced reComparisons of clinical status and QT parameters lation between QT duration and RR intervals. This relabetween patients with AN and those without were tion is significantly correlated with the bradycardia
based on unpaired Student’s t tests for continuous data associated with the eating disorder. Furthermore, these
and on chi-square tests for categorical data. The paired abnormalities seem to be reversible after medically conStudent’s t test was used to compare parameters at trolled refeeding. The role of autonomic cardiac regulafollow-up. Statistical significance was defined at the p tion in ventricular repolarization adaptation seems to be
⬍0.05 level. Pearson’s product moment procedure very important9: pharmacologic sympathetic activation
was used to determine the correlation between vari- or blockade has been shown to greatly alter QT length
ables and the QT/RR shape.
and QT-rate adaptation. This suggests a direct influence
Twenty-hour-hour electrocardiographic findings at of cholinergic activity on the repolarization of ventricular
TABLE 1 Evolution of the 24-hour Electrocardiographic Holter Parameters of the
Population With AN Before and After Refeeding and Comparison Controls
278 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
focused on QT dispersion in patients
with AN.14 –16 This method, however, has several operator-dependent
limitations, and its prognostic impliAN Before
AN After
Controls
cations remain uncertain in high-risk
Measurement
(n ⫽ 10)
(n ⫽ 10)
(n ⫽ 10)
populations, such as patients who
SDNN (ms)
207 ⫾ 64*
148 ⫾ 76†
173 ⫾ 20‡
have had myocardial infarctions. UnpNN50 (%)
27 ⫾ 15
16 ⫾ 16
23 ⫾ 4
usual autonomic imbalance characRMSSD (ms)
57 ⫾ 16
41 ⫾ 27
52 ⫾ 8
terizes the severe starvation status.
2,982 ⫾ 1,467
2,195 ⫾ 1,776
3,868 ⫾ 1,016
Total PSD (ms2/Hz)
VLF PSD (ms2/Hz)
1,378 ⫾ 579
877 ⫾ 625
1,978 ⫾ 684
Kollai et al,17 using baroreflex sensiLF PSD (ms2/Hz)
602 ⫾ 253
487 ⫾ 327
1,066 ⫾ 419
tivity,
and others,18 –20 using valiHF PSD (ms2/Hz)
873 ⫾ 705
646 ⫾ 839
576 ⫾ 171
dated
HRV
analysis, have reported
2
LFnu PSD (ms /Hz)
45 ⫾ 18
55 ⫾ 18
55 ⫾ 11
enhanced parasympathetic control of
HFnu PSD (ms2/Hz)
49 ⫾ 17
45 ⫾ 18
31 ⫾ 9
HR probably associated with sympaLF/HF ratio
1.4 ⫾ 1.4
1.6 ⫾ 1.1
2.0 ⫾ 1.0
QT/RR slope
⫺1.8 ⫾ 0.7*
⫺1.1 ⫾ 0.8†
⫺1.4 ⫾ 0.3
thetic withdrawal in patients with
QT mean (ms)
374 ⫾ 29
345 ⫾ 55
374 ⫾ 28
AN. It should be noted that in our
QT SD (ms)
39 ⫾ 10*
31 ⫾ 7†
31 ⫾ 5
population, HF values as well as norQTc Bazett (ms)
388 ⫾ 20*
410 ⫾ 28
412 ⫾ 14
malized HF power spectral density
*p ⬍0.05, AN before versus controls; †p ⬍0.05, AN before versus AN after refeeding; ‡p ⬍0.05,
tended to be greater in patients with
AN after versus controls.
AN before treatment but without staHFnu ⫽ HF computed in normalized units; LFnu ⫽ LF computed in normalized units; pNN50 ⫽
tistically significant differences beproportion of consecutive beats with RR intervals showing a ⬎50-ms difference; PSD ⫽ power spectral
density; RMSSD ⫽ root mean square of successive differences; SDNN ⫽ SD of all normal RR intervals
cause of the small sample of patients.
in the entire recording; VLF ⫽ very LF.
None of the HRV parameters, except
SDNN, seems to have significantly
altered after refeeding.
myocardium independently from rate adaptation of QT
In conclusion, the QT/RR relation is found to be
intervals after the elimination of cardiac parasympathetic enhanced in patients with AN, reflecting in part the
activity. Because the prolongation of QT interval has specific autonomic imbalance encountered in this popubeen associated with sudden cardiac death, the relation lation, and this abnormal pattern seems to be reversible
between semistarvation periods, the development of after refeeding. The clinical implications of such findings
non–rate-related QT-interval prolongation, and subse- need to be discussed, because an equivalent enhancequent clinical outcomes have been examined by several ment of the QT/RR slope has been described in patients
investigators first during therapeutic starvation for with life-threatening ventricular arrhythmias.
morbid obesity and then in patients with AN.10 QT
prolongation had been described by Rasmussen and
Andersen11 in 12 of 22 women in association with
Acknowledgment: We are indebted to Christian
weight reduction after gastroplasty. Thurston and Marks2 Roche, MD, Maryse Victoire, BSc, and Delphine
described minor degrees of QT-interval prolongation in 5 Maudoux, BSc, for their expert technical assistance.
of 9 women with AN but did not correct for HR, and
Isner et al10 retrospectively reported the prolongation of
QTc intervals in 3 adult anorexic women who died 1. American Psychiatric Association. Practice guidelines for eating disorders.
suddenly. The absence of histologic abnormalities and of Am J Psychiatry 1993;150:212–228.
2. Thurston J, Marks P. Electrocardiographic abnormalities in patients with
electrolyte imbalance within the heart suggests that some anorexia
nervosa. Br Heart J 1974;36:719 –723.
extracardiac influence may explain the potential QT pro- 3. Isner JM, Roberts WC, Heymsfield SB, Yager J. Anorexia nervosa and sudden
longation observed in such patients. It is not inconceiv- death. Ann Intern Med 1985;102:49 –52.
4. Cooke RA, Chambers JB, Singh R, Todd GJ, Smeeton NC, Treasure J,
able that autonomic inflow to the heart might be de- Treasure
T. QT interval in anorexia nervosa. Brit Heart J 1994;72:69 –73.
ranged, particularly under conditions of stress, and this 5. Pringle TH, Scobie IN, Murray RG, Kesson CM, Maccuish AC. Prolongation
should limit the rate adaptation of the QT interval in such of the QT interval during therapeutic starvation: a substrate for malignant arInt J Obesity 1983;7:253–261.
conditions. In contrast, it should be noted that other rhythmias.
6. Viitasalo M, Karjalainen J. QT intervals at heart rates from 50 to 120 beats per
12,13
investigators
have been unable to document QT- minute during 24-hour electrocardiographic recordings in 100 healthy men.
interval prolongation in their patients with AN. In our Circulation 1992;86:1439 –1442.
7. Roche F, Estour B, Kadem M, Millot L, Pichot V, Duverney D, Gaspoz JM,
population, we could not confirm a significantly large Barthélémy
JC. Alteration of the QT rate dependence in anorexia nervosa. PACE
prevalence of abnormal QT prolongation. Although the 2004;27:1099 –1104.
mean QTc interval, which should be considered as the 8. Task Force of the European Society of Cardiology and the North American
of Pacing and Electrophysiology. Heart rate variability: standards of
most discriminating parameter in the evaluation of de- Society
measurements, physiological interpretation, and clinical use. Circulation 1996;
polarization in several heart disorders, was not signifi- 93:1043–1065.
cantly altered by AN in our population, only 1 patient 9. Ahnve S, Vallin H. Influence of heart rate and inhibition of autonomic tone on
QT interval. Circulation 1982;65:435– 439.
with AN had an abnormal QTc interval, which was the
10. Isner JM, Roberts WC, Heymsfield SB, Yager J. Anorexia nervosa and
evident after refeeding only.
sudden death. Ann Intern Med 1985;102:49 –52.
This study is the first to report a 24-hour mean QT 11. Rasmussen LH, Andersen T. The relationship between QTc changes and nutrition during weight loss after gastroplasty. Acta Med Scand 1985;21:271–275.
duration and variation in a population of patients with 12. Powers PS, Schocken DD, Feld J, Holloway JD, Boyd F. Cardiac function during
AN before and after refeeding. Recent studies have weight restoration in anorexia nervosa Int J Eating Disord 1991;10:521–530.
TABLE 2 QT and HRV Measurements Obtained Using 24-hour
Electrocardiographic Holter Monitoring in Patients With AN (Before and After
Refeeding) and Control Group
BRIEF REPORTS
279
13. Gottdiener JS, Gross HA, Henry WL, Borer JS, Ebert MH. Effects of
self-induced starvation on cardiac size and function in anorexia nervosa. Circulation 1978;58:421– 433.
14. Galetta F, Franzoni F, Cupisti A, Belliti D, Prattichizzo F, Rolla M. QT interval
dispersion in young women with anorexia nervosa. J Pediatr 2002;140:456 – 460.
15. Swenne I, Larsson PT. Heart risk associated with weight loss in anorexia
nervosa and eating disorders: risk factors for QTc interval prolongation and
dispersion. Acta Paediatr 1999;88:304 –309.
16. Swenne I. Heart risk associated with weight loss in anorexia nervosa and
eating disorders: risk factors for QTc interval prolongation and dispersion. Acta
Paediatr 2000;89:447– 452.
17. Kollai M, Bonyhay I, Jokkel G, Szonyi L. Cardiac vagal hyperactivity in
adolescent anorexia nervosa. Eur Heart J 1994;15:1113–1118.
18. Petretta M, Bonaduce D, Scalfi L, De Filippo E, Marciano F, Migaux ML,
Themistoclakis S, Ianniciello A, Contaldo F. Heart rate variability as a measure
of autonomic nervous system function in anorexia nervosa. Clin Cardiol 1997;
20:219 –224.
19. Rechlin T, Weis M, Ott C, Bleichner F, Joraschky P. Alterations of autonomic
cardiac control in anorexia nervosa. Biol Psychiatry 1998;43:358 –363.
20. Kreipe RE, Goldstein B, Deking DE, Tipton R, Kempski MH. Heart rate
power spectrum analysis of autonomic dysfunction in adolescents with anorexia
nervosa. Int J Eating Disord 1994;16:159 –165.
Left Ventricular Structure and Function in Sedentary
and Physically Active Subjects With Left Ventricular
Hypertrophy (the LIFE Study)
Kurt Boman, MD, PhD, Mona Olofsson, BS, Björn Dahlöf, MD, PhD, Eva Gerdts,
Markku S. Nieminen, MD, PhD, Vasilios Papademetriou, MD, PhD,
Kristian Wachtell, MD, PhD, and Richard B. Devereux, MD, PhD
Patients with left ventricular hypertrophy had higher
heart rates, left ventricular mass, and left atrial size
independent of gender, blood pressure, or body mass
index than physically active patients. 䊚2005 by
Excerpta Medica Inc.
(Am J Cardiol 2005;95:280 –283)
the echocardiographic substudy of the Losartan
for End Point Reduction in HypertenIsionnIntervention
(LIFE) study, we examined left ventricular (LV)
structure and function in sedentary and physically
active hypertensive patients with LV hypertrophy.
Our hypothesis was that physically active patients
would have more favorable LV structure and function
than sedentary patients.
•••
The present analysis was conducted in 958 patients
aged 55 to 80 years with stage II or III hypertension
enrolled in the LIFE echocardiographic substudy who
underwent yearly echocardiograms during 3 years of
antihypertensive treatment1– 4 and had data on exercise
status at baseline. Mean seated systolic and diastolic
blood pressures after a 1- to 2-week placebo run-in
period were 160/95 to 200/115 mm Hg. Exercise
status was obtained by questionnaire and was categorized into 3 groups: group 1 (sedentary): never exerFrom the Department of Medicine, Skellefteå County Hospital and
Umeå University, Skellefteå; the Department of Medicine, Sahlgrenska
University Hospital/Östra, and University of Gothenburg, Gothenburg, Sweden; the Department of Cardiology, Haukeland University
Hospital, Bergen, Norway; the Department of Cardiology, Helsinki
University Central Hospital, Helsinki, Finland; the Department of Cardiology, Veterans Administration Hospital, Washington, DC; the Department of Medicine, Glostrup University Hospital, Glostrup,
Denmark; and Weill Medical College of Cornell University, New
York, New York. This study was supported in part by a grant from
Merck & Co., Inc., West Point, Pennsylvania. Dr. Boman’s address is:
Skellefteå County Hospital and Umeå University, 93186 Skellefteå,
Sweden. E-mail: kurt.boman@vll.se. Manuscript received March 30,
2004; revised manuscript received and accepted September 1,
2004.
280
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MD, PhD,
cise; group 2: exercise ⱕ30 minutes twice per week;
and group 3 (physically active): exercise ⬎30 minutes
twice per week.
The LIFE echocardiographic substudy was carried
out in echocardiographic centers in Denmark, Finland,
Iceland, Norway, Sweden, the United Kingdom, and the
United States. Geographic distribution, blood pressure,
body mass index, and prevalences of diabetes and vascular disease resembled the entire LIFE population, with
the exception of enrolling more men and African-Americans. Patients gave informed consent, and ethical committees in participating countries approved this study.
Before enrollment, all patients had screening electrocardiograms showing LV hypertrophy by either genderadjusted Cornell voltage-duration product ⱖ2,440 mV
⫻ ms or Sokolow-Lyon voltage ⬎38 mV.5 Exclusion
criteria included myocardial infarction or stroke ⱕ6
months before the study, current heart failure or known
LV ejection fraction ⬍40%, significant aortic stenosis, or
overt renal insufficiency (serum creatinine ⬎160 ␮mol/L
or 1.8 mg/dl).
Echocardiographic procedures for this study have
been previously described.1– 4 End-diastolic LV dimensions were used to calculate LV mass by an anatomically
validated formula (r ⫽ 0.90 vs necropsy LV mass).6 LV
hypertrophy was considered present when LV mass/
body surface area was ⬎116 g/m2 for men and ⬎104
g/m2 for women.7 Relative wall thickness, calculated as
diastolic posterior wall thickness/LV internal radius,8
was increased when this ratio was ⬎0.430.9 Normal
geometry was present when LV mass index and relative
wall thickness were normal, increased relative wall
thickness and normal LV mass index identified concentric LV remodeling, increased LV mass index with normal relative wall thickness identified eccentric LV hypertrophy, and increases of the 2 variables identified
concentric LV hypertrophy.10
Endocardial fractional shortening (percent) was
calculated from LV internal dimensions in diastole
and systole.11 To assess LV contractility, we used the
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.019
TABLE 1 Patient Characteristics According to Exercise Status
Variable
Age (yrs)
Women
Body mass index (kg/m2)
Systolic blood pressure (mm Hg)
Diastolic blood pressure (mm Hg)
Pulse pressure (mm Hg)
Heart rate (beats/min)
Smoker
Framingham risk score (%)
Diabetes mellitus
Prevalence of myocardial infarction
and/or angina pectoris
Cerebral vascular disease
Peripheral vascular disease
Group 1
(n ⫽ 212)
Group 2
(n ⫽ 236)
Group 3
(n ⫽ 510)
ANOVA
p Value
66.4 ⫾ 7.5
42.5%
28.6 ⫾ 5.2
174 ⫾ 15
99 ⫾ 10
75 ⫾ 16
74 ⫾ 11
26.4%
23.8 ⫾ 9.6
17.9%
15.1%
66.2 ⫾ 7.2
46.8%
27.6 ⫾ 5
174 ⫾ 14
98 ⫾ 9
76 ⫾ 16
74 ⫾ 11
19.1%
22.9 ⫾ 9.6
10.2%
12.8%
65.6 ⫾ 6.6
38.4%
26.6 ⫾ 3.6
173 ⫾ 14
98 ⫾ 9
75 ⫾ 15
71 ⫾ 11
18.0%
22.6 ⫾ 9.2
9.0%
13.1%
0.285
0.087
⬍0.001
0.600
0.690
0.777
⬍0.001
0.035
0.257
0.002
0.727
12.7%
8.0%
10.6%
4.3%
5.5%
4.9%
0.002
0.158
Group 1 (sedentary), never exercise; group 2, exercise ⱕ30 minutes twice per week; group 3 (physically active), exercise ⬎30 minutes twice per week. ANOVA ⫽
analysis of variance.
relation between mid-wall shortening and mid-wall
circumferential end-systolic stress measured at the
level of the LV minor axis,12,13 using a previously
reported method to identify the location in the LV wall
at end-systole of the surface between the inner and
outer myocardial shell volumes at end-diastole.14 Circumferential end-systolic stress, as the primary measure of myocardial afterload, was estimated at the
mid-wall from LV linear dimensions using a cylindrical model.13,15 Previously published equations relating
endocardial and mid-wall shortening to circumferential end-systolic stress in 140 normotensive adults
were used to derive predicted endocardial fractional
and mid-wall shortening.12 Stress-corrected endocardial and midwall shortening were then calculated as
the ratios to the predicted value.
SPSS version 11.0.1 (SPSS, Inc., Chicago, Illinois)
was used for data management and analysis. Continuous variables are expressed as mean ⫾ SD and
categorical data as percentages. Differences in continuous and categorical variables among the 3 exercise
groups were assessed by 1-way analysis of variance
followed by the Scheffé post hoc test and chi-square
analyses, respectively. Bivariate correlations between
physical activity and LV geometric variables were
assessed with Spearman’s correlation coefficients. A
full factorial model was used for analysis of covariance; multiple simple contrasts were obtained after
adjustment of the confidence intervals with Sidak’s
method, whereby main effects were adjusted for potential confounders (body mass index, heart rate,
smoking, and the effects of coexisting disease as diabetes or previous cerebral vascular disease). Twotailed p ⬍0.05 was considered statistically significant.
A total of 958 patients who had echocardiograms
and provided information on exercise habits at enrollment in the LIFE Study constituted the present study
population (Table 1).
No significant difference (p ⫽ 0.081) in exercise
categories was observed between the 397 women and
561 men. Group 1 included 90 women and 122 men,
group 2 included 111 women and 125 men, and group
3 included 196 women and 314 men.
Body mass index was significantly greater in groups
1 and 2 than in group 3 (28.6 and 27.6 vs 26.6 kg/m2,
p ⬍0.001 and p ⫽ 0.008, respectively), without a significant difference between groups 1 and 2 (p ⫽ 0.067).
Body surface area was larger in group 1 than in group 3
(1.92 vs 1.89 m2, p ⬍0.001) but did not differ between
groups 1 and 2 (p ⫽ 0.067). Mean body weight was
greater in group 1 than in groups 2 and 3 (82.0 vs 78.4
and 76.9 kg, p ⫽ 0.024 and p ⬍0.001, respectively) but
was statistically indistinguishable between groups 2 and
3 (p ⫽ 0.385). There were no significant differences
among the groups in height or systolic and diastolic
blood pressures.
LV mass was greater in group 1 than in groups 2 and
3 (p ⫽ 0.015 and p ⫽ 0.040, respectively), without a
significant difference between groups 2 and 3. LV mass/
height2.7 was greater in group 1 than in group 3 (p ⫽
0.004), but LV mass/body surface area did not differ
significantly among the groups (Table 2). Heart rate
during the echocardiograms was lower in group 3 than in
group 1 (p ⫽ 0.001). Interventricular septal thickness
was lower in group 3 than in group 1 (p ⫽ 0.012), with
an intermediate mean value in group 2. The LV posterior
wall was thicker in group 1 than in groups 2 and 3 (p ⫽
0.045 and p ⫽ 0.016, respectively). Left atrial diameter
was larger in group 1 than in groups 2 and 3 (p ⫽ 0.006
and p ⫽ 0.005, respectively). No differences were found
in LV internal dimension or relative wall thickness
among the 3 groups, as a result of which there was also
no significant intergroup difference in the patterns of LV
geometry.
There were no significant differences among the 3
groups in LV systolic function measured as endocardial fractional shortening, ejection fraction, circumferential systolic stress, or mid-wall shortening. The
atrial filling fraction was greater in group 1 than in
groups 2 and 3 (p ⫽ 0.003 and p ⫽ 0.018, respectively), but there were no significant differences in
other measures of LV diastolic function among the
BRIEF REPORTS
281
TABLE 2 LV Structure and Function and Left Atrial Size According to Exercise Status
Group 1
(n ⫽ 212)
Variable
Interventricular septum thickness (cm)
Posterior wall thickness (cm)
LV internal dimension in diastole (cm)
LV mass (g)
LV mass/surface area (g/m2)
LV mass/height2.7 (g/m2.7)
Relative wall thickness
Left atrial dimension (cm)
Fractional shortening (%)
Ejection fraction (%)
Midwall shortening (%)
Circumferential end-systolic stress (kdynes/cm2)
Stress-corrected mid-wall shortening (%)
Isovolumic relaxation time (ms)
Mitral valve E/A ratio
Mitral valve deceleration time (ms)
Atrial filling fraction
Stroke volume (ml)
Cardiac output (ml/min)
Cardiac index (ml/min/m2)
Total peripheral resistance (dynes · sec · cm⫺5)
Total peripheral resistance index (dynes · sec · cm⫺2)
1.18
1.09
5.3
244
126
58.8
0.42
4.1
33.1
60.8
15.2
182
95.7
116
0.86
211
0.44
76.3
5,313
2,782
1,948
3,696
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
0.2
0.2
0.6
68
30
15.7
0.08
0.6
5.9
8.7
2.3
53
13.7
24
0.41
63
0.12
17.6
1,378
701
609
1,120
Group 2
(n ⫽ 236)
1.16
1.06
5.2
229
122
56
0.41
3.9
33.4
61.2
15.3
182
96.1
113
0.84
210
0.43
76.2
5,254
2,815
1,962
3,659
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
0.1
0.1
0.6
51
25
12
0.07
0.6
6.1
8.9
2.2
48
13.1
23
0.41
64
0.11
16.3
1,274
664
552
1,052
Group 3
(n ⫽ 510)
ANOVA
p Value
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
0.012
0.011
0.249
0.011
0.208
0.004
0.123
0.001
0.823
0.806
0.120
0.730
0.109
0.303
0.685
0.102
0.001
0.052
0.241
0.384
0.388
0.518
1.15
1.06
5.3
232
123
55.3
0.41
3.9
33.4
61.3
15.6
185
97.7
115
0.87
220
0.41
79.1
5,136
2,739
2,010
3,758
0.1
0.1
0.6
53
24
11.4
0.06
0.5
5.6
8.1
2
48
12.5
23
0.33
67
0.1
17.4
1,252
666
558
1,028
Explanation of groups in Table 1. Abbreviation as in Table 1.
groups, including the isovolumic relaxation time, E- or
A-wave velocities, E/A ratio, and deceleration time. No
difference was found in stroke volume, cardiac output,
and total peripheral resistance among the groups.
Multivariate analysis in the general linear model
showed that LV mass was greater in group 1 than in
groups 2 and 3 (adjusted mean 241 vs 229 and 232 g, p
⫽ 0.032), independent of body mass index, male gender,
and pulse pressure (all p ⬍0.001). In similar models,
ventricular septal thickness was significantly (p ⫽ 0.007)
less in physically active than in sedentary patients, as
was LV posterior wall thickness (p ⫽ 0.015), whereas
the adjusted mean atrial filling fraction was smaller in
physically active than in sedentary patients (p ⫽ 0.026).
With addition of age to these covariates, mean adjusted
LV mass was smaller in patients in group 2 and tended
to be smaller in group 3 than in group 1 (229 and 232 vs
241 g, p ⬍0.05 and p ⫽ 0.12, respectively). In similar
models, interventricular septal thickness was smaller in
physically active than sedentary patients (p ⫽ 0.015), as
was LV posterior wall thickness (1.06 vs 1.09 cm, p ⫽
0.024). The adjusted mean atrial filling fraction remained
significantly smaller in physically active than in sedentary patients (0.41 vs 0.44, p ⫽ 0.04).
•••
To our knowledge, this is the largest study showing
that patients with hypertension and sedentary lifestyles have increased LV mass due to larger LV wall
thicknesses. It extends the results of smaller studies
that reported the regression of pathologic LV hypertrophy with exercise training in selected hypertensive
patients.16 The finding that LV mass was greater in
sedentary patients independent of the effect of body
mass index is especially important in view of the
strong association between relative body weight and
arterial pressure and the known benefit of weight loss
282 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
for regression of LV hypertrophy. There was a significant association between body mass index and LV
mass. Hinderliter and coworkers17 found that exercise
and weight loss reduced blood pressure and induced
favorable changes in LV structure in a cohort of
overweight, sedentary men and women. Blood pressure decreased by 7/6 mm Hg in the weight management group and by 3/4 mm Hg in the aerobic exercise
group. In association with these decreases in blood
pressure, participants in the intervention groups had
significant decreases in relative wall thickness (p ⫽
0.003), posterior wall thickness (p ⫽ 0.05), and septal
thickness (p ⫽ 0.004), with a trend toward a decrease
in indexed LV mass (p ⫽ 0.08) relative to the control
group.17 In our study, the difference between physically active and sedentary hypertensive patients in LV
mass was due to greater ventricular septal and posterior wall thicknesses, without a significant betweengroup difference in LV chamber size. In the present
study, the greater LV hypertrophy, due to increased
LV wall thicknesses, in sedentary patients was not
associated with abnormalities of LV systolic function.
In contrast, the atrial filling fraction of diastolic transmitral blood flow was significantly greater in sedentary patients.
The abnormalities of LV geometry and filling associated with a sedentary lifestyle among patients in
the LIFE Study occurred despite the absence of significant differences among groups in arterial blood
pressure. Although hypertension is the most important
cause of LV hypertrophy, the relation between the
degree of LV hypertrophy and blood pressure at rest is
generally weak and less close than that with pressure
measured by ambulatory monitor. Papademetriou et
al18 found that systolic blood pressure at a low level of
exercise (5 METs), corresponding to usual daily acJANUARY 15, 2005
tivities, was the closest correlate of LV hypertrophy in
hypertensive patients. Endurance exercise training can
reduce blood pressure in older adults with mild
(grade I) hypertension. It also improves exercise capacity and the quality of life and can induce a modest
but significant regression of LV hypertrophy and remodeling in older hypertensive adults.
1. Wachtell K, Rokkedal J, Bella JN, Aalto T, Dahlöf B, Smith G, Roman MJ,
Ibsen H, Aurigemma GP, Devereux RB. Effect of electrocardiographic left
ventricular hypertrophy on left ventricular systolic function in systemic hypertension (the LIFE Study). Am J Cardiol 2001;87:54 – 60.
2. Wachtell K, Bella JN, Liebson PR, Gerdts E, Dahlöf B, Aalto T, Roman MJ,
Papademetriou V, Ibsen H, Rokkedal J, et al. Impact of different partition values
on prevalences of left ventricular hypertrophy and concentric geometry in a large
hypertensive population: the LIFE study. Hypertension 2000;35:6 –12.
3. Wachtell K, Smith G, Gerdts E, Dahlöf B, Nieminen MS, Papademetriou V,
Bella JN, Ibsen H, Rokkedal J, Devereux RB. Left ventricular filling patterns in
patients with systemic hypertension and left ventricular hypertrophy (the LIFE
study). Am J Cardiol 2000;85:466 – 472.
4. Devereux RB, Bella JN, Boman K, Gerdts E, Nieminen MS, Rokkedal J,
Papademetriou V, Wachtell K, Wright J, Paranicas M, et al. Echocardiographic
left ventricular geometry in hypertensive patients with electrocardiographic left
ventricular hypertrophy: the LIFE Study. Blood Press 2001;10:74 – 82.
5. Dahlöf B, Devereux RB, de Faire U, Fyhrquist F, Hedner T, Ibsen H, Julius S,
Kjeldsen S, Kristianson K, Lederballe-Pedersen O, et al. The Losartan Intervention For Endpoint Reduction (LIFE) in Hypertension Study. Am J Hypertens
1997;10:705–713.
6. Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, Reichek
N. Echocardiographic assessment of left ventricular hypertrophy: comparison to
necropsy findings. Am J Cardiol 1986;57:450 – 458.
7. Devereux RB, Dahlöf B, Levy D, Pfeffer MA. Comparison of enalapril versus
nifedipine to decrease left ventricular hypertrophy in systemic hypertension (the
PRESERVE trial). Am J Cardiol 1996;78:61– 65.
8. Reichek N, Devereux RB. Reliable estimation of peak left ventricular systolic
pressure by M-mode echographic-determined end-diastolic relative wall thickness: identification of severe valvular aortic stenosis in adult patients. Am Heart J
1982;103:202–209.
9. Roman MJ, Pickering TG, Schwartz JE, Pini R, Devereux RB. Association of
carotid atherosclerosis and left ventricular hypertrophy. J Am Coll Cardiol
1995;25:83–90.
10. Ganau A, Devereux RB, Roman MJ, de Simone G, Pickering TG, Saba PS,
Vargiu P, Simongini I, Laragh JH. Patterns of left ventricular hypertrophy and
geometric remodeling in essential hypertension. J Am Coll Cardiol 1992;19:
1550 –1558.
11. Gutgesell HP, Paquet M, Duff DF, McNamara DG. Evaluation of left
ventricular size and function by echocardiography. Results in normal children.
Circulation 1977;56:457– 462.
12. de Simone G, Devereux RB, Roman MJ, Ganau A, Saba PS, Alderman MH,
Laragh JH. Assessment of left ventricular function by the midwall fractional
shortening/end-systolic stress relation in human hypertension. J Am Coll Cardiol
1994;23:1444 –1451.
13. Gaasch WH, Zile MR, Hoshino PK, Apstein CS, Blaustein AS. Stressshortening relations and myocardial blood flow in compensated and failing canine
hearts with pressure-overload hypertrophy. Circulation 1989;79:872– 883.
14. Shimizu G, Hirota Y, Kita Y, Kawamura K, Saito T, Gaasch WH. Left
ventricular midwall mechanics in systemic arterial hypertension. Myocardial
function is depressed in pressure-overload hypertrophy. Circulation 1991;83:
1676 –1684.
15. Gaasch WH, Battle WE, Oboler AA, Banas JS Jr, Levine HJ. Left ventricular
stress and compliance in man. With special reference to normalized ventricular
function curves. Circulation 1972;45:746 –762.
16. Hagberg JM, Park JJ, Brown MD. The role of exercise training in the
treatment of hypertension: an update. Sports Med 2000;30:193–206.
17. Hinderliter A, Sherwood A, Gullette EC, Babyak M, Waugh R, Georgiades
A, Blumenthal JA. Reduction of left ventricular hypertrophy after exercise and
weight loss in overweight patients with mild hypertension. Arch Intern Med
2002;162:1333–1339.
18. Papademetriou V, Notargiacomo A, Sethi E, Costello R, Fletcher R, Freis ED.
Exercise blood pressure response and left ventricular hypertrophy. Am J Hypertens 1989;2:114 –116.
Comparison of Angiotensin-Converting Enzyme
Inhibitors in the Treatment of Congestive Heart Failure
Karen Tu,
MD, MSc,
Muhammad Mamdani, PharmD,
Douglas Lee, MD
The effectiveness of the different angiotensin-converting enzyme (ACE) inhibitors in the treatment of patients with congestive heart failure (CHF) was compared by performing a retrospective cohort study
using linked administrative databases on elderly patients admitted to the hospital for the treatment of
CHF. Relative to those initiated on enalapril, no significant differences in the combined end point of readmission to the hospital for CHF or mortality were
observed among users of lisinopril, ramipril, or other
From the Institute for Clinical Evaluative Sciences, Toronto, Ontario;
University Health Network, Toronto Western Hospital Family Medicine
Centre, Toronto, Ontario; University Health Network, Division of Cardiology, Toronto, Ontario; and the Departments of Family and Community Medicine, Family Healthcare Research Unit, Health Policy
Management and Evaluation, Pharmacy, and Medicine, University of
Toronto, Toronto, Ontario, Canada. Dr. Lee holds research fellowships
from the Heart and Stroke Foundation of Canada and the Canadian
Institutes of Health Research, Ottawa, Ontario, Canada. Dr. Tu’s
address is: Institute for Clinical Evaluative Sciences, G106-2075
Bayview Avenue, Toronto, Ontario M4N 3M5, Canada. E-mail:
karen.tu@ices.on.ca. Manuscript received July 8, 2004; revised
manuscript received and accepted August 26, 2004.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MA, MPH,
Alex Kopp,
BA,
and
ACE inhibitors. In terms of effectiveness for the treatment of patients with CHF, the findings of this study
suggest a class effect among ACE inhibitors. 䊚2005
by Excerpta Medica Inc.
(Am J Cardiol 2005;95:283–286)
ngiotensin-converting enzyme (ACE) inhibitors
have been demonstrated to reduce morbidity and
A
mortality in the treatment of patients with congestive
heart failure (CHF).1 Although the classic and strongest evidence supporting the use of ACE inhibitors in
CHF has involved 1 specific agent, enalapril,2,3 the
findings are often extrapolated to other ACE inhibitors, and a class effect is often presumed despite
differences in pharmacokinetic and pharmacodynamic
properties among agents.4,5 The publication of the
Heart Outcomes Prevention Evaluation Study,6 a
study showing the cardiovascular protective effects of
ramipril, resulted in a dramatic increase in the use of
ramipril in patients with CHF, although such patients
were excluded from the trial.7 Furthermore, there have
been no published head-to-head trials comparing the
effectiveness of the different ACE inhibitors in pa0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.097
283
TABLE 1 Hospitalizations, Procedures, and Drug Use History
Variable
Enalapril
(n ⫽ 2,976)
Lisinopril
(n ⫽ 1,192)
Ramipril
(n ⫽ 1,138)
Other ACE Inhibitors
(n ⫽ 1,447)
79.6
54%
6%
78.4
54%
4%
78.7
51%
5%
79.3
53%
6%
7%
15%
3%
4%
8%
16%
3%
4%
9%
15%
4%
4%
7%
15%
3%
5%
1%
⬍1%
20%
2%
⬍1%
2%
1%
23%
3%
⬍1%
3%
1%
25%
4%
⬍1%
1%
⬍1%
22%
2%
⬍1%
3%
5%
11%
18%
2%
1%
23%
21%
31%
17%
16%
⬍1%
21%
9%
29%
25%
18%
1%
2%
4%
6%
13%
19%
2%
⬍1%
22%
23%
30%
18%
15%
⬍1%
20%
12%
26%
26%
19%
⬍1%
2%
4%
5%
12%
18%
2%
1%
23%
24%
32%
16%
14%
⬍1%
18%
12%
27%
25%
15%
⬍1%
1%
3%
5%
11%
19%
2%
⬍1%
24%
19%
34%
17%
15%
⬍1%
22%
9%
31%
25%
19%
1%
2%
Age, mean (yrs)
Women
Long-term care
Hospitalizations
Acute myocardial infarction
Coronary heart disease
Renal failure
Stroke
Procedures
Coronary angiography
Coronary bypass
Echocardiography
Coronary angioplasty
Cardiac valve surgery
Previous drug use
␣ blockers
Antiarrhythmics
Anticoagulants
Antidiabetics
Antiplatelets
Antirheumatics
Aspirin
␤ blockers
Calcium channel blockers
Digoxin
Diuretics
Hydralazine
Inhalers
Lipid-lowering drugs
Loop diuretics
Nitrates
Nonsteroidal anti-inflammatory drugs
Other antihypertensives
Spironolactone
TABLE 2 Readmissions or Mortality, Readmissions Alone, or Mortality Alone
Variable
Readmissions or mortality
Follow-up time, mean (d)
Death or readmission
Death or readmission per
100 person-yrs
Crude HR (95% CI)
Adjusted HR (95% CI)
Readmissions
Follow-up time, mean (d)
Readmission
Readmission per 100
person-yrs
Crude HR (95% CI)
Adjusted HR (95% CI)
Mortality
Follow-up time, mean (d)
Death
Death per 100 person-yrs
Crude HR (95% CI)
Adjusted HR (95% CI)
Enalapril
259
669 (22%)
31.8
1.00
1.00
259
390 (13%)
18.5
1.00
1.00
280
360 (12%)
15.8
1.00
1.00
Lisinopril
Ramipril
286
290 (24%)
31.0
285
266 (23%)
30.0
1.00 (0.88–1.15)
1.08 (0.94–1.23)
286
175 (15%)
18.7
0.97 (0.84–1.12)
1.06 (0.92–1.24)
285
170 (15%)
19.1
Other ACE
Inhibitors
247
324 (22%)
33.1
1.04 (0.91–1.19)
1.02 (0.90–1.17)
247
211 (15%)
21.6
1.06 (0.88–1.26)
1.11 (0.92–1.32)
1.08 (0.90–1.29)
1.19 (0.99–1.45)
1.15 (0.97–1.36)
1.13 (0.96–1.34)
313
158 (13%)
15.5
0.98 (0.81–1.18)
1.09 (0.90–1.31)
311
137 (12%)
14.1
0.90 (0.74–1.09)
0.97 (0.78–1.20)
268
158 (11%)
14.9
0.95 (0.78–1.14)
0.94 (0.78–1.13)
CI ⫽ confidence interval; HR ⫽ hazard ratio.
tients with CHF. Whether the effects are unique to
specific ACE inhibitors or all ACE inhibitors show
equivalent benefits has not been firmly established.8,9
284 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
The objective of this study was to examine the relative
association between ACE inhibitors and readmission
and mortality rates of patients with CHF.
JANUARY 15, 2005
•••
We performed a retrospective cohort study using
administrative databases, which were linked anonymously through encrypted unique patient identifiers,
on the ⬎1.4 million elderly residents in Ontario. We
identified patients discharged alive from the hospital
with CHF from April 1, 1997, to March 31, 2000,
using the International Classification of Diseases,
Ninth Revision, Clinical Modification code 428.x recorded in the Canadian Institute for Health Information hospital discharge database, which contains detailed information on all hospitalizations in Ontario. A
recent chart abstraction study validating the accuracy
of coding for CHF admission in our databases revealed positive predictive values of 90% to 96%,
depending on the criteria used to define CHF.10
We excluded patients ⬍66 years of age and those
with admissions for CHF in the 3 years before the
index date. To minimize the potential for survival
treatment selection bias,11 we limited our analysis to
patients surviving ⱖ30 days after discharge and those
initiated on ACE inhibitors during this interval. To
minimize the potential for channeling bias, we limited
our analysis to ACE inhibitor users. Furthermore, we
excluded patients who received ACE inhibitors or
angiotensin II receptor blockers in the year before the
hospital admission to limit our analysis to treatmentnaive patients. Enalapril users served as the reference
group. The Ontario Drug Benefit database records
prescriptions filled by elderly residents aged ⱖ65
years in Ontario and served as the source of drug use
data. Patients were divided into those initiated on
enalapril, lisinopril, ramipril, or other ACE inhibitors
(benazepril, captopril, cilazapril, fosinopril, perindopril, quinapril, and trandolapril).
For each of the study drug cohorts, we defined the
duration of exposure as the period of continuous,
exclusive enrollment in the study medication group
after the index date. The “days supply” variable of the
pharmacy claims database allowed us to estimate the
intended duration of each prescription. If patients
were dispensed drugs before the end of this period, the
excess drug supplies were carried over to the next
prescription’s day supply estimation. Patients were
allowed a 20% grace period on the previous day
supply to refill the next prescriptions. If they did not
refill their prescriptions for the study drugs within
these successive time windows, they were deemed to
have discontinued the study drugs.
The primary outcome of interest was the combined
end point of readmission for CHF as a primary diagnosis or mortality. The secondary outcomes were CHF
readmission alone and mortality alone. Patients were
followed for ⱕ2 years and were censored if they were
readmitted for CHF, died, switched medications, discontinued the initial ACE inhibitor, or reached the end
of the observation period (March 31, 2002).
Time-to-event analyses were conducted for CHF
hospitalization using Cox proportional-hazards models with the enalapril group as the reference, controlling for all covariates outlined in Table 1, as well as
the fiscal quarter of the index date to account for
temporal effects. As an overall measure of co-morbidity, we controlled for the number of distinct drugs
dispensed in the year before the index date,12 the
Charlson co-morbidity index,13 and the presence of
diabetes.14 The proportional-hazards assumption for
each exposure variable was assessed in each analysis.
All analyses were performed using SAS for Unix,
version 8.2 (SAS Institute Inc., Cary, North Carolina).
All statistical tests were performed at the 5% level of
significance and were 2 sided.
We identified 6,753 patients with newly diagnosed
CHF admitted to the hospital who survived ⱖ30 days
after discharge and were initiated on ACE inhibitors
during this time. The average age of the patients receiving the different types of ACE inhibitors was 79.2 years,
and in total, 53.3% of the cohort was women and 5.5%
were long-term care patients (Table 1). There did not
appear to be any consistently meaningful clinical differences among the study drug groups with respect to the
covariates, as outlined in Table 1.
Relative to enalapril users, no significant differences in the combined end point of readmission for
CHF or mortality were observed in users of lisinopril
(adjusted hazard ratio 1.08, 95% confidence interval
0.94 to 1.23), ramipril (adjusted hazard ratio 1.06,
95% confidence interval 0.92 to 1.24), or the other
ACE inhibitors (adjusted hazard ratio 1.02, 95% confidence interval 0.90 to 1.17). No significant differences among groups were observed in the secondary
end points of readmission for CHF or mortality independently (Table 2).
•••
We observed no significant difference among the
various ACE inhibitors for patients with newly diagnosed CHF initiated on ACE inhibitors in terms of
CHF readmission or mortality, readmission alone, or
mortality alone. Regardless of the varying pharmacologic properties of ACE inhibitors,5 including differences in chemical structure, lipophilicity, half-life,
and tissue-binding capacity, our results suggest a class
effect of ACE inhibitors in the treatment of patients
with CHF.
Several limitations must be acknowledged. Our
analysis considered the duration of the prescriptions,
but we were unable to determine actual patient compliance with their medications or dosages taken. Our
databases did not contain important information on
other variables that may have been associated with the
outcome of interest, including body mass index, dietary salt intake, alcohol consumption, hypertension,
and smoking. It is possible that patients varied in the
proportion of left ventricular dysfunction or the cause
of CHF (ischemic, nonischemic, diastolic, systolic),
but it is unlikely that this measurement was substantially different among the different drug groups. Although bias is often problematic in observational studies, the distribution of the observed clinical covariates
among the study groups was very similar, and the
unadjusted and adjusted hazard ratios were virtually
identical, implying that this was probably not a major
concern for this study. In attempts to minimize bias in
our study through strict exclusion criteria, the number of
BRIEF REPORTS
285
patients in each study group may have been relatively
low, which may have compromised statistical power in
detecting a clinically meaningful difference. However,
given the results of our study, we estimate ⱖ80% power
to detect a clinically meaningful relative risk reduction of
30% in the primary outcome among groups. Our study is
the first direct comparison of ACE inhibitors in terms of
heart failure effectiveness. Our results suggest no significant differences among patient ACE inhibitors in clinically meaningful outcomes for treating patients with
CHF. Thus, when prescribing ACE inhibitors, consideration should likely be given to dosing convenience and cost.
1. Garg R, Yusuf S. Overview of randomized trials of angiotensin-converting
enzyme inhibitors on mortality and morbidity in patients with heart failure. JAMA
1995;273:1450 –1456.
2. The CONSENSUS Trial Study Group. Effects of enalapril on mortality in
severe congestive heart failure. N Engl J Med 1987;316:1429 –1435.
3. The SOLVD Investigators. Effect of enalapril on survival in patients with
reduced left ventricular ejection fractions and congestive heart failure. N Engl
J Med 1992;395:293–302.
4. Kostis JB. Pharmacological differentiation of angiotensin-converting enzyme
inhibitors. J Human Hyperten 1989;3:119 –125.
5. Brown NJ, Vaughan DE. Angiotensin-converting enzyme inhibitors. Circula-
tion 1998;97:1411–1420.
6. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. Effects of an
angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in
high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med 2000;342:145–153.
7. Tu K, Mamdani MM, Jacka RM, Forde NJ, Rothwell DM, Tu JV. The striking
effect of the Heart Outcomes Prevention Evaluation (HOPE) on ramipril prescribing in Ontario. Can Med Assoc J 2003;168:553–557.
8. Furberg CD, Pitt B. Are all angiotensin-converting enzyme inhibitors interchangeable? J Am Coll Cardiol 2001;37:1456 –1460.
9. Sica DA. Class effects of angiotensin-converting enzyme inhibitors. Am J
Manag Care 2000;6:S85–S108.
10. Jong P, Gong Y, Liu PP, Austin PC, Lee DS, Tu JV. Care and outcomes of
patients newly hospitalized for heart failure in the community treated by cardiologists compared with other specialists. Circulation 2003;108:184 –191.
11. Glesby MJ, Hoover DR. Survivor treatment selection bias in observational
studies: examples from the AIDS literature. Ann Intern Med 1996;124:999 –1005.
12. Schneeweiss S, Seeger JD, Maclure M, Wang PS, Avorn J, Glynn RJ.
Performance of comorbidity scores to control for confounding in epidemiologic
studies using claims data. Am J Epidemiol 2001;154:854 – 864.
13. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of
classifying prognostic comorbidity in longitudinal studies: development and
validation. J Chron Dis 1987;40:373–383.
14. Hux JE, Ivis F, Flintoft V, Bica A. Diabetes in Ontario: determination of
prevalence and incidence using a validated administrative data algorithm. Diabetes Care 2002;25:512–516.
Simplified Peak Power Reserve in Patients With an
Implantable Cardioverter-Defibrillator and Advanced
Heart Failure
William T. Katsiyiannis, MD, Alan D. Waggoner, MHS, Benico Barzilai, MD,
Brian F. Gage, MD, MS, Jose M. Sanchez, MD, Joseph G. Rogers, MD,
Bruce D. Lindsay, MD, and Marye J. Gleva, MD
The prognostic ability of simplified peak power (SPP)
reserve, a novel measure of left ventricular systolic
performance, was prospectively studied in patients
with advanced heart failure (HF) and implantable
cardioverter-defibrillators. Reduced SPP reserve identified patients who are at high risk for experiencing
progressive HF. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:286 –288)
ome patients with left ventricular dysfunction and
ventricular arrhythmias derive limited benefit
S
from the placement of implantable cardioverter-defibrillators (ICDs), because they die of progressive
heart failure (HF). The accurate identification of this
group could guide the selection of patients who may
benefit from more directed HF therapy. Unfortunately,
the tools that are currently available to stratify this
population are imprecise. Patients with the most advanced HF symptoms are more likely to die from
From the Cardiovascular Division, Department of Internal Medicine,
Washington University School of Medicine, St. Louis, Missouri. Dr. Katsiyiannis was supported by the Michael Bilitch fellowship of the North
American Society of Pacing and Electrophysiology, Natick,
Massachusetts. Dr. Katsiyiannis’s address is: Minneapolis Heart Institute,
920 East Twenty-Eighth St., Suite 300, Minneapolis, Minnesota 55407.
E-mail: wkatsiyiannis@mplsheart.com. Manuscript received June 2,
2004; revised manuscript received and accepted September 8, 2004.
286
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
progressive pump failure as opposed to arrhythmic
death. Contractile reserve determined by invasively
measured peak power has been used as a prognostic
indicator in patients with HF.1 Ventricular reserve
using noninvasive simplified peak power (SPP) is
easily obtained, afterload independent, can be preload
adjusted, and may add additional predictive power to
traditional prognostic measures. We conducted a prospective study to determine whether noninvasive SPP
reserve can identify ICD candidates with rapidly progressive HF. Our hypothesis was that in patients with
ICDs in New York Heart Association functional class
III HF, those with limited SPP reserve would have a
greater number of adverse HF events.
•••
We prospectively enrolled patients in New York
Heart Association class III HF who underwent ICD
implantation for an American College of Cardiology–
American Heart Association class I indication at
Barnes-Jewish Hospital–Washington University School
of Medicine. Patients with either ischemic or nonischemic cardiomyopathy were included. Patients were excluded for the inability to complete a dobutamine stress
echocardiogram (i.e., the development of chest pain,
arrhythmia, or hypotension or hypertension necessitating
the early discontinuation of the study). A cohort of agematched control patients without a history of HF or
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.020
SPP was calculated according to
the methods described by Armstrong
SPP Reserve
and colleagues,2 where peak power
is equal to the product of peak aortic
⬍1.5
⬎1.5
flow and mean arterial pressure. Peak
Characteristic
(n ⫽ 8)
(n ⫽ 10)
p Value
aortic flow was defined as the prodAge (yrs)
66
58
0.41
uct of peak aortic velocity and aortic
African-American
0 (0%)
2 (20%)
0.48
annulus area. Aortic velocity was
Men
6 (75%)
7 (70%)
0.81
Hypertension
3 (38%)
8 (80%)
0.145
measured by continuous-wave DoppDiabetes mellitus
4 (50%)
6 (60%)
0.67
ler in the apical 5-chamber view. The
Coronary artery disease
7 (88%)
7 (70%)
0.59
aortic annulus diameter was meaLeft ventricular ejection fraction (%)
27 ⫾ 13
37 ⫾ 8
0.06
sured from the parasternal long-axis
Mitral deceleration time (ms)
166 ⫾ 42
221 ⫾ 81
0.11
Sodium (mmol/L)
140 ⫾ 3
139 ⫾ 3
0.53
view at rest. Mean arterial pressure
Creatinine (mg/dl)
2.0 ⫾ 1.4
1.1 ⫾ 0.2
0.08
was obtained by a sphygmomanomAngiotensin-converting enzyme inhibitor
8 (100%)
9 (90%)
1.0
eter at the brachial artery. SPP reor angiotensin receptor blocker
serve, as described in the following
␤-adrenergic blocker
5 (63%)
5 (50%)
0.66
equations, equals the difference beAntiarrhythmic
3 (38%)
4 (40%)
1.0
Digoxin
4 (50%)
5 (50%)
1.0
tween SPP at maximal dobutamine
Diuretic
8 (100%)
9 (90%)
1.0
stress and SPP at rest divided by the
2
0.7 ⫾ 0.4
5.7 ⫾ 2.7
0.0001
SPP reserve (W/ml )
square of end-diastolic volume, an
estimate of preload: (1) SPP reserve
⫽ SPP maximum dobutamine ⫺
SPP baseline; (2) SPP ⫽ aortic flow ⫻ mean arterial
pressure; (3) aortic flow ⫽ aortic annulus area ⫻ peak
aortic velocity; and (4) mean arterial pressure ⫽ ([2 ⫻
systolic pressure] ⫹ diastolic pressure)/3.
Patients were followed in the Washington University Medical Center Arrhythmia Clinic every 3
months. At each follow-up visit, clinical information
was obtained, including a history of shocks, ICD
interrogation, HF hospitalizations, or cardiac transplantation. Mortality data were collected from hospital
records and family interviews.
The composite end point of HF hospitalizations,
cardiac transplantation, and all-cause mortality was
analyzed by the Kaplan-Meier method. Analyses were
performed with SPP reserve dichotomized at 1.5
W/ml2. This value was chosen a priori on the basis of
a previous study.1 Continuous variables were compared using the unpaired Student’s t test, and categorFIGURE 1. Kaplan-Meier curves for the composite end point in
ical variables were compared using Fisher’s exact test.
patients with ICDs and New York Heart Association class III HF.
Analyses were performed using SPSS version 10.0 for
Windows (SPSS, Inc., Chicago, Illinois) statistical
software.
Twelve age-matched normal control patients unarrhythmia was also enrolled and underwent noninvasive
SPP reserve measurements only. The Human Studies derwent dobutamine stress echocardiography and
Committee of Washington University School of Medi- the determination of SPP reserve. Their mean age
cine approved all aspects of the study, and written in- was 61 years. Their mean baseline ejection fraction
formed consent was obtained from all participants before was 73 ⫾ 11%, and their mean SPP reserve was
35.2 ⫾ 19 W/ml2. Eighteen patients in New York
enrollment.
Dobutamine stress echocardiograms were performed Heart Association class III HF who had ICDs were
beginning at an initial dose of 5 ␮g · kg⫺1 · min⫺1 for 3 enrolled. Their mean age was 61 years. Their mean
minutes, followed by 10 ␮g · kg⫺1 · min⫺1 for 3 min- baseline ejection fraction was 32 ⫾ 11%, and their
utes, with subsequent dose increments of 10 ␮g · kg⫺1 · mean SPP reserve was 3.5 ⫾ 3.2 W/ml2.
min⫺1 every 3 minutes, up to a maximum of 40 ␮g ·
SPP reserve discriminated HF patients from conkg⫺1 · min⫺1. The infusions were discontinued when the trols without overlap (mean 3.5, range 0.05 to 10.34 vs
patients achieved 85% of their target heart rate for their mean 35.2, range 21.8 to 51.3, respectively; p
age group or if chest pain, ST-segment depression, or ⬍0.0001). The study population had a mean follow-up
new regional wall motion abnormalities developed. of 15.5 months. There were no differences in the
Echocardiograms at rest and dobutamine stress echocar- baseline characteristics of the 2 groups of patients
diograms were used to determine the left ventricular with HF and ICDs: those with adequate (⬎1.5 W/ml2)
and poor (⬍1.5 W/ml2) SPP reserve (Table 1). There
ejection fraction and SPP reserve.
TABLE 1 Characteristics of Patients With HF and ICDs
BRIEF REPORTS
287
was 1 death, 1 heart transplantation, and 3 HF hospitalizations in the group with poor SPP reserve and no
end points in the group with adequate SPP reserve.
There were 4 ICD shocks in the group with poor SPP
reserve and 1 ICD shock in the group with adequate
SPP reserve. Kaplan-Meier analysis of the composite
end point of death, heart transplantation, or HF hospitalization showed a significantly (p ⫽ 0.02) greater
event rate in the group with poor SPP reserve (Figure
1). SPP reserve discriminated the 2 groups of patients
with HF and ICDs without overlap, whereas the ejection fraction showed considerable overlap. The mean
SPP reserve of the group with death or transplantation
was 0.63 ⫾ 0.4 W/ml2, and the mean SPP reserve of
the surviving patients was 4.9 ⫾ 2.3 W/ml2 (p ⫽
0.02).
predict survival in patients with ICDs and is not easily
measured in patients with more advanced HF and those
unable to reach aerobic threshold.
Therefore, an ideal prognostic tool would have the
predictive power of VO2 max, be noninvasive, and be
easy to use in patients with advanced HF. One such
instrument that has emerged as a novel echocardiographic measure of ventricular function is SPP reserve. It has an advantage over other echocardiographic measures in that it is not significantly affected
by afterload, can be adjusted for preload,8 and correlates well with VO2 max.2 The present study demonstrates that SPP reserve identifies patients with advanced HF and ICDs who are more likely to have
progressive HF.
•••
For ICDs to show a continued survival benefit in
patients with advanced HF, the risk for sudden cardiac
death from life-threatening ventricular arrhythmias
must be sufficiently greater than the risk for dying
from other causes.3 As ICD indications continue to
expand, it is increasingly important to discriminate
patients who are less likely to die of HF, thus identifying a subpopulation more likely to enjoy a survival
benefit from ICD therapy.
The prognostic values of the left ventricular systolic
ejection fraction, mitral deceleration time, and serum
sodium concentration for mortality have been well established.4 – 6 These measurements are insensitive and are
relatively poor predictors of survival in patient with
advanced HF. Peak oxygen consumption stress testing
(VO2 max) has greater prognostic ability than the left
ventricular ejection fraction and has been useful for the
risk stratification of ambulatory patients who may benefit
from cardiac transplantation.7 It has not been used to
288 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
1. Marmor A, Schneeweiss A. Prognostic value of noninvasively obtained left
ventricular contractile reserve in patients with severe heart failure. J Am Coll
Cardiol 1997;29:422– 428.
2. Armstrong GP, Carlier SG, Thomas JD. Estimation of cardiac reserve by peak
power: validation and initial application of a simplified index. Heart 1999;82:
357–364.
3. Fogoros RN. Impact of the implantable defibrillator on mortality: the axiom of
overall implantable cardioverter-defibrillator survival. Am J Cardiol 1996;
78(suppl):57A– 61A.
4. Rihal CS, Nishimura RA, Hatle LK. Systolic and diastolic function in patients
with clinical diagnosis of dilated cardiomyopathy: relation to symptoms and
prognosis. Circulation 1994;90:2772–2779.
5. Xie GY, Berk MR, Smith MD. Prognostic value of Doppler transmitral flow
patterns in patients with congestive heart failure. J Am Coll Cardiol 1994;24:
132–139.
6. Lee WH, Packer M. Prognostic importance of serum sodium concentration and
its modification by converting enzyme inhibition in patients with severe chronic
heart failure. Circulation 1986;73:257–267.
7. Mancini DM, Eisen H, Kussmaul W. Value of peak exercise oxygen consumption for optimal timing of cardiac transplantation in ambulatory patients with
heart failure. Circulation 1991;83:778 –786.
8. Sharir T, Feldman MD, Haber H. Ventricular systolic assessment in patients
with dilated cardiomyopathy by preload-adjusted maximal power. Validation and
noninvasive application. Circulation 1994;89:2045–2053.
JANUARY 15, 2005
Effect of Hormone Therapy on Mortality Rates Among
Women With Heart Failure and Coronary
Artery Disease
Kirsten Bibbins-Domingo, PhD, MD, Feng Lin, MS, Eric Vittinghoff, PhD,
Elizabeth Barrett-Connor, MD, Stephen B. Hulley, MD, MPH, Deborah Grady, MD,
and Michael G. Shlipak, MD, MPH
Randomized, controlled trial data from the Heart and
Estrogen-progestin Replacement Study were used to
evaluate the effect of estrogen plus progestin use on
all-cause mortality in women with heart failure and
coronary disease. Over the 4.1-year follow-up, estrogen plus progestin use had no effect on all-cause
mortality (hazard ratio 1.0, 95% confidence interval
0.7 to 1.4, p ⴝ 0.8) in women with heart failure and
coronary disease. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:289 –291)
bservational studies suggest that hormone therapy may reduce mortality risk in women with
O
heart failure. We used data from the Heart and Estrogen-progestin Replacement Study (HERS), a randomized, double-blind, placebo-controlled trial of the effects
of postmenopausal hormone therapy on cardiovascular
outcomes in women with established coronary artery
disease, to evaluate the effect of randomization to estrogen plus progesterone or placebo on all-cause mortality
in women with heart failure and coronary artery disease.
Of the 2,763 women with coronary artery disease enrolled in HERS, 644 women had heart failure. Over
the 4.1 years of follow-up in this double-blinded trial,
estrogen plus progestin use had no effect on all-cause
mortality (hazard ratio [HR] 1.0, 95% confidence interval [CI] 0.7 to 1.4, p ⫽ 0.8) in women with heart
failure and coronary artery disease.
•••
HERS methods have been previously described.1,2
Briefly, participants were postmenopausal women
with known coronary artery disease defined by previous myocardial infraction, coronary artery bypass surgery, percutaneous transluminal coronary angioplasty,
or angiographic evidence of ⱖ50% narrowing of ⱖ1
major coronary artery. Women were excluded if they
were ⬎79 years of age or had had hysterectomies, had
From the Division of General Internal Medicine, San Francisco General Hospital, University of California, San Francisco, San Francisco,
California; the Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, California; the Department of Family and Preventive Medicine, University of California,
San Diego, San Diego, California; and the Section of General Medicine, San Francisco VA Medical Center, University of California,
San Francisco, San Francisco, California. Dr. Bibbins-Domingo’s address is: Box 1364, San Francisco General Hospital, University of
California, San Francisco, San Francisco, California 94143-1364.
E-mail: bibbinsk@medicine.ucsf.edu. Manuscript received July 14,
2004; revised manuscript received and accepted September 3,
2004.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MPH,
coronary events ⱕ6 months before randomization,
had serum triglyceride levels ⬎3.39 mmol/L (300
mg/dl), had used hormones ⱕ3 months before the
study, or had a history of conditions that would contraindicate estrogen therapy.2 Women who were found
to be in New York Heart Association class IV at
baseline physical examination were also excluded
from HERS. Participants in HERS were randomly
assigned within the 20 HERS clinical centers to conjugated equine estrogens 0.625 mg plus medroxyprogesterone acetate 2.5 mg in 1 tablet daily (n ⫽ 1,380)
or a placebo of identical appearance (n ⫽ 1,383). The
HERS randomized, controlled trial was conducted
over 4.1 years. After the trial, women were unblinded
to treatment assignment and observed for an additional 2.7 years on average (HERS II).3
Women were defined as having heart failure on the
basis of a self-reported history of heart failure at entry
into HERS (n ⫽ 343) or diagnoses of heart failure during
HERS follow-up on the basis of signs or symptoms at
annual (n ⫽ 143) visits or hospital admissions for heart
failure (n ⫽ 158). Participants were included in this
analysis and contributed follow-up information only after they met 1 of these eligibility criteria. The left ventricular ejection fraction was determined by chart review
and was recorded in 442 women (69%) in this analysis.
We used an unadjusted, intention-to-treat Cox
model to assess the effect of assignment to hormone
therapy on mortality risk during the HERS main trial.
Using the methods given in Therneau and Grambsch,4
the 114 deaths during the HERS trial provided 80%
power to detect HRs of 1.7 in 2-sided tests with an ␣
level of 5%. We performed an unadjusted intentionto-treat Cox model using data from the HERS main
trial and the HERS II follow-up period. We also
undertook an “as-treated” analysis in which the
follow-up time for women receiving hormone therapy
was ascribed to placebo when they stopped using
hormone therapy, and the follow-up time for women
receiving placebo was ascribed to hormone therapy if
they used open-label hormone therapy. This “astreated” analysis was adjusted for other factors associated with mortality in women with heart failure and
coronary artery disease (New York Heart Association
classification; atrial fibrillation; diabetes; creatinine
clearance ⬍60 ml/min; ⬎1 myocardial infarction, age
⬎70 years, current tobacco use, limited exercise, left
ventricular ejection fraction, race, left bundle branch
block, and the use of aspirin, angiotensin-converting
enzyme inhibitors, ␤ blockers, statins, calcium chan0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.021
289
TABLE 1 Baseline Characteristics of 644 Women With Heart Failure and
Coronary Disease
Characteristic
Age (mean yrs)
Race
White
Black
Atrial fibrillation
Diabetes
Previous myocardial infarction
1
⬎1
Systolic blood pressure ⬎140 mm Hg
Creatinine clearance ⬍60 ml/min
Total cholesterol (mean mg/dl)
Low-density lipoprotein cholesterol
(mean mg/dl)
Body mass index (mean kg/m2)
Exercise less than peers
Smoking
Former smoker
Current smoker
Left ventricular ejection fraction
ⱕ50%
⬎50%
New York Heart Association classification
II
III or IV
Angiotensin-converting enzyme
inhibitor use
␤-blocker use
Aspirin use
Digoxin use
Estrogen Plus
Progestin
(n ⫽ 324)
67.7 ⫾ 6.6
271
41
12
112
(84)
(13)
(4)
(35)
Placebo
(n ⫽ 320)
67.9 ⫾ 6.6
269
37
5
93
(84)
(12)
(2)
(29)
159 (49)
20 (6)
136 (42)
177 (55)
229.5 ⫾ 41.7
147.3 ⫾ 37.6
160 (50)
28 (9)
129 (40)
194 (61)
227.1 ⫾ 42.6
143.2 ⫾ 39.9
29.7 ⫾ 6.0
104 (32)
29.4 ⫾ 5.7
142 (44)
155 (48)
39 (12)
166 (52)
42 (13)
p Value
0.7
0.8
0.09
0.1
0.5
0.7
0.1
0.5
0.2
0.5
0.001
0.4
95% CI 0.7 to 1.2, p ⫽ 0.4 for HERS
plus HERS II). Results were also unchanged when we restricted the analysis to the pure randomized design,
including only the 343 women with
heart failure at the time of randomization (HR 0.9, 95% CI 0.5 to 1.4, p
⫽ 0.6 for HERS; HR 0.9, 95% CI 0.6
to 1.3, p ⫽ 0.5 for HERS plus HERS
II).
Of the 442 women with recorded
left ventricular ejection fractions,
282 (60%) had left ventricular ejection fractions ⱖ50%. We tested for
interactions in each of the models
described previously and found no
evidence for interaction between
hormone treatment and the left ventricular ejection fraction as determinants of survival.
•••
Two recent observational studies that used data collected from
89 (28)
93 (29)
138 (43)
122 (38)
clinical trials of interventions for
0.8
heart failure have found that the
42 (13)
41 (13)
use of postmenopausal hormone
9 (3)
9 (3)
therapy was associated with de99 (31)
100 (31)
0.8
creased mortality in women with
103 (32)
100 (31)
0.9
heart failure.5,6 Postmenopausal
238 (74)
231 (72)
0.7
hormones have been postulated to
76 (24)
71 (22)
0.7
be of particular benefit to women
with heart failure through their effects on neurohormonal activanel blockers, digoxin, and diuretics). We tested for the tion,7–10 endothelial function,9,11 and cardiac repossibility of interactions between hormone therapy modeling.12 Animal studies have found that
and the left ventricular ejection fraction in each of the estrogen is associated with reductions in pre- and
models described previously using a significance level afterload.13–16
of p ⬍0.2 to identify interactions.
The investigators of previous observational studOf the 2,763 women in HERS, 644 had heart ies that have reported an association between postfailure. Of these, 324 had been randomly assigned at menopausal hormone therapy and reduced mortality
the beginning of HERS to receive estrogen plus pro- in women with heart failure have called for a rangestin and 320 had been assigned to placebo. The domized, controlled trial to confirm these results.
women assigned to hormone therapy did not differ Our study uses data from a randomized, controlled
significantly in their demographic characteristics, co- trial that compared the cardiovascular effects of
morbid conditions, tobacco use, and medication use estrogen plus progestin versus placebo, and we
from the women assigned to placebo (Table 1). found no survival advantage associated with horWomen in the placebo group were significantly more mone use in the subset of trial participants with
likely than those in the hormone group to report ex- heart failure. Our study has the major advantage of
ercising less often than their peers.
a randomized, double-blind, placebo-controlled
Over the 4.1-year HERS randomized trial, 114 of trial design, allowing the evaluation of the hormone
the 644 women with heart failure died (18%). Hor- effect independent of measured and unmeasured
mone use did not affect mortality rates during the trial confounding factors.
(Table 2). During the 2.7 years of the HERS II obserThe likely explanation for the difference between
vation period after the conclusion of the trial, an the previous observational studies of hormone therapy
additional 99 women with heart failure died (total in women with heart failure and the results of our
mortality 25%). Hormone use was not associated with randomized trial is the inability of observational studmortality risk during the HERS II observation period ies to exclude or adjust for residual confounding facor during the combined 6.8 years of the HERS trial tors that put hormone users at less risk than nonusers.
plus HERS II follow-up (Figure 1).
Another possibility for the different results of previous
The use of an as-treated analysis and adjustment studies is that our study population may have had less
for other risk factors did not change the findings (HR severe heart failure than those in the heart failure
0.9, 95% CI 0.6 to 1.4, p ⫽ 0.7 for HERS; HR 0.9, intervention trials, and many in our study had heart
290 THE AMERICAN JOURNAL OF CARDIOLOGY姞
0.5
VOL. 95
JANUARY 15, 2005
TABLE 2 HR Associated With Estrogen Use During the HERS
and HERS II Follow-up Periods
Period
HR (95% CI)
p Value
HERS trial (4.1 yrs)
HERS-II observation study
(2.7 yrs)
Combined HERS and HERS-II
study (6.8 yrs)
1.0 (0.7–1.4)
0.9 (0.6–1.3)
0.8
0.6
0.9 (0.7–1.2)
0.6
power to detect the small benefit of hormone use on
mortality in women with heart failure. Our study
does not address the impact of hormone use in
women with nonischemic heart failure. However,
the absence of an effect of hormone treatment in
HERS women with ischemic heart failure, coupled
with the results of large randomized trials showing
that hormone therapy increases the risk for coronary
artery disease events, stroke, and pulmonary embolism1,18 argues for caution in the pursuit of another
trial of hormone therapy for women with cardiovascular disease.
1. Hulley S, Grady D, Bush T, Furberg C, Herrington D, Riggs B, Vittinghoff E.
FIGURE 1. Kaplan-Meier survival curve by the use of estrogen
plus progestin over the HERS plus HERS II follow-up period.
failure with preserved systolic function. We cannot
directly address whether a potential mortality benefit
of hormone therapy is limited to those with severe
heart failure with depressed systolic function. Our
study population also had coexisting coronary artery
disease. In a trial of bucindolol for the treatment of
patients with heart failure, the use of postmenopausal
hormone therapy was associated with reduced mortality in women with nonischemic heart failure but not
those with ischemic disease.5 However, this interaction has been questioned17 and has not been consistently observed.6 A limitation of our study is that the
assessment of the left ventricular ejection fraction was
based on chart review at baseline, the measures were
not standardized, and a significant percentage was
missing.
In calling for a randomized, controlled trial of
hormone therapy in women with heart failure, the
investigators of these observational studies suggested that such trials offer clear advantages to
observational data when exploring the association
between therapies and outcomes. We believe that
our finding of no mortality benefit of hormone therapy in women with heart failure using data from a
randomized, controlled trial makes it less likely that
such a trial would find a positive effect, thus making
its design and conduct more difficult to justify.
Although we cannot exclude the possibility of a
small positive or negative effect of hormone use on
mortality, the CIs that we observed suggest that
only a small effect (⬍30% benefit or ⬍40% harm)
could have been missed by our study. Therefore, a
very large trial would be required to have sufficient
Randomized trial of estrogen plus progestin for secondary prevention of coronary
heart disease in postmenopausal women. Heart and Estrogen/progestin Replacement Study (HERS) Research Group. JAMA 1998;280:605– 613.
2. Grady D, Applegate W, Bush T, Furberg C, Riggs B, Hulley SB. Heart and
Estrogen/progestin Replacement Study (HERS): design, methods, and baseline
characteristics. Control Clin Trials 1998;19:314 –335.
3. Grady D, Herrington D, Bittner V, Blumenthal R, Davidson M, Hlatky M, Hsia
J, Hulley S, Herd A, Khan S, et al. Cardiovascular disease outcomes during 6.8
years of hormone therapy: Heart and Estrogen/progestin Replacement Study
follow-up (HERS II). JAMA 2002;288:49 –57.
4. Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox
Model. Berlin, Germany: Springer-Verlag; 2000:62.
5. Lindenfeld J, Ghali JK, Krause-Steinrauf HJ, Khan S, Adams K, Goldman S,
Peberdy MA, Yancy C, Thaneemit-Chen S, Larsen RL, et al. Hormone replacement therapy is associated with improved survival in women with advanced heart
failure. J Am Coll Cardiol 2003;42:1238 –1245.
6. Reis SE, Holubkov R, Young JB, White BG, Cohn JN, Feldman AM.
Estrogen is associated with improved survival in aging women with congestive heart failure: analysis of the vesnarinone studies. J Am Coll Cardiol
2000;36:529 –533.
7. Komesaroff PA, Sudhir K, Esler MD. Effects of estrogen on stress responses
in women. J Clin Endocrinol Metab 1999;84:4292– 4293.
8. Weitz G, Elam M, Born J, Fehm HL, Dodt C. Postmenopausal estrogen
administration suppresses muscle sympathetic nerve activity. J Clin Endocrinol
Metab 2001;86:344 –348.
9. Lieberman EH, Gerhard MD, Uehata A, Walsh BW, Selwyn AP, Ganz P,
Yeung AC, Creager MA. Estrogen improves endothelium-dependent, flowmediated vasodilation in postmenopausal women. Ann Intern Med 1994;121:
936 –941.
10. Nogawa N, Sumino H, Ichikawa S, Kumakura H, Takayama Y, Nakamura T,
Kanda T, Mizunuma H, Kurabayashi M. Effect of long-term hormone replacement therapy on angiotensin-converting enzyme activity and bradykinin in postmenopausal women with essential hypertension and normotensive postmenopausal women. Menopause 2001;8:210 –215.
11. Saitta A, Altavilla D, Cucinotta D, Morabito N, Frisina N, Corrado F, D’Anna
R, Lasco A, Squadrito G, Gaudio A, et al. Randomized, double-blind, placebocontrolled study on effects of raloxifene and hormone replacement therapy on
plasma no concentrations, endothelin-1 levels, and endothelium-dependent vasodilation in postmenopausal women. Arterioscler Thromb Vasc Biol 2001;
21:1512–1519.
12. Lim WK, Wren B, Jepson N, Roy S, Caplan G. Effect of hormone
replacement therapy on left ventricular hypertrophy. Am J Cardiol 1999;83:
1132–1134.
13. Sharkey LC, Holycross BJ, Park S, Shiry LJ, Hoepf TM, McCune SA, Radin
MJ. Effect of ovariectomy and estrogen replacement on cardiovascular disease in
heart failure-prone SHHF/Mcc-fa cp rats. J Mol Cell Cardiol 1999;31:1527–
1537.
14. Nekooeian AA, Pang CC. Estrogen restores role of basal nitric oxide in
control of vascular tone in rats with chronic heart failure. Am J Physiol 1998;
274:H2094 –H2099.
15. Nekooeian AA, Pang CC. Effects of estrogen on venous function in rats
with chronic heart failure. Am J Physiol Heart Circ Physiol 2000;278:H1941–
H1947.
16. Sudhir K, Chou TM, Mullen WL, Hausmann D, Collins P, Yock PG,
Chatterjee K. Mechanisms of estrogen-induced vasodilation: in vivo studies in
canine coronary conductance and resistance arteries. J Am Coll Cardiol 1995;
26:807– 814.
17. Petitti DB. New hope for hormone replacement and the heart? J Am Coll
Cardiol 2003;42:1246 –1248.
18. Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Kooperberg C,
Stefanick ML, Jackson RD, Beresford SA, Howard BV, Johnson KC, et al.
WGftWsHI. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized
controlled trial. JAMA 2002;288:321–333.
BRIEF REPORTS
291
Aortic Dimensions in Patients With Bicuspid Aortic
Valve Without Significant Valve Dysfunction
Moreno Cecconi, MD, Marcello Manfrin, MD, Alessandra Moraca, MD,
Raffaele Zanoli, PhD, Pier Luigi Colonna, MD, Maria Grazia Bettuzzi, MD,
Stefano Moretti, MD, Domenico Gabrielli, MD, and Gian Piero Perna, MD
The dimensions of the entire aorta at different anatomic
levels were measured by transthoracic 2-dimensional
echocardiography in 162 consecutive patients with isolated bicuspid aortic valves (BAVs) without significant
aortic valve dysfunction. Aortic dilation involved the
aortic root and the ascending aorta but was not present
in the descending and abdominal aorta. A significant
increase in the dimensions of the aortic arch was found
in patients with BAVs aged >40 years. Ascending aortic
diameter and the extension of aortic dilation were significantly correlated with age, but no correlation was
found between aortic dimensions and aortic valve
morphology. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:292–294)
icuspid aortic valves (BAVs) are associated with
dilation of the aortic root and ascending aorta.
B
However, limited data exist regarding the dimensions
1–12
of the aorta at other anatomic levels in patients with
BAVs. In addition, the influence of age and BAV
morphology on aortic dilation is not clearly determined.7,11 The purpose of this study was to analyze
the size of the entire aorta at different anatomic levels
in a large group of consecutive patients with isolated
BAVs without significant aortic valve dysfunction to
better characterize the associated aortic dilation and to
determine whether it is related to age and aortic valvular morphology.
•••
Three hundred six consecutive patients with unequivocal diagnosis of BAVs on 2-dimensional transthoracic
echocardiography were retrospectively identified. To
eliminate any factor influencing the aortic dimensions,
patients were excluded if they had aortic stenosis, more
than mild aortic regurgitation, or any other congenital
anomalies of the heart or great vessels, including aortic
coarctation, mitral and tricuspid valve disease, aortic
dissection, Marfan’s syndrome, a family history of
Marfan’s syndrome, a history of aortic surgery, systemic
arterial hypertension, and a history of systemic arterial
hypertension. Patients with left ventricular systolic dysfunction (left ventricular ejection fraction on 2-dimensional echocardiography ⱕ50%) were also excluded. Of
the 306 patients initially screened, 144 were excluded. A
total of 162 patients (113 males and 49 females) aged 1
to 68 years (mean age 24 ⫾ 20) were thus included in the
From Ospedale Cardiologico “G. M. Lancisi,” Ancona; and the
Polytechnic University of Marche, Ancona, Italy. Dr. Cecconi’s address is: Via Guazzatore 66, 60027 Osimo (AN), Italy. E-mail:
morcecconi@tiscalinet.it. Manuscript received March 30, 2004; revised manuscript received and accepted August 31, 2004.
292
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
study. One hundred sixty-two subjects without known
heart disease and with normal 2-dimensional echocardiographic and Doppler studies, matched for age,
gender, and body surface area, were included as a
control group.
Comprehensive transthoracic M-mode, 2-dimensional, and Doppler echocardiographic studies were
performed using commercially available equipment.
Aortic valve morphology and, particularly, the number and the opening motion of the cusps during systole
were carefully examined using multiple cardiac views.
A BAV was identified according to previously described criteria.13,14 On the basis of commissural orientation and cusp position in the parasternal short-axis
view, 2 morphologic types of BAVs were identified15,16: type 1, with the commissures oriented in a
medial–lateral direction, resulting in anterior and posterior cusps, and type 2, with the commissures oriented in an anteroposterior direction, resulting in left
and right cusps. The presence and position of raphe,
when visualized, was reported. Cusps were examined
with regard to relative size, and an eccentricity index
was obtained according to previously published criteria.16 A BAV was considered eccentric if the eccentricity index was ⱖ1.2.16 The presence and degree of
aortic valvular disease was assessed by Doppler echocardiography. Aortic stenosis was defined as present if
the peak aortic flow velocity, assessed by continuouswave Doppler, was ⱖ2.5 m/s.17 Aortic regurgitation
was graded according to a previously described algorithm.18 Patients with trivial aortic regurgitation were
classified as having mild aortic regurgitation. A comprehensive evaluation of the entire aorta was performed for patients with BAVs and controls using
multiple views, including left parasternal, left high
parasternal, right parasternal, right high parasternal,
apical, subcostal, suprasternal, and abdominal views.
The aortic dimensions were assessed at end-diastole at
the following levels: (1) the annulus, (2) the mid-point
of the sinuses of Valsalva, (3) the sinotubular junction,
(4) the ascending aorta at the level of its largest
diameter, (5) the transverse arch, (6) the descending
aorta posterior to the left atrium, and (7) the abdominal aorta just distal to the origin of the renal arteries.
Measurements were made using 2-dimensional imaging perpendicular to the long axis of the aorta in views
showing the largest aortic diameter. The maximal
luminal width was measured at each aortic level.
Three to 5 measurements were made and averaged at
each aortic level, and results were indexed for body
surface area. The presence and prevalence of aortic
dilation was established in comparison with the aortic
dimensions measured in the control group; dilation
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.098
0.7
1.1
1.3
1.2
0.8
0.9
0.9
⫾
⫾
⫾
⫾
⫾
⫾
⫾
11.7
17.3
15.1
16.8
12.8
11.5
9.9
1.1†
3‡
3‡
3.1‡
2.1*
1.1
1.4
⫾
⫾
⫾
⫾
⫾
⫾
⫾
12.8
20.7
18.2
22.1
13.8
11.7
9.6
0.7
0.9
1
1.1
0.8
1.2
1.1
⫾
⫾
⫾
⫾
⫾
⫾
⫾
11.9
16.7
14.8
16
12.9
11.4
8.9
0.9*
2.0†
2.4†
3.4‡
2.3
1.6
1.1
⫾
⫾
⫾
⫾
⫾
⫾
⫾
12.8
18.2
16.4
21.2
13.2
10.8
8.7
1.8
2.4
2.1
2.5
2
1.5
1.5
⫾
⫾
⫾
⫾
⫾
⫾
⫾
12.2
16
13.6
15.4
11.8
9.6
8.3
1.9
2.9*
3.1*
4.3*
1.4
1.3
1.5
⫾
⫾
⫾
⫾
⫾
⫾
⫾
13.5
18.6
16.3
18.7
12
10.1
8.2
2.2
2.7
1.8
1.6
1.7
1.6
1.1
⫾
⫾
⫾
⫾
⫾
⫾
⫾
15.3
19.5
16.4
18.1
13.3
11.1
9.6
2.8
3.8
3.5
5.1†
3.2
1.3
1
Controls
(n ⫽ 27)
BAV
(n ⫽ 27)
7–12
0–6
Values are expressed as mean ⫾ SD.
*p ⬍0.05; †p ⬍0.01; ‡p ⬍0.001.
Age Group (yrs)
was defined as a diameter greater than the mean ⫹ 2
SDs of the value found in controls for a given aortic
level. The extension of dilation was defined as the
number of dilated aortic segments. To analyze the
relation between age and aortic dimensions, the patients were divided in 5 groups: group 1, aged 0 to 6
years; group 2, aged 7 to 12 years; group 3, aged 13 to
18 years; group 4, aged 19 to 40 years; and group 5,
aged ⬎40 years. We used this division because of
the discontinuous distribution of ages in our study
population.
Data are presented as mean ⫾ 1 SD. Differences in
mean values between patients with BAVs and controls
were assessed using the independent-samples Student’s t test; differences in variances were accounted
for using Levene’s test. A 2-tailed p value of ⬍0.05
was considered statistically significant. Because of the
nonlinear relation between aortic size and body surface area, the relation between age and ascending
aortic dimensions was assessed only in patients with
BAVs aged ⬎12 years. The relations among ascending aortic dimensions, age, systolic and diastolic blood
pressures, and peak aortic flow velocity were assessed
by multiple regression analysis.
Baseline characteristics of patients with BAVs and
controls are listed in Table 1. Of the entire BAV group,
58 patients (36%) had normally functioning valves,
whereas 104 patients (64%) had mild aortic regurgitation. The morphologic analysis of BAVs showed that 89
patients (55%) had type 1 and 73 patients (45%) type 2
BAVs. Raphes were identified in 91 patients (56%); they
TABLE 3 Aortic Dimensions at Different Levels in Patients With BAVs and Controls According to Age
Values are expressed as mean ⫾ SD.
*p ⬍0.01; †p ⬍0.001.
⫾
⫾
⫾
⫾
⫾
⫾
⫾
4
5.1
4.1
4.7
3.8
2.9
2.5
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
⫾
14.4
19.5
16.7
18.6
14.4
12.3
10.5
Controls
(n ⫽ 24)
4.7
6.5*
6.3†
7.3†
4.7
3.2
3
BAV
(n ⫽ 24)
⫾
⫾
⫾
⫾
⫾
⫾
⫾
Controls
(n ⫽ 32)
15.4
22
19.5
23.7
15.1
12.4
10.2
BAV
(n ⫽ 32)
Annulus (mm/m2)
Sinuses of Valsalva (mm/m2)
Sinotubular junction (mm/m2)
Ascending aorta (mm/m2)
Aortic arch (mm/m2)
Descending aorta (mm/m2)
Abdominal aorta (mm/m2)
Controls
(n ⫽ 162)
13–18
Variable
Patients With
BAVs
(n ⫽ 162)
BAV
(n ⫽ 40)
TABLE 2 Aortic Dimensions at Different Levels in Patients
With BAVs and Controls
15.9
20.8
17.9
22.1
14.8
11.8
9.6
19 – 40
Values are expressed as mean ⫾ SD or number of patients.
*p ⬍0.001.
3.3
5.5
4.7
5.7
4.5
4
3.5
1.3 ⫾ 0.2
20.9
27.6
23.2
26
20.4
16.4
14
1.7 ⫾ 0.5*
5.3
8.2
8.3
9.4*
6.5
4.1
4
68 ⫾ 13
22.1
30.1
27
31.9
20.5
16.4
13.8
65 ⫾ 12
Annulus (mm/m2)
Sinuses of Valsalva (mm/m2)
Sinotubular junction (mm/m2)
Ascending aorta (mm/m2)
Aortic arch (mm/m2)
Descending aorta (mm/m2)
Abdominal aorta (mm/m2)
23.6 ⫾ 19.8
113/49
1.44 ⫾ 0.5
111 ⫾ 20
BAV
(n ⫽ 39)
23.6 ⫾ 19.8
113/49
1.43 ⫾ 0.5
108 ⫾ 18
Controls
(n ⫽ 40)
Controls
(n ⫽ 162)
Variable
Age (yrs)
Male/female
Body surface area (m2)
Systolic blood pressure
(mm Hg)
Diastolic blood pressure
(mm Hg)
Peak aortic velocity (m/s)
Patients With BAVs
(n ⫽ 162)
⬎40
Variable
Controls
(n ⫽ 39)
TABLE 1 Baseline Characteristics of Patients With BAVs and
Controls
BRIEF REPORTS
293
abdominal aorta. We also found dilation of the aortic
arch in patients aged ⬎40 years. This is the first study to
demonstrate the involvement of the aortic arch in patients with BAVs. Previous studies6,7,10 showed that the
greater aortic dimension in patients with BAVs was at
the level of the sinuses of Valsalva, with comparable size
at the sinuses and the proximal ascending aorta. We
found that the largest aortic segment was the ascending
aorta. Discrepancies with previous data are probably due
to the different site of sizing of the ascending aorta; we
measured the ascending aortic size at the maximum
diameter, which is usually at or near the mid-ascending
aorta instead of the proximal ascending aorta, as done in
previous studies. These results are in accordance with
those of 2 recent studies.9,11 Consequently, the maximum
aortic dimensions and the prevalence of aortic dilation
found in this study are greater than those previously
reported.2,6,7,10
Acknowledgment: We thank Francesco Antonini
Canterin, MD, for his critical review and suggestions.
FIGURE 1. Extension of aortic dilatation in patients with BAVs
according to age. Bars, means; error bars, 95% confidence intervals of means.
were located in the anterior cusp in type 1 BAVs and in
the right cusp in type 2 BAVs. Eccentric BAVs were
found in 86 patients (53%). The aortic dimensions at
different anatomic levels are listed in Table 2. In the
BAV group, no significant difference in ascending aortic
diameter was found between male and female patients
(23.3 ⫾ 7.7 vs 24.7 ⫾ 5.6 mm/m2) and between those
with and without mild aortic regurgitation (24.1 ⫾ 7.6 vs
23.5 ⫾ 7.1 mm/m2). No correlation was found between
aortic dimensions and systolic blood pressure, diastolic
blood pressure, and peak aortic flow velocity. Aortic
dimensions in the age groups are listed in Table 3. In
patients with BAVs who were aged ⬎12 years, age was
significantly correlated with ascending aortic diameter
(r ⫽ 0.73, p ⬍0.0001). In the entire BAV group, age was
significantly correlated with the extension of aortic dilation (r ⫽ 0.42, p ⬍0.001). Figure 1 shows the extension
of aortic dilation in patients with BAVs according to age.
The prevalence of dilation of the ascending aorta, assessed in relation to normal values found in the control
group, was 28% in group 1, 44% in group 2, 42% in
group 3, 65% in group 4, and 77% in group 5. Only age
showed a significant independent correlation with ascending aortic dimensions in multiple regression analysis
(p ⬍0.01). No relation was found between the morphologic characteristics of BAVs and aortic dimensions. In
particular, the ascending aortic diameter was not statistically different in type 1 and type 2 valves (23.5 ⫾ 6.1
vs 24.4 ⫾ 7.6 mm/m2), in valves with and without
raphes (23.9 ⫾ 7.2 vs 23.5 ⫾ 6.6 mm/m2), and in
eccentric and symmetric valves (24.2 ⫾ 7.9 vs 23.4 ⫾
5.9 mm/m2).
•••
The results of this study confirm that aortic dilation in
patients with BAVs involves the aortic root and the
ascending aorta, but is not present in the descending and
294 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
1. Pachulski RT, Weinberg AL, Chan KL. Aortic aneurysm in patients with
functionally normal or minimally stenotic bicuspid aortic valve. Am J Cardiol
1991;67:781–782.
2. Hahn RT, Roman MJ, Mogtader AH, Devereux RB. Association of aortic
dilation with regurgitant, stenotic and functionally normal bicuspid aortic valves.
J Am Coll Cardiol 1992;19:283–288.
3. Burks JM, Illes RW, Keating EC, Lubbe WJ. Ascending aortic aneurysm and
dissection in young adults with bicuspid aortic valve: implications for echocardiographic surveillance. Clin Cardiol 1998;21:439 – 443.
4. Ando M, Okita Y, Morota T, Takamoto S. Thoracic aortic aneurysm associated
with congenital bicuspid aortic valve. Cardiovasc Surg 1998;6:629 – 634.
5. Sabet HY, Edwards WD, Tazelaar HD, Daly RC. Congenitally bicuspid aortic
valves: a surgical pathology study of 542 cases (1991 through 1996) and a
literature review of 2,715 additional cases. Mayo Clin Proc 1999;74:14 –26.
6. Nistri S, Sorbo MD, Marin M, Palisi M, Scognamiglio R, Thiene G. Aortic root
dilatation in young men with normally functioning bicuspid aortic valves. Heart
1999;82:19 –22.
7. Keane MG, Wiegers SE, Plappert T, Pochettino A, Bavaria JE, St. John Sutton
MG. Bicuspid aortic valves are associated with aortic dilatation out of proportion
to coexistent valvular lesions. Circulation 2000;102(suppl):III-35–III-39.
8. Fedak PWM, Verma S, David TE, Leask RL, Weisel RD, Butany J. Clinical
and pathophysiological implications of a bicuspid aortic valve. Circulation 2002;
106:900 –904.
9. Alegret JM, Duran I, Palazon O, Vernis JM, Ameijide A, Rabassa A, Masana
L. Prevalence of and predictors of bicuspid aortic valves in patients with dilated
aortic roots. Am J Cardiol 2003;91:619 – 622.
10. Nkomo VT, Enriquez-Sarano M, Ammash NM, Melton LJ III, Bailey KR,
Desjardins V, Horn RA, Tajik AJ. Bicuspid aortic valve associated with aortic
dilatation. A community-based study. Arterioscler Thromb Vasc Biol 2003;
23:351–356.
11. Novaro GM, Tiong IY, Pearce GL, Grimm RA, Smedira N, Griffin BP.
Features and predictors of ascending aortic dilatation in association with a
congenital bicuspid aortic valve. Am J Cardiol 2003;92:99 –101.
12. Ferencik M, Pape LA. Changes in size of ascending aorta and aortic valve
function with time in patients with congenitally bicuspid aortic valves. Am J
Cardiol 2003;92:43– 46.
13. Brandenburg RO Jr, Tajik AJ, Edwards WD, Reeder GS, Shub C, Seward JB.
Accuracy of 2-dimensional echocardiographic diagnosis of congenitally bicuspid
aortic valve: echocardiographic-anatomic correlation in 115 patients. Am J Cardiol 1983;51:1469 –1473.
14. Chan KL, Stinson WA, Veinot JP. Reliability of transthoracic echocardiography in the assessment of aortic valve morphology: pathological correlation in
178 patients. Can J Cardiol 1999;15:48 –52.
15. Roberts WC. The congenitally bicuspid aortic valve: a study of 85 autopsy
cases. Am J Cardiol 1970;26:72– 83.
16. Beppu S, Suzuki S, Matsuda H, Ohmori F, Nagata S, Miyatake K. Rapidity
of progression of aortic stenosis in patients with congenital bicuspid aortic valves.
Am J Cardiol 1993;71:322–327.
17. Otto CM, Lind BK, Kitzman DW, Gersh BJ, Siscovick DS. Association of
aortic valve sclerosis with cardiovascular mortality and morbidity in the elderly.
N Engl J Med 1999;341:142–147.
18. Padial LR, Oliver A, Vivaldi M, Sagie A, Freitas N, Weyman AE, Levine
RA. Doppler echocardiographic assessment of progression of aortic regurgitation.
Am J Cardiol 1997;80:306 –314.
JANUARY 15, 2005
Is Transesophageal Echocardiography Overused in the
Diagnosis of Infective Endocarditis?
Molly Thangaroopan,
MD,
Because of its greater sensitivity, transesophageal
echocardiography (TEE) is often misused as a screening tool for the exclusion of infective endocarditis (IE)
in patients with small clinical probability of the disease. This study examined the role of using TEE exclusively at a Canadian tertiary care center for the
diagnosis of IE and determined which clinical variables are most often associated with positive or negative echocardiographic results supporting or refuting
the diagnosis. 䊚2005 by Excerpta Medica Inc.
(Am J Cardiol 2005;95:295–297)
uke’s criteria, which form the basis of current
guidelines, incorporate echocardiographic docuD
mentation of vegetations in establishing a diagnosis of
infective endocarditis (IE).1–3 Because of its greater
sensitivity, transesophageal echocardiography (TEE)
is often misused as a screening tool for the exclusion
of IE in patients with small clinical probability of the
disease.4 – 6 This study examined the role of using TEE
exclusively at a Canadian tertiary care center for the
diagnosis of IE and determined which clinical variables are most often associated with positive or negative echocardiographic results supporting or refuting
the diagnosis.
•••
From August 31, 2001, to August 30, 2002, 104
consecutive inpatients referred to the University of
Alberta Hospital Adult Echocardiography Laboratory
for the primary indication of “rule out IE” were reviewed. The study protocol was approved by the Ethics Committee at the University of Alberta and
Stollery Children’s Center. Patients’ charts were reviewed to extract clinical findings associated with IE,
including intravenous drug use, the presence of an
indwelling catheter, high-risk invasive procedures, a
history of IE, immunocompromised status, and highrisk predisposition to heart disease. Clinical findings
such as fever (⬎38.0°C), documentation of a new
regurgitant murmur, and peripheral manifestations
(Roth spots, Janeway lesions, Osler’s nodes, splinter
hemorrhages) and splenomegaly were also included.
Laboratory data were collected, including the number
of blood cultures, the number of positive blood culture
results, micro-organisms identified, and alternative
potential microbiologic evidence of infection. A disFrom the Division of Cardiology, University Health Network, Toronto,
Ontario; and the Division of Cardiology, University of Alberta Hospital, Edmonton, Alberta, Canada. Dr. Choy’s address is: Division of
Cardiology, 2C2, WM Health Sciences Center, 8440-112 Street,
Edmonton, AB T6G 2B7, Canada. E-mail: jchoy@cha.ab.ca. Manuscript received May 26, 2004; revised manuscript received and
accepted August 31, 2004.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
and Jonathan B. Choy,
MD
charge diagnosis of endocarditis and alternative diagnoses were obtained from medical records.
Each case was categorized by the von Reyn criteria
into high, medium, or low probability of endocarditis
using clinical criteria based on previously published
definitions.7 TEE was performed using a Philips Sonos 5500 ultrasound machine (Philips Medical Systems, Andover, Massachusetts) with a 4.0- to 7.0MHz multiplane transducer. TEE was performed by
American Society of Echocardiography level III–
trained cardiologists.
Statistical analysis was performed using SAS version 8.2 (SAS Institute Inc., Cary North Carolina).
Fisher’s exact test was used to compare the frequencies of prespecified clinical criteria in the 2 groups,
positive versus negative echocardiographic results,
with a p value of 0.05 defining significance. Forward
logistic regression was used to determine independent
predictors of having positive echocardiographic results for vegetations.
From August 31, 2001, to August 30, 2002, 114
inpatients underwent TEE in our institution for suspected IE. Ten patients already known to have vegetations were having repeat TEE for reassessment and
were excluded. Table 1 lists the demographic characteristics of the patients. Seventy-seven patients (74%)
had fevers and 70 (67%) had positive blood culture
results, of whom only 20 (19%) had persistently positive blood culture results.
According to the von Reyn criteria, 87 patients were
low, 5 were intermediate, and 12 were high probability
for IE. In the high-probability group, 9 of 12 echocardiographic results (75%) were positive for vegetations
(Figure 1). Three of 5 patients (60%) in the intermediateprobability group had positive electrocardiographic results. In contrast, only 2 of 87 patients (2.2%) in the
low-probability group had vegetations verified on TEE,
both of whom had 3 characteristics in common: positive blood culture results for Staphylococcus aureus
(S. aureus) infection, immunosuppression, and fever
(⬎38.0°C). The remainder of the patients in the lowprobability group had negative echocardiographic results
and firm alternative diagnoses for the source of suspected
infection. The most common sources of infections in the
low-risk group were soft tissue or bone (46%), abscess
(15%), pneumonia (14%), and line sepsis (10%). The
remaining discharge diagnoses were connective tissue
disorder (7%), malignancy (5%), and viral illness (2%).
Of the 14 positive echocardiographic results, the
causative micro-organism from blood cultures was S.
aureus (n ⫽ 5), Staphylococcus epidermidis (n ⫽ 1),
Enterococcus fecalis (n ⫽ 4), Streptococcus viridans
(n ⫽ 2), Streptococcus bovis (n ⫽ 1), and Streptococcus
mitis (n ⫽ 1; Table 2). Gram negative rods were not
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.022
295
TABLE 1 Baseline Demographic Characteristics of Patients
Who Underwent TEE for Suspected IE
Parameter
TABLE 3 Clinical Criteria in Patients With and Without
Evidence of Endocarditis on TEE
Value
Mean age (yrs)
Men
Fever
Positive blood culture results
Persistently positive blood culture results
Immunocompromised
Intravenous drug user
Indwelling catheter
Predisposing heart disease
Previous endocarditis
54
70
77
70
20
54
8
46
24
7
(17–89)
(67%)
(74%)
(67%)
(19%)
(52%)
(8%)
(44%)
(23%)
(7%)
Clinical Criterion
Previous IE
Fever
Intravenous drug use
Indwelling catheter
Rheumatic heart disease
Immunosuppression
New regurgitant murmur*
Persistently positive blood
culture results*
Streptococcus aureus
cultured
Streptococcus viridans
cultured
Peripheral manifestations
Positive blood culture results
Positive
(n ⫽ 14)
2
14
3
6
1
7
8
10
(14)
(100)
(21)
(43)
(7)
(50)
(57)
(71)
Negative
(n ⫽ 90)
5
63
5
40
6
47
9
10
(6)
(70)
(6)
(44)
(7)
(52)
(10)
(11)
Univariate
p Value
NS
0.02
0.07
NS
NS
NS
⬍0.001
⬍0.001
5 (35)
25 (28)
NS
2 (14)
2 (2)
0.09
9 (64)
14 (100)
11 (12)
56 (62)
⬍0.001
0.004
Values are numbers (percentages).
*Independent predictors by multivariate analysis were new regurgitant murmur (p ⫽ 0.001) and persistently positive blood culture results (p ⬍0.001).
FIGURE 1. Comparison of positive and negative TEE results
according to risk group.
TABLE 2 Type of Organism Cultured With Positive Blood
Culture Results
Organism Cultured
S. aureus
Staphylococcus epidermis
Enterococcus fecalis
Streptococcus bovis
Streptococcus viridans
Gram negative bacilli
Streptococcus pneumonia
Fungal
Other
Positive TEE
Results
Negative TEE
Results
5
1
4
0
2
0
0
0
2
24
6
6
1
2
10
2
2
5
found to be a causative agent for endocarditis. There
were 2 cases of prosthetic valve endocarditis involving
the aortic valve, 1 case of pacemaker lead tip infection, 6
cases of indwelling catheter–associated endocarditis, and
3 cases in patients with a history of intravenous drug use.
Of the high-risk patients, 4 (33%) had peripheral
embolization (i.e., septic pulmonary emboli, splenic
abscess, and infarct); 3 (25%) had central nervous
system complications, including stroke, brain abscess, and mycotic aneurysm with rupture; and 1
(8.3%) had prosthetic valve dehiscence and aortic
ring abscess with subsequent right coronary artery
embolism resulting in acute myocardial infarction
and cardiogenic shock.
The univariate predictors of positive TEE results
were fever, new regurgitant murmur, peripheral man296 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
ifestations, positive blood culture results, persistently
positive blood culture results, intravenous drug use,
and Streptococcus viridans bacteremia (Table 3). In a
forward stepwise logistic regression model, only persistently positive blood culture results (p ⬍0.001) and
new regurgitant murmur (p ⫽ 0.001) remained significant independent predictors of positive TEE results.
The presence of an intravascular catheter, rheumatic
heart disease, and S. aureus infection were not considered independent predictors of positive TEE
results.
•••
We found that the most TEE studies were performed in patients with small risk for IE, and the
results were therefore negative. Many of these TEE
studies could probably have been avoided with more
judicious screening and the appropriate application of
clinical criteria to select patients with a moderate to
large risk for IE.
Because of the large incidence of complications in
our high-risk group (67%), our results support the use
of TEE in high-risk patients to acquire prognostic
information and for assessment of the risk for complications related to IE, such as heart failure, risk for
embolization, valve dehiscence, and abscess formation. TEE in high-risk patients offers additional information, such as characteristics of the vegetation (e.g.,
size and mobility), the presence of a perivalvular
abscess, prosthetic valve dehiscence, or multivalvular
involvement. These have been shown in previous
studies to help with the prediction of prognosis and
tailor the course of management accordingly.8 TEE in
the intermediate-risk group, in contrast, likely has the
greatest diagnostic utility, confirming the results of
previous studies.9,10
In the low-risk group, our study suggests that most
TEE studies are unnecessary unless S. aureus bacteremia, immunosuppression, and fever are present. S.
aureus bacteremia is a condition associated with a
JANUARY 15, 2005
high mortality rate if inadequately treated.11 In the
current era, the most common mode of acquisition is
hospital-acquired S. aureus bacteremia, not community acquired, as specified in Duke’s criteria.12 With
the increasing use of intravascular catheters and prosthetic materials and attendant increases in the risk for
staphylococcal infections related to foreign material,
the revision of Duke’s original criteria is required to
better select patients for TEE with suspected IE who
currently are being misclassified as having a small
probability of having this syndrome.13
In conclusion, it is well known that diagnostic
tests, such as myocardial perfusion imaging, performed in patients with very small or large clinical
likelihood of coronary disease offer little additional
information that alters patient management.14 Similarly, in patients with either large or small clinical
likelihood of IE, the practical role of TEE for diagnostic purposes is also small. Although Duke’s criteria
essentially suggest that echocardiography be used as a
screening test for IE, our study does not support this
strategy.
Acknowledgment: We would like to thank EPICORE
at the University of Alberta for aid in statistical analysis.
1. Durack DT, Lukes BS, Bright DK, for the Duke Endocarditis Service. New
criteria for the diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Am J Med 1994;96:200 –209.
2. Olaison L, Hogevik H. Comparison of the von Reyn and Duke criteria for the
diagnosis of infective endocarditis: a critical analysis of 161 episodes. Scand
J Infect Dis 1996;28:399 – 406.
3. Bayer AS, Bolger AF, Taubert KA, Wilson W, Steckelberg J, Karchmer AW,
Levison M, Chambers HF, Dajani AS, Gewitz MH, et al. Diagnosis and management of infective endocarditis and its complications. Circulation 1998;98:
2936 –2948.
4. Erbel R, Rohmann S, Drexler M. Improved diagnostic value of echocardiography in patients with infective endocarditis by transesophageal approach: a
prospective study. Eur Heart J 1988;9:43–53.
5. Birminghan GD, Rahko PS, Ballantyne F. Improved detection of infective
endocarditis with transesophageal echocardiography. Am Heart J 1992;123:
774 –781.
6. Shapiro SM, Young E, De Guzman S. Transesophageal echocardiography in
the diagnosis of infective endocarditis. Chest 1994;105:377–382.
7. von Reyn CF, Levy BS, Arbeit RD, Friedland G, Crumpacker CS. Infective
endocarditis: an analysis based on strict case definitions. Ann Intern Med 1981;
94:505–518.
8. Sanfilippo AJ, Picard MH, Newell JB, Rosas E, Davidoff R, Thomas JD,
Weyman AE. Echocardiographic assessment of patients with infectious endocarditis: prediction of risk of complications. J Am Coll Cardiol 1991;18:1191–1199.
9. Kuruppu JC, Corretti M, Mackowiak P, Roghmann MC. Overuse of transthoracic echocardiography in the diagnosis of native valve endocarditis. Arch Int
Med 2002 1992;162:1715–1720.
10. Lindner JR, Case RA, Dent JM, Abbott RD, Scheld WM, Kaul S. Diagnostic
value of echocardiography in suspected endocarditis: an evaluation based on the
pretest probability of disease. Circulation 1996;93:730 –736.
11. Rosen AB, Fowler VG, Corey GR, Downs SM, Biddle AK, Li J, Jollis JG.
Cost effectiveness of transesophageal echocardiography to determine the duration
of therapy for intravascular catheter associated staphylococcus aureus bacteremia.
Ann Intern Med 1999;130:810 – 820.
12. Fowler VG, Li J, Corey R, Boley J, Marr KA, Gopal AK, Kong LK, Gottlieb
G, Donovan CL, Sexton DJ, et al. Role of echocardiography in evaluation of
patients with Staphylococcus aureus bacteremia: experience in 103 patients. J Am
Coll Cardiol 1997;30:1072–1078.
13. Lamas CC, Eykyn SJ. Suggested modifications to the Duke criteria for the
clinical diagnosis of native valve and prosthetic valve endocarditis: analysis of
118 pathologically proven cases. Clin Infect Dis 1997;25:713–719.
14. Diamond GA, Denton TA, Berman DS, Cohen A. Prior restraint: a Bayesian
perspective on the optimization of technology utilization for diagnosis of coronary artery disease. Am J Cardiol 1995;76:82– 86.
Symptomatic Patients Have Similar Outcomes
Compared With Asymptomatic Patients After Carotid
Artery Stenting With Emboli Protection
Michael H. Yen, MD, David S. Lee, MD, Samir Kapadia, MD, Ravish Sachar,
Deepak L. Bhatt, MD, Christopher T. Bajzer, MD, and Jay S. Yadav, MD
In a single-center cohort of 174 consecutive patients,
we sought to evaluate whether the use of emboli
protection devices (EPDs) results in equivalent rates of
adverse events in symptomatic and asymptomatic
patients after carotid artery stenting (CAS) with EPDs.
Death or stroke occurred in 3.3% in the symptomatic
group and in 3.5% of the asymptomatic group at 30
days (p ⴝ NS). At 6 months, there was also no
significant difference in the rate of stroke or death
between the groups. Unlike surgical revascularization, symptomatic patients did not have a greater risk
From the Cleveland Clinic Foundation, Cleveland, Ohio. Dr. Yadav’s
address is: Department of Cardiovascular Medicine / F25, Cleveland
Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195.
E-mail: yadavj@ccf.org. Manuscript received May 25, 2004; revised
manuscript received and accepted August 31, 2004.
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MD,
for stroke and death compared with asymptomatic
patients after CAS with EPDs. 䊚2005 by Excerpta
Medica Inc.
(Am J Cardiol 2005;95:297–300)
everal randomized clinical trials have shown that
surgical revascularization is superior to medical
S
management for symptomatic and asymptomatic patients with severe carotid stenoses.1,2 However, most
patients in these trials had few medical co-morbidities
or anatomic features that are known to increase the
risk for major adverse clinical events after carotid
endarterectomy (CEA).3 In the major surgical studies,
symptomatic patients had substantially higher stroke
and death rates than asymptomatic patients, which is
reflected in the published CEA practice guidelines.4,5
The introduction of carotid artery stenting (CAS) with
emboli protection devices (EPDs) offers an alternative, less invasive method of intervention for patients
considered to be at high risk for CEA.6,7 We sought to
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.09.023
297
TABLE 1 Adverse Outcomes at 30 Days and 6 Months
Outcomes
30-d
Stroke
Intracranial hemorrhage
Death
Stroke or death
6-mo
Stroke
Intracranial hemorrhage
Death
Stroke or death
Symptomatic
(n ⫽ 62)
Asymptomatic
(n ⫽ 112)
1
1
1
2
(1.6%)
(1.6%)
(1.6%)
(3.2%)
1
0
3
4
(0.9%)
(0%)
(2.7%)
(3.6%)
1
1
2
3
(1.6%)
(1.6%)
(3.2%)
(4.8%)
1
0
5
6
(0.9%)
(0%)
(4.5%)
(5.4%)
FIGURE 1. Kaplan-Meier 30-day and 6-month estimates of survival free of stroke or death in symptomatic and asymptomatic
patients who underwent CAS with EPDs.
determine whether the hazard associated with symptomatic disease in patients who undergo CEA persists
in those treated with CAS in the setting of EPDs.
•••
We prospectively followed 174 consecutive patients who underwent CAS of 193 arteries with EPDs
from January 2000 to September 2002. Symptomatic
patients were defined as patients who had stenoses
ⱖ60% as determined by the North American Symptomatic Endarterectomy Trial criteria1 and had either
ipsilateral transient ischemic attacks or strokes ⱕ6
months before their carotid stenting procedures. A
transient ischemic attack was characterized by distinct
focal neurologic dysfunction or monocular blindness
298 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
that resolved ⱕ24 hours after onset. Strokes were
defined as the persistence of neurologic symptoms or
signs for ⬎24 hours. All asymptomatic patients had
carotid stenoses ⱖ80%. A neurologist evaluated all
patients with complete neurologic examinations and a
standardized stroke scale assessment before CAS. All
patients who underwent CAS gave written informed
consent and were enrolled in protocols for CAS approved by the Cleveland Clinic Foundation Institutional Review Board.
Procedural details regarding the CAS procedure
have been previously described.8 The AngioGuard
emboli capture guidewire (Cordis Corporation, Miami
Lakes, Florida) was used in 137 patients, GuardWire
(Medtronic Vascular, Santa Rosa, California) in 22
patients, FilterWire EX (Boston Scientific Corporation, Natick, Massachusetts) in 11 patients, Accunet
(Guidant Corporation, Indianapolis, Indiana) in 3 patients, and Emboshield (Abbott Vascular Devices,
Redwood City, California) in 1 patient. Procedural
success was defined as a ⬍30% residual stenosis as
determined by angiography immediately after stent
placement.
The primary end point was death or any stroke at
30-day and 6-month follow-up. Postoperative stroke
events were defined as any neurologic deficit persisting for ⬎24 hours and were adjudicated by a neurologist who did not participate in the CAS procedure.
Mortality was determined by the Social Security
Death Index. Event-free survival between patient
groups was assessed using Kaplan-Meier analysis, and
differences between the curves were assessed by the
log-rank test. Categorical values are expressed as values and percentages, whereas continuous variables are
presented as mean ⫾ SD. Differences between categorical variables were analyzed with the chi-square
test or Fisher’s exact test. For continuous variables,
2-tailed t tests were used to assess differences. A p
value ⬍0.05 was considered statistically significant.
Statistical analyses were performed using SPSS statistical software version 9.0 (SPSS, Inc., Chicago,
Illinois).
In our cohort, 36% of the patients were symptomatic (n ⫽ 62) and 64% were asymptomatic (n ⫽ 112).
Baseline characteristics were similar between the 2
study groups except for a higher incidence of previous
transient ischemic attacks (81% vs 13%, p ⫽ 0.001)
and a lower incidence of previous CEA (16% vs 34%,
p ⫽ 0.03) in the symptomatic group. Procedural success was achieved in 100% of patients in the 2 groups.
There were no significant differences between the
groups in either the 30-day or 6-month outcomes
(Table 1 and Figure 1). There was only 1 stroke in
each group at 6-month follow-up, each of which was
attributed to periprocedural embolic events (Table 2).
There was 1 myocardial infarction in each group at
30-day follow-up and, at 6-month follow-up, there
was 1 myocardial infarction in the symptomatic group
and 4 in the asymptomatic group (p ⫽ NS).
•••
Although symptomatic and asymptomatic patients
benefit from surgical revascularization for significant
JANUARY 15, 2005
results compare favorably with those
of recent reports in a global registry7
Time After
and other studies,18 in which the 30Carotid
Index
day stroke and death rates were relaAge (yrs)
Artery
Procedure
Event
tively low in symptomatic and asympSymptomatic
tomatic patients who receive cerebral
58
Right internal
Periprocedural Right hemispheric embolic stroke
protection at time of CAS.
76
Right internal
3.5 d
Intracranial hemorrhage/death
There are certain limitations with
83
Right internal
38 d
Death (cardiac)
this single-center study. Patients who
86
Left internal
143 d
Death (unknown cause)
Asymptomatic
underwent CAS with EPDs did not
78
Left internal
1h
Left hemispheric embolic stroke
have concomitant transcranial Dopp80
Left internal
2.6 d
Death (presumed cardiac)
ler to compare the degree and fre69
Left internal
25 d
Death (cardiac)
quency of microembolization be69
Right internal
22 d
Death (after cardiac bypass)
tween the 2 groups at the 3 phases of
77
Left internal
36 d
Death (unknown, after cardiac bypass)
62
Left internal
45 d
Death (acute renal failure)
the procedure associated with increased embolic events: predilatation, stent deployment, and postdilacarotid stenosis, the risk for stroke or death from CEA tation.19 Because it was not required of patients to
is increased in symptomatic patients. In a systemic have a more sensitive evaluation of embolic events
overview of 25 trials of CEA in symptomatic and after the procedure, such as diffusion-weighted magasymptomatic patients, Rothwell et al9 reported that netic resonance imaging, potential silent neurologic
the overall risk for stroke and/or death within 30 days embolic events were not detected.20 Although the Anin symptomatic patients was 5.18% (95% confidence gioGuard device was the most commonly used at our
interval 4.30 to 6.06) compared with 3.35% (95% center, this study was not designed to evaluate differconfidence interval 2.38 to 4.31) in asymptomatic ences in efficacy among various cerebral EPDs.
patients. This difference in short-term outcome may Whether similar results can be attained in a lower risk
be possibly explained by the increased risk for embo- patient cohort with severe carotid stenoses remains to
lization during CEA in symptomatic patients.
be determined. Last, given the low incidence of major
Atherosclerotic plaque in patients with significant adverse clinical events in our cohort, our study was
carotid artery stenosis frequently has vulnerable char- underpowered to demonstrate statistical equivalence
acteristics, such as ulceration, thrombus, platelet ag- between symptomatic and asymptomatic patients.
gregates, or fibrin clots that may predispose the plaque However, the low event rate in either group provides
to embolize during either surgical or percutaneous reassuring data regarding the safety of CAS with
revascularization, resulting in periprocedural cerebral EPDs.
ischemia.9,10 Previous studies have demonstrated that
In summary, the use of emboli protection to premicroembolism as assessed by transcranial Doppler vent periprocedural and postprocedural neurologic
occurs more frequently in symptomatic than in asympevents results in equivalent major adverse events in
tomatic patients during CEA.11 The higher rate of
symptomatic patients with significant carotid stenoses
stroke after CEA in the surgical trials of symptomatic
patients may be related to a greater burden of unstable who are considered to be at high risk for CEA comatherosclerotic plaques in symptomatic patients, lead- pared with asymptomatic patients with similar coing to increased risk for embolization during revascu- morbidities. By preventing embolization into the celarization.12,13 Before the development of EPDs for rebral circulation, the use of EPDs during CAS
CAS, percutaneous revascularization was associated appears to neutralize the increased risk for stroke in
with a greater risk for microembolization as detected symptomatic patients that was reported in the CEA
by transcranial Doppler than CEA.14 Recent studies trials. Whether CAS with EPDs can be safely perthe completion of
have suggested that this greater embolic risk during formed in lower risk patients awaits
15
large,
randomized
clinical
trials.
CAS can be attenuated with cerebral embolic protection,15 with a corresponding decrease in major clinical
events.7,16 The decrease in cerebrovascular events
with the use of EPDs in CAS has recently been con- 1. North American Symptomatic Carotid Endarterectomy Trialists’ Collaborative
Group. Beneficial effect of carotid endarterectomy in symptomatic patients with
firmed by the Endarterectomy Versus Angioplasty in high-grade carotid stenosis. N Engl J Med 1991;325:445– 453.
Patients with Symptomatic Severe Carotid Stenosis 2. Executive Committee for the Asymptomatic Carotid Atherosclerosis Study
trial, in which the Safety Committee recommended Group. Endarterectomy for asymptomatic carotid artery stenosis. JAMA 1995;
273:1421–1428.
the cessation of unprotected CAS because of the three- 3. Tu JV, Wang H, Bowyer B, Green L, Fang J, Kucey D. Risk factors for death
fold greater 30-day rate of stroke compared with CAS or stroke after carotid endarterectomy: observations from the Ontario Carotid
Endarterectomy Registry. Stroke 2003;34:2568 –2573.
with EPDs.17
4. Moore WS, Mohr JP, Najafi H, Robertson JT, Stoney RJ, Toole JF. Carotid
In the present study, we have demonstrated that CAS endarterectomy: practice guidelines. Report of the Ad Hoc Committee to the Joint
with EPDs results in similar rates of stroke or death in Council of the Society for Vascular Surgery and the North American Chapter of the
symptomatic compared with asymptomatic patients. International Society for Cardiovascular Surgery. J Vasc Surg 1992;15:469 – 479.
5. Biller J, Feinberg WM, Castaldo JE, Whittemore AD, Harbaugh RE,
Only 1 periprocedural embolic stroke occurred in either Dempsey RJ, Caplan LR, Kresowik TF, Matchar DB, Toole J, et al. Guidegroup in the setting of cerebral distal protection. Our lines for carotid endarterectomy: a statement for healthcare professionals from
TABLE 2 Adverse Clinical Events
BRIEF REPORTS
299
a special writing group of the Stroke Council, American Heart Association.
Stroke 1998;29:554 –562.
6. Roubin GS, New G, Iyer SS, Vitek JJ, Al-Mubarak N, Liu MW, Yadav J,
Gomez C, Kuntz RE. Immediate and late clinical outcomes of carotid artery
stenting in patients with symptomatic and asymptomatic carotid artery stenosis:
a 5-year prospective analysis. Circulation 2001;103:532–537.
7. Wholey MH, Al-Mubarek N. Updated review of the global carotid artery stent
registry. Catheter Cardiovasc Interv 2003;60:259 –266.
8. Yadav JS, Roubin GS, Iyer S, Vitek J, King P, Jordan WD, Fisher WS. Elective
stenting of the extracranial carotid arteries. Circulation 1997;95:376 –381.
9. Rothwell PM, Slattery J, Warlow CP. A systematic comparison of the risks of
stroke and death due to carotid endarterectomy for symptomatic and asymptomatic
stenosis. Stroke 1996;27:266 –269.
10. Valton L, Larrue V, le Traon AP, Massabuau P, Geraud G. Microembolic
signals and risk of early recurrence in patients with stroke or transient ischemic
attack. Stroke 1998;29:2125–2128.
11. Stork JL, Kimura K, Levi CR, Chambers BR, Abbott AL, Donnan GA.
Source of microembolic signals in patients with high-grade carotid stenosis.
Stroke 2002;33:2014 –2018.
12. Golledge J, Greenhalgh RM, Davies AH. The symptomatic carotid plaque.
Stroke 2000;31:774 –781.
13. Liapis CD, Kakisis JD, Kostakis AG. Carotid stenosis: factors affecting
symptomatology. Stroke 2001;32:2782–2786.
14. Jordan WD Jr, Voellinger DC, Doblar DD, Plyushcheva NP, Fisher WS,
McDowell HA. Microemboli detected by transcranial Doppler monitoring in
patients during carotid angioplasty versus carotid endarterectomy. Cardiovasc
Surg 1999;7:33–38.
15. Hobson RW II. CREST (Carotid Revascularization Endarterectomy Versus
Stent Trial): background, design, and current status. Semin Vasc Surg 2000;13:
139 –143.
16. Kastrup A, Groschel K, Krapf H, Brehm BR, Dichgans J, Schulz JB. Early
outcome of carotid angioplasty and stenting with and without cerebral protection
devices: a systematic review of the literature. Stroke 2003;34:813– 819.
17. Mas JL, Chatellier G, Beyssen B. Carotid angioplasty and stenting with and
without cerebral protection: clinical alert from the Endarterectomy Versus Angioplasty in Patients With Symptomatic Severe Carotid Stenosis (EVA-3S) trial.
Stroke 2004;35:e18 – e20.
18. Cremonesi A, Manetti R, Setacci F, Setacci C, Castriota F. Protected carotid
stenting: clinical advantages and complications of embolic protection devices in
442 consecutive patients. Stroke 2003;34:1936 –1941.
19. Al-Mubarak N, Roubin GS, Vitek JJ, Iyer SS, New G, Leon MB. Effect of
the distal-balloon protection system on microembolization during carotid stenting. Circulation 2001;104:1999 –2002.
20. Schluter M, Tubler T, Steffens JC, Mathey DG, Schofer J. Focal ischemia of
the brain after neuroprotected carotid artery stenting. J Am Coll Cardiol
2003;42:1007–1013.
Prognostic Importance of Isolated
T-Wave Abnormalities
Takuya Yamazaki,
MD,
Jonathan Myers,
Computerized electrocardiograms recorded on 46,950
male veterans were analyzed to demonstrate the prognostic value of T-wave amplitude in a general medical
population. There were 3,926 cardiovascular deaths
over 6 years. Multivariate survival analysis allowed the
development of a T-wave amplitude graphic that
provides clinicians with a simple method of estimating
the relative risk for cardiovascular mortality from
T-wave amplitude in limb lead I. 䊚2005 by Excerpta
Medica Inc.
(Am J Cardiol 2005;95:300 –304)
lthough population studies have highlighted the
prognostic value of static T-wave abnormalities in
A
asymptomatic community cohorts,
the clinical im1–3
pression is that such findings are labile and nonspecific.
This is at least partially because most of the studies have
included ST measurements and have been based on older
electrocardiographic technology and scoring systems designed for epidemiologic research. This study was performed in a general medical population to evaluate simple T-wave amplitude measurements from computerized
12-lead electrocardiograms (ECGs) and to compare the
prognostic power of isolated T-wave abnormalities to
other established electrocardiographic measurements
and interpretations.
•••
All ECGs obtained at the Palo Alto Veterans Affairs Medical Center from March 1987 to January
From the Division of Cardiovascular Medicine, Stanford University
Medical Center and Veterans Affairs Health Care System, Palo Alto,
California. Dr. Froelicher’s address is: Veterans Affairs Palo Alto Health
Care System, 3801 Miranda Ave, 111C, Palo Alto, California 94304.
E-mail: vicmd@aol.com. Manuscript received July 14, 2004; revised
manuscript received and accepted August 31, 2004
300
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
PhD,
and Victor F. Froelicher,
MD
2000 were digitally recorded and stored in the General
Electric MUSE ECG management system (GE
Healthcare, Waukesha, Wisconsin). When a patient
had ⬎1 ECG in the database, only the earliest ECG
was considered. Computerized measurements from
the ECGs as well as several computerized electrocardiographic interpretations were extracted. ECGs obtained in an inpatient or outpatient setting or in the
emergency room were coded because they could possibly be from acute clinical events. Women (n ⫽
4,616) and patients with ECGs exhibiting electronic
pacing (n ⫽ 309) or Wolff-Parkinson-White syndrome (n ⫽ 44) were excluded from all analyses. To
make the target population specific for isolated repolarization abnormalities, a subset was created by excluding ECGs exhibiting atrial fibrillation, diagnostic
Q waves, ST depression (ⱕ0.5 mm), left ventricular
hypertrophy, QRS duration ⬎120 ms, left bundle
branch block, right bundle branch block, and intraventricular conduction delay. This primary analysis group
consisted of 27,335 patients, and those remaining (n ⫽
14,662) with abnormal ECGs were considered separately.
Standardized computerized electrocardiographic
criteria as described by the General Electric 12-lead
electrocardiographic analysis program were used for
diagnostic Q waves and bundle branch blocks (see
MUSE/12-lead ECG Physician Program Manual at
http://www.gemedicalsystems.com). Although these
interpretations are usually read by a cardiologist before being released as “confirmed,” we used only the
computer interpretation. T-wave amplitude in 12 leads
was taken as the first deflection (either positive or
negative) after the QRS complex and the maximum
deviation from the PR isoelectric baseline. In addition
to simple voltage measurements, the Minnesota Code
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.099
TABLE 1 Demographics and Electrocadiographic Findings of the Total Study Cohort Classified as to CV Death Status
Variable
Age (yrs)
Body mass index
Inpatient
Heart rate (beats/min)
QRS duration (ms)
Minor T-wave change (MC 5-3, 5-4)
Major T-wave change (MC 5-1, 5-2)
Minor ST change (MC 4-3)
Major ST change (MC 4-1, 4-2)
QT interval (ms)
QTc (ms)
QT dispersion (ms)
QTc dispersion (ms)
Total
(n ⫽ 41,997)
CV Death
(n ⫽ 3,926)
Free of CV
Death
(n ⫽ 38,071)
p Value*
56.9 ⫾ 14.4
27.3 ⫾ 5.4
12,246 (29.2%)
74.4 ⫾ 16.7
96.1 ⫾ 16.7
5,583 (13.3%)
3,378 (8.0%)
24,366 (58.0%)
5,723 (13.6%)
386.5 ⫾ 41.5
422.9 ⫾ 27.4
29.1 ⫾ 21.2
30.7 ⫾ 22.3
67.7 ⫾ 11.2
26.8 ⫾ 5.3
1,661 (42.3%)
77.9 ⫾ 17.5
103.9 ⫾ 22.7
824 (21.0%)
757 (19.3%)
2,336 (59.5%)
918 (23.4%)
391.2 ⫾ 45.9
437.6 ⫾ 32.3
28.2 ⫾ 24.1
30.2 ⫾ 25.9
55.7 ⫾ 14.2
27.4 ⫾ 5.5
10,585 (27.8%)
74.1 ⫾ 16.6
95.3 ⫾ 15.7
4,759 (12.5%)
2,621 (6.9%)
22,030 (57.9%)
4,805 (12.6%)
386.0 ⫾ 41.0
421.4 ⫾ 26.3
29.1 ⫾ 20.9
30.7 ⫾ 21.9
⬍0.0001
0.0001
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
0.05
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.01
0.22
*p values are for comparisons between the subjects who experienced CV death and those free from CV death.
MC ⫽ Minnesota code.
TABLE 2 Demographic and Clinical Characteristics of Primary Normal and Abnormal ECG Groups
Variable
CV death
Annual mortality
Age (yrs)
Body mass index
Inpatient
Heart rate (beats/min)
QRS duration (ms)
QT interval (ms)
QTc (ms)
QT dispersion (ms)
QTc dispersion (ms)
Minor T-wave change (MC 5-3, 5-4)
Major T-wave change (MC 5-1, 5-2)
Minor ST change (MC 4-3)
Major ST change (MC 4-1, 4-2)
Atrial fibrillation
Left ventricular hypertrophy
Pathologic Q wave
QRS duration ⬎120 ms
Left bundle branch block
Right bundle branch block
Intraventricular conduction delay
Primary Normal Group
(n ⫽ 27,335)
Abnormal ECG Group*
(n ⫽ 14,662)
1,431 (5.2%)
0.7
53.6 ⫾ 14.0
27.2 ⫾ 5.4
7,364 (26.9%)
73.0 ⫾ 15.2
91.5 ⫾ 9.4
384.3 ⫾ 37.7
417.4 ⫾ 22.7
29.6 ⫾ 19.0
31.1 ⫾ 19.8
2,817 (10.3%)
799 (2.9%)
16,286 (59.6%)
—
—
—
—
2,495 (17.0%)
2.9
63.0 ⫾ 13.1
27.5 ⫾ 5.6
4,882 (33.3%)
77.0 ⫾ 19.0
104.7 ⫾ 22.7
390.6 ⫾ 47.6
433.2 ⫾ 32.0
28.1 ⫾ 25.0
29.9 ⫾ 26.4
2,766 (18.9%)
2,579 (17.6%)
8,080 (55.1%)
5,723 (39.0%)
1,266 (8.6%)
2,004 (13.7%)
5,431 (37.0%)
—
—
—
565 (3.9%)
1,615 (11.0%)
1,435 (9.8%)
p Value
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0005
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
⬍0.0001
*Includes ECGs exhibiting atrial fibrillation, pathological Q waves, left ventricular hypertrophy, ST depression, left bundle branch block, right bundle branch block,
intraventricular conduction delay, and QRS duration ⬎120 ms.
Abbreviation as in Table 1.
was determined using computer algorithms (4-1 and
4-2 as major ST depression, 4-3 as minor ST depression, 5-1 and 5-2 as a major T-wave abnormalities,
and 5-3 and 5-4 as minor T-wave abnormalities).4
Left ventricular hypertrophy was defined using the
Romhilt-Estes point score as ⬎3.5
The Social Security Death Index and the California
Health Department Service were used to ascertain
vital status as of December 30, 2000. The main outcome measure was cardiovascular (CV) mortality.
Descriptive statistics were used to find mean values
for continuous variables and to test for normality.
Bivariate associations between those who experienced
CV death and all others were tested using chi-square
tests for categorical data and Student’s t test for con-
tinuous variables. A p value ⬍0.05 was considered
significant.
Cox proportional-hazards testing was performed to
assess the significance and independence of predictors
of CV mortality, and models were adjusted for age
and heart rate. The T-wave amplitude of standard
leads, as well as other established electrocardiographic indicators, was considered in the model. Limb
lead aVR was excluded from all analysis because it is
a reciprocal lead, and the remaining 11 leads were
considered to determine the representative lead. Multivariate models included T-wave amplitude, QRS duration, QT interval, QT dispersion, and minor ST
depression (Minnesota code 4-3) for the analyses in
the primary normal electrocardiographic group. The
BRIEF REPORTS
301
had the greatest predictability, limb
lead I was the most powerful predictor in the 2 groups. If lead I was
Hazard Ratio for
removed, lead V5 took its place. For
CV Death
each
1-mm (0.1-mV) decrease in amVariable
(95% CI)
p Value
plitude, the risk for CV death inPrimary normal ECG group
creased 33%. The histogram of TT-wave ampitude in limb lead I†
0.71 (0.68–0.74)
⬍0.0001
wave amplitude in the primary group
QT interval (per 5 ms)
1.03 (1.03–1.05)
⬍0.0001
was distributed as a bell-shaped
QRS duration (per 5 ms)
1.04 (1.01–1.06)
⬍0.0001
Minor ST depression
—
NS
curve with average amplitude of 1.8
QT dispersion
—
NS
⫾ 1.1 mm.
Abnormal ECG group
Other electrocardiographic mea†
T-wave ampitude in limb lead I
0.82 (0.80–0.84)
⬍0.0001
surements, such as ST and T changes
QRS duration ⬎120 ms
1.52 (1.39–1.67)
⬍0.0001
based on the Minnesota code, QRS
Pathologic Q wave
1.48 (1.36–1.61)
⬍0.0001
Left ventricular hypertrophy
1.36 (1.22–1.51)
⬍0.0001
duration, QT interval, and QT disQT interval
—
NS
persion, were also significantly assoQT dispersion
—
NS
ciated with outcome when consid*Data are from the Cox proportional-hazards model with age and heart rate adjustment. The variables
ered univariately adjusted for age
are listed in the order of the model selection in both groups.
and heart rate. Because limb lead I
†
T-wave amplitude was evaluated as a continuous variable in millimeters.
demonstrated significant ability to
predict CV death, it was entered into
multiple models to examine the efentire analysis was repeated in the abnormal electro- fect and interaction with other electrocardiographic
cardiographic group with the common electrocardio- characteristics (Table 3). It was found to be signifigraphic abnormalities and indicators in the model. The cantly and independently associated with CV mortalsubset analysis was also repeated only in outpatients ity, with a hazard ratio of 0.71 (95% confidence into remove all possible ECGs associated with acute terval [CI] 0.68 to 0.74) for the normal ECG group.
clinical events.
Surprisingly, in the abnormal ECG group, simple TRelative risks (RRs) were calculated for candidate wave amplitude outperformed commonly used elecamplitudes to determine criteria for practical clinical trocardiographic indicators, including QRS duration,
utility. The RR of each threshold was calculated to diagnostic Q waves, left ventricular hypertrophy, and
demonstrate incremental risk per each downward step ST depression (hazard ratio 0.82, 95% CI 0.80 to
in amplitude. Kaplan-Meier survival curves were plot- 0.84). The predictive value of simple T-wave amplited accordingly.
tude was demonstrated in the absence of other elecDemographics of the total study cohort (41,997 male trocardiographic abnormalities or when they were
veterans, mean age 57 years) are listed in Table 1. Most present. In addition, even when T-wave amplitude
were outpatients when their ECGs were obtained. relative to QRS voltage was taken into account, the
There were 3,926 CV deaths during a mean follow-up results were similar.
of 6 years. Patients who experienced CV death were
RRs for lead I T-wave amplitude thresholds were
older and had higher heart rates and longer QRS
plotted using a T-wave amplitude of 1.0 to 2 mm as
duration and QT intervals compared with patients free
the reference group. Examination of this plot showed
from CV death. ST- and T-wave abnormalities were
that the RR did not decrease significantly for T waves
more prevalent in those who experienced CV death.
The characteristics of electrocardiographic indica- ⬎1 mm, nor did the risk increase significantly for T
tors and measurements in the subsets with otherwise waves ⬍0.5 mm. These thresholds were then used for
primary and abnormal ECGs are listed in Table 2. The Figure 1, which provides a simple illustration of our
primary normal ECG group consisted of 27,335 male main findings for clinical application showing the RR
veterans (mean age 54 years) who experienced 1,431 of each T-wave amplitude cut point with ⬎1.0 mm as
CV deaths during a mean follow-up of 6 years. Av- the reference group. There was a progressive increase
erage annual CV mortality was 0.7% per year in the of risk as amplitude decreased, flattened, and inverted.
primary group compared with 2.9% in the abnormal From ⬎1.0 to 0.5 to 1.0 mm, the RR CI was 1.2 to 1.6;
group. The abnormal ECG group had greater mortal- to 0.0 to 0.5 mm, the RR was CI 1.9 to 2.5; and to an
ity, age, heart rate, QRS duration, and prevalence of inverted T wave, the RR CI was 2.9 to 3.8 (p ⬍
T-wave abnormalities. Most inpatient ECGs were for 0.0001).
Survival plots are shown in Figure 2 using the
routine presurgical evaluation, and hospital status did
T-wave amplitude cut points of ⬍0 mm ⫽ 1, 0 to 0.5
not affect the results of analyses.
Adjusting for age and heart rate in the Cox model, mm ⫽ 2, 0.5 to 1.0 mm ⫽ 3, and ⬎1.0 mm ⫽ 4. A
the T-wave amplitude in each lead was considered significant and quantitative separation was demonunivariately. All leads but leads III and V2 were sig- strated among those groups with amplitude categories
nificantly associated with CV mortality in the primary in the primary and abnormal groups. The annual CV
and abnormal ECG groups. When the leads were mortality for inverted T waves was 3.3% in the normal
considered in the same model to determine which lead ECG group and 5.4 in the abnormal ECG group,
TABLE 3 Multivariate Analyses of Electrocardiographic Criteria Adjusted for Age
and Heart Rate*
302 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
FIGURE 1. RR of T-wave amplitude, using 0.5 to 1.0 mm as the
reference group. From >1.0 to 0.5 to 1.0 mm, RR CI was 1.2 to
1.6; to 0.0 to 0.5 mm, RR CI was 1.9 to 2.5; and to inverted T
wave, RR CI was 2.9 to 3.8 (p <0.0001).
FIGURE 2. Kaplan-Meier survival curves for cut points of T-wave
amplitude in limb lead I in the primary (normal ECG) and abnormal ECG groups. In all analyses, the stratification according to
T-wave amplitude discriminated among groups of subjects with
significantly different mortality rates; that is, the survival rate
was lower as T-wave amplitude decreased (p <0.001).
whereas the annual CV mortality associated with amplitude ⬎1.0 mm was 0.5% and 1.6%, respectively.
•••
T-wave abnormalities are found in up to 15% of
apparently healthy subjects and more frequently in
general hospital and clinical settings. In our clinical
study, 21% of patients were found to have T-wave
abnormalities in the entire cohort, with 13% of those
with otherwise normal ECGs and 40% of those with
abnormal ECGs.
From the Framingham study, isolated T-wave flattering or inversion carried a significant increased risk
for morbidity and mortality.1 Dutch researchers used
the Minnesota code to classify the ST-T changes of
nearly 10,000 apparently healthy men and women.2
They found that ST changes were significantly associated with CV mortality, and isolated T-wave abnormalities had borderline significance. The prognostic
impact of isolated minor ST and/or T-wave abnormalities has also been demonstrated in the Chicago Western Electric Company Study cohort, which included
annual ECGs over a 5-year baseline period.6 It was
found that the repeated occurrence of minor ST-T
abnormalities was associated with greater risk than
abnormalities that were only present transiently.7 Epidemiologists have used data from the Multiple Risk
Factor Intervention Trial to demonstrate the prognostic significance of isolated minor T-wave abnormalities after excluding other coexisting abnormalities.8
Researchers in The Netherlands reported the predictive value of T-wave amplitude and ST-segment level
using lead I in a community population.9 T-wave
amplitude in lead I had a hazard ratio of 2 for ⬍0.5
mm and 1/3 for 1.5 mm after adjustment for clinical
risk factors. Our study confirms the findings of these
community epidemiologic studies but extends the importance of T-wave amplitude to general medical populations.
A recent review focused on the ventricular repolarization components on surface ECGs, including ST
segments and T and U waves.10 The investigators
concluded that the T wave represents the transmural
dispersion of repolarization.
Clinicians consider T-wave abnormalities as “nonspecific” because so many factors affect T-wave amplitude. Serum electrolytes, medications, hyperventilation, intracranial lesions, and recent food ingestion
can all produce T-wave amplitude variation. Large
T-wave amplitude can be seen in athletes,11 whereas
the experimental blocking of parasympathetic activity
by the administration of atropine results in a decrease
of T-wave amplitude.12 A high prevalence of early
repolarization (elevated ST and T wave) has been
demonstrated in those with spinal cord injuries, most
likely because of enhanced vagal tone or the loss of
sympathetic tone.13
Our ECGs were obtained from inpatients and outpatients and represent findings from a very broad
range of patients similar to the type seeking general
medical evaluation. This contrasts with previous studies of patients, which focused on patients with specific
conditions or considered community epidemiologic
cohorts. Our sample is representative of everyday
medical practice, in which electrocardiography is
commonly used as the first assessment tool for possible cardiac disease. Although baseline clinical and
laboratory data were not available, it has been well
validated that electrocardiography14 and specifically
the T wave9 add additional prognostic information to
baseline clinical and laboratory data. Furthermore,
requiring other tests, comparing T-wave amplitude
with the results of other tests, or using other test
results as surrogate end points would be very biased
because these tests are performed for more specific
clinical reasons other than electrocardiography.
1. Kannel WB, Anderson K, McGee DL, Degatano LS, Stampfer MJ. Nonspecific
electrocardiographic abnormality as a predictor of coronary heart disease: the
Framingham Study. Am Heart J 1987;113:370 –376.
2. De Bacquer D, De Backer G, Kornitzer M, Myny K, Doyen Z, Blackburn H.
Prognostic value of ischemic electrocardiographic findings for cardiovascular
mortality in men and women. J Am Coll Cardiol 1998;32:680 – 685.
3. Harlan WR, Cowie CC, Oberman A, Mitchell RE, MacIntyre NR. Prediction
of subsequent ischemic heart disease using serial resting electrocardiograms.
Am J Epidemiol 1984;119:208 –217.
BRIEF REPORTS
303
4. Prineas RJ, Crow RS, Blackburn H. The Minnesota Code Manual of Electrocar-
9. Dekker JM, Schouten EG, Klootwijk P, Pool J, Kromhout D. ST segment and
diographic Findings. Boston, Massachusetts: John Wright PSG; 1982:223–229.
5. Romhilt DW, Estes EH Jr. A point-score system for the ECG diagnosis of left
ventricular hypertrophy. Am Heart J 1968;75:752–758.
6. Daviglus ML, Liao Y, Greenland P, Dyer AR, Liu K, Xie X, Huang CF,
Prineas RJ, Stamler J. Association of nonspecific minor ST-T abnormalities with
cardiovascular mortality: the Chicago Western Electric Study. JAMA 1999;10:
281:530 –536.
7. Greenland P, Xie X, Liu K, Colangelo L, Liao Y, Daviglus ML, Agulnek AN,
Stamler J. Impact of minor electrocardiographic ST-segment and/or T-wave
abnormalities on cardiovascular mortality during long-term follow-up. Am J
Cardiol 2003;91:1068 –1074.
8. Prineas RJ, Grandits G, Rautaharju PM, Cohen JD, Zhang ZM, Crow RS,
MRFIT Research Group. Long-term prognostic significance of isolated minor
electrocardiographic T-wave abnormalities in middle-aged men free of clinical
cardiovascular disease (the Multiple Risk Factor Intervention Trial [MRFIT]).
Am J Cardiol 2002;90:1391–1395.
T wave characteristics as indicators of coronary heart disease risk: the Zutphen
Study. J Am Coll Cardiol 1995;25:1321–1326.
10. Yan GX, Lankipalli RS, Burke JF, Musco S, Kowey PR. Ventricular repolarization components on the electrocardiogram: cellular basis and clinical significance. J Am Coll Cardiol 2003;42:401– 409.
11. Bjornstad H, Storstein L, Meen HD, Hals O. Electrocardiographic findings in
athletic students and sedentary controls. Cardiology 1991;79:290 –305.
12. Annila P, Yli-Hankala A, Lindgren L. Effect of atropine on the QT interval
and T-wave amplitude in healthy volunteers. Br J Anaesth 1993;71:736 –737.
13. Marcus RR, Kalisetti D, Raxwal V, Kiratli BJ, Myers J, Perkash I, Froelicher
VF. Early repolarization in patients with spinal cord injury: prevalence and
clinical significance. J Spinal Cord Med 2002;25:33–38.
14. Ashley EA, Raxwal VK, Froelicher VF. The prevalence and prognostic
significance of electrocardiographic abnormalities. Curr Probl Cardiol
2000;25:1–72.
Relation Between Effects of Adenosine on Brachial
Artery Reactivity and Perfusion Pattern in Patients
With Known or Suspected Coronary Artery Disease
Deval Mehta, MD, Gurpreet Baweja, MD, Rajesh Venkataraman, MD,
Gilbert J. Zoghbi, MD, Thein Htay, MD, Jaekyeong Heo, MD, Navin C. Nanda,
and Ami E. Iskandrian, MD
This study examined the changes in brachial artery
diameter and flow velocity in response to intravenous
adenosine and compared the results to cuff occlusion
and single-photon emission computed tomographic
(SPECT) images. The change in diameter was less with
adenosine than with cuff occlusion. There was no
correlation between the presence of abnormal SPECT
images and the responses to adenosine or cuff occlusion in either diameter or flow velocity. 䊚2005 by
Excerpta Medica Inc.
(Am J Cardiol 2005;95:304 –307)
he purpose of this study was to examine the
changes in the brachial artery dimensions and flow
T
velocity during adenosine infusion and compare the
results to those obtained after cuff occlusion of the
brachial artery and to myocardial perfusion imaging.
•••
The study group consisted of patients with known or
suspected coronary artery disease (CAD) referred to our
stress test laboratory for adenosine stress single-photon
emission computed tomographic (SPECT) imaging for
clinical indications. Pertinent demographic data and
medical histories were documented at enrollment in the
study. Our institutional review board approved the study
protocol, and all patients signed informed consent forms
From the Division of Cardiovascular Diseases, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama;
and the Sarver Heart Institute, Tucson, Arizona. This study was supported by an unrestricted grant from Fujisawa Healthcare, Inc., Deerfield, Illinois. Dr. Iskandrian’s address is: University of Alabama at
Birmingham, 318 LHRB, 1900 University Blvd., Birmingham,
Alabama 35294-0006. E-mail: aiskand@uab.edu. Manuscript received June 28, 2004; revised manuscript received and accepted
August 26, 2004.
304
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MD,
before participating in the study. There were no complications related to the study.
Brachial artery reactivity was assessed at the time
of adenosine SPECT imaging, using the technique
described by Celermajer et al.1 Patients fasted for ⱖ6
hours, and all medications were withheld the morning
of the study. Patients were placed in a supine position,
and the right arm was used for the brachial artery
measurements. The brachial artery was scanned longitudinally 2 to 5 cm above the elbow with a 10.0MHz linear array transducer using the Acuson-Aspen
ultrasound system (Siemens, Palo Alto, California) to
find the clearest images of the anterior and posterior
walls of the artery. This area was marked on the skin,
and all subsequent measurements were performed at
the same position. The baseline brachial artery diameter and peak velocity (using color Doppler-guided
pulsewave Doppler) were measured, along with blood
pressure and heart rate. These measurements were
repeated at 1-minute intervals for 5 minutes during
intravenous adenosine infusion at a dose of 0.140
mg · kg⫺1 · min⫺1.
Five minutes after the completion of adenosine
infusion, a blood pressure cuff was placed around the
arm proximal to the position of the ultrasound transducer and inflated to a pressure of 50 mm Hg greater
than the measured systolic pressure. After 5 minutes
of occlusion, the cuff was deflated, and diameter and
velocity measurements were performed immediately
and repeated every minute for a total of 5 measurements. All studies were recorded on S-VHS tapes.
Off-line quantitative analysis was performed using
calibrated electronic calipers to measure brachial artery diameter. The maximal end-diastolic diameter
was measured at the same position for all the readings.
Adenosine-mediated change in diameter was calcu0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.100
TABLE 1 Baseline Characteristics of the Patients
Category
Age (yrs)
Men
Body mass index (kg/m2)
Diabetes mellitus
Systemic hypertension
Hyperlipidemia*
Smoker
Systolic blood pressure at rest (mm Hg)
Systolic blood pressure peak (mm Hg)
Heart rate at rest (beats/min)
Heart rate peak (beats/min)
Total
(n ⫽ 93)
Normal SPECT Images
(n ⫽ 59)
Abnormal SPECT Images
(n ⫽ 34)
p Value
60.4 ⫾ 12.7
56 (60%)
27.1 ⫾ 8.5
43 (46%)
73 (79%)
41 (44%)
27 (29%)
135 ⫾ 32
140.0 ⫾ 41.0
75 ⫾ 13.0
88 ⫾ 16
59.0 ⫾ 12.8
32 (54%)
27.7 ⫾ 7.5
28 (47%)
45 (76%)
21 (36%)
14 (24%)
138 ⫾ 35
143 ⫾ 47.3
76.0 ⫾ 13
89 ⫾ 16
62.9 ⫾ 12.5
24 (71%)
26.1 ⫾ 10.1
15 (44%)
28 (82%)
20 (59%)
13 (38%)
130 ⫾ 25
135 ⫾ 27
73 ⫾ 13
86 ⫾ 16
0.16
0.13
0.36
0.83
0.60
0.03
0.16
0.20
0.36
0.25
0.33
*Total cholesterol ⬎200 mg/dl or taking lipid-lowering medicines.
Results are mean ⫾ SD or numbers and percentages (in parentheses).
FIGURE 1. Correlation between adenosine- and cuff occlusion–
induced changes in brachial artery diameter.
lated as: (peak diameter during adenosine infusion ⫺
baseline diameter)/baseline diameter ⫻ 100. Change
in diameter after cuff occlusion was calculated as:
(peak diameter after cuff occlusion ⫺ baseline diameter)/baseline diameter ⫻ 100. The percentage
changes in velocity with adenosine and after cuff
occlusion were also measured in a similar fashion.
Intraobserver variation assessments for the brachial
artery measurements were performed in 29 patients
using correlation coefficients, and the variation was
found to be small (r ⫽ 0.98, p ⬍0.001). The initial 40
studies were performed by 1 observer and the remaining by a second observer, both of whom were blinded
to the SPECT results.
Patients were injected with technetium-99m tetrofosmin (10 to 35 mCi) at the 3-minute point during the
adenosine infusion. Gated SPECT images were obtained 60 minutes later using an ADAC dual-head
gamma camera (ADAC Laboratories, Milipitas, California) according to methods previously described
FIGURE 2. Correlation between adenosine- and cuff occlusion–
induced changes in brachial artery velocity.
FIGURE 3. Time distribution of peak dilation in brachial artery
with adenosine and cuff occlusion.
from our laboratory.2 Images at rest were obtained
later during the day or on a separate day in patients
with abnormal stress images. The images were interpreted without knowledge of the brachial reactivity
results. The images were interpreted as normal or
BRIEF REPORTS
305
Brachial artery images were of
suboptimal quality in 5 of 98 patients, and these were excluded. The
Normal SPECT
Abnormal SPECT
baseline characteristics of the reImages
Images
maining 93 patients are listed in TaCategory
(n ⫽ 59)
(n ⫽ 34)
p Value
ble 1. Thirty-four patients had known
Baseline diameter (mm)
4.1 ⫾ 0.1
4.0 ⫾ 0.1
0.79
CAD. Twenty-four had a history of
Percentage diameter change with
8.6 ⫾ 6.4
8.9 ⫾ 6.4
0.84
myocardial infarction, 18 had undercuff occlusion
gone coronary bypass surgery, and
Percentage diameter change with
4.6 ⫾ 5.7
5.1 ⫾ 6.9
0.70
adenosine
10 had undergone percutaneous corBaseline velocity (cm/s)
58.5 ⫾ 32.0
67.2 ⫾ 34.8
0.23
onary intervention. There was a high
Percentage velocity change with
21.0 ⫾ 18.0
26.0 ⫾ 26.5
0.79
prevalence of hypertension, diabetes,
cuff occlusion
smoking, obesity, and family history
Percentage velocity change with
21.0 ⫾ 21.8
19.4 ⫾ 19.6
0.79
of CAD in patients with normal or
adenosine
abnormal images.
The baseline diameter was larger in
men than women (4.2 ⫾ 0.7 vs 3.8 ⫾
1.0 mm, p ⫽ 0.04), but the change in diameter was
similar in the 2 genders with adenosine (5.9 ⫾ 7.3% in
men vs 4.0 ⫾ 5.1% in women, p ⫽ 0.15) and after cuff
occlusion (8.9 ⫾ 7.0% in men vs 8.6 ⫾ 5.9% in women,
p ⫽ 0.85). There was no correlation between age and the
changes in diameter with adenosine (r2 ⫽ 0.002, p ⫽
0.66) or after cuff occlusion (r2 ⫽ 0.013, p ⫽ 0.28). The
changes in diameter with adenosine were not different in
patients with and without diabetes (4.7 ⫾ 6.0% vs 4.8 ⫾
6.3% p ⫽ 0.95), hypertension (4.4 ⫾ 5.7% vs 6.3 ⫾
7.4%, p ⫽ 0.23), or hyperlipidemia (3.8 ⫾ 5% vs 5.5 ⫾
FIGURE 4. Cuff occlusion- and adenosine-mediated brachial ar6.7%, p ⫽ 0.16). The changes in diameter after cuff
tery vasodilatation in patients with (n ⴝ 19) and without (n ⴝ
occlusion were also not different in patients with and
74) evidence of ischemia on their adenosine myocardial perfusion images.
without diabetes (8.2 ⫾ 7.1% vs 9.2 ⫾ 5.7%, p ⫽ 0.45),
hypertension (8.5 ⫾ 5.8% vs 9.7 ⫾ 8.2%, p ⫽ 0.43), or
hyperlipidemia (8.6 ⫾ 7.4% vs 8.6 ⫾ 5.8%, p ⫽ 0.85).
The percentage change in diameter was less with adenosine than with cuff occlusion (4.8 ⫾ 6.1% vs 8.7 ⫾
6.4%, p ⬍0.0001). There was a weak correlation between the change in diameter with adenosine and with
cuff occlusion (r2 ⫽ 0.20, p ⬍0.001; Figure 1). The
changes in velocity were similar with adenosine and cuff
occlusion (20.4 ⫾ 21% vs 23 ⫾ 21.5%, p ⫽ 0.32). There
was also a weak correlation between adenosine- and cuff
occlusion–induced changes in velocity (r2 ⫽ 0.17, p
FIGURE 5. Prevalence of abnormal SPECT images in different
⬍0.001; Figure 2). The maximum change in diameter
quartiles of patients on the basis of adenosine- and cuff occluoccurred
in 65 patients (70%) ⱕ2 minutes after adenosion–induced brachial artery dilatation.
sine infusion and in 71 patients (76%) ⱕ1 minute after
the release of cuff occlusion (Figure 3).
Thirty-four patients had abnormal SPECT scan
abnormal, showing either reversible defect (ischemia),
fixed defect (scar), or both, as previously described. results and 19 patients had evidence of ischemia. The
The extent of ischemia was assessed as the number of baseline measurements of diameter and velocity and
their responses to adenosine and cuff occlusion were
coronary artery territories with reversible defects.
Continuous variables are reported as mean ⫾ SD similar in patients with normal (n ⫽ 59) and abnormal
and discrete data as percentages. Continuous variables (n ⫽ 34) SPECT images (Table 2). The changes in
were analyzed by the unpaired Student’s t test and brachial artery diameter were also similar in patients
discrete variables by the chi-square test. Bonferroni with and without ischemia (Figure 4).
The patients were divided into quartiles on the
correction was used to adjust for multiple comparisons. Pearson’s method was used to estimate the cor- basis of the changes in diameter with adenosine or
relation coefficients between continuous variables. A cuff occlusion. There was no difference in the number
p value ⬍0.05 was considered statistically significant. of patients with abnormal SPECT images in the highAll data were entered into an Excel spreadsheet (Mi- est or lowest quartiles with either adenosine (p ⫽ 0.7)
crosoft Corporation, Redmond, Washington), and the or cuff occlusion (p ⫽ 0.3; Figure 5).
In the 19 patients with myocardial ischemia on the
analysis was performed using the SPSS version 10.1
SPECT images, there was no correlation between the
(SPSS, Inc., Chicago, Illinois) statistical package.
TABLE 2 Comparison Between Patients With Normal and Abnormal SPECT
Images
306 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
JANUARY 15, 2005
extent of ischemia and baseline diameter (r ⫽ 0.04, p ⫽
0.69), change in diameter with adenosine (r ⫽ 0.16, p ⫽
0.88), change in diameter with cuff occlusion (r ⫽ 0.14,
p ⫽ 0.19), baseline velocity (r ⫽ 0.05, p ⫽ 0.67), and
change of velocity with adenosine (r ⫽ 0.05, p ⫽ 0.62)
or with cuff occlusion (r ⫽ 0.15, p ⫽ 0.16).
•••
This is the first report on brachial artery response
during intravenous adenosine infusion performed simultaneously at the time of adenosine gated SPECT
imaging. Our data show that the response of diameter
to adenosine is less than that with cuff occlusion,
suggesting a smaller increase in flow with adenosine
than with cuff occlusion. Moreover, the adenosineinduced changes in velocity were only a fraction of
those observed in the coronary circulation.3–5 These
differences between different vascular beds might be
due to differences in receptor subtype (A-2a) density
in the coronary versus systemic circulation or blood
levels of adenosine in coronary versus brachial artery
after intravenous infusion.6
In our study, we measured diameter and velocity.
The percentage change in velocity was much greater
with adenosine and cuff occlusion than the change in
diameter. However, the SD for the velocity measurement was much wider than that for the change in
diameter measurements. This may suggest that the
change in diameter may be more robust and reproducible than the change in velocity measurements. The
arm position remained stable during the adenosine
study, perhaps making such measurements more reproducible.
Previous studies have demonstrated endothelial
dysfunction in patients with established CAD7 as well
as those with preclinical atherosclerosis, such as those
with coronary risk factors.8 –11 We did not find a
significant difference in adenosine- or cuff occlusion–
induced change in diameter or velocity in patients
with and without diabetes mellitus, hypertension, or
hyperlipidemia. Most previous studies compared pa-
tients with 1 or 2 risk factors with healthy subjects,
whereas most of our patients had multiple risk factors.
Previous studies have suggested that endothelial dysfunction worsens with an increasing number of risk
factors.11,12 Thus, the presence of multiple risk factors
in our cohort may explain the lack of difference in the
brachial artery responses in different subgroups.
1. Celermajer DS, Sorensen KE, Gooch VM, Spiegelhalter DJ, Miller OI,
Sullivan ID, Lloyd JK, Deanfield JE. Non-invasive detection of endothelial
dysfunction in children and adults at risk of atherosclerosis. Lancet 1992;340:
1111–1115.
2. Iskandrian AS, Heo J, Nguyen T, Beer S, Cave V, Cassel D, Iskandrian BB.
Tomographic myocardial perfusion imaging with technetium-99m teboroxime
during adenosine-induced coronary hyperemia: correlation with thallium-201
imaging. J Am Coll Cardiol 1992;19:307–312.
3. Wilson RF, Wyche K, Christensen BV, Zimmer S, Laxson DD. Effects of
adenosine on human coronary arterial circulation. Circulation 1990;82:1595–1606.
4. Iskandrian AS, Verani MS, Heo J. Pharmacologic stress testing: mechanism of
action, hemodynamic responses, and results in detection of coronary artery
disease. J Nucl Cardiol 1994;1:94 –111.
5. Shryock JC, Belardinelli L. Adenosine and adenosine receptors in the cardiovascular system: biochemistry, physiology, and pharmacology. Am J Cardiol
1997;79:2–10.
6. Neunteufl T, Katzenschlager R, Hassan A, Klaar U, Schwarzacher S, Glogar D,
Bauer P, Weidinger F. Systemic endothelial dysfunction is related to the extent
and severity of coronary artery disease. Atherosclerosis 1997;129:111–118.
7. Balletshofer BM, Rittig K, Stock J, Lehn-Stefan A, Overkamp D, Dietz K,
Haring HU. Insulin resistant young subjects at risk of accelerated atherosclerosis
exhibit a marked reduction in peripheral endothelial function early in life but not
differences in intima-media thickness. Atherosclerosis 2003;171:303–309.
8. Celermajer DS, Sorensen KE, Georgakopoulos D, Bull C, Thomas O, Robinson J, Deanfield JE. Cigarette smoking is associated with dose-related and
potentially reversible impairment of endothelium-dependent dilation in healthy
young adults. Circulation 1993;88:2149 –2155.
9. Clarkson P, Celermajer DS, Powe AJ, Donald AE, Henry RM, Deanfield JE.
Endothelium-dependent dilatation is impaired in young healthy subjects with a
family history of premature coronary disease. Circulation 1997;96:3378 –3383.
10. Jensen-Urstad K, Johansson J, Jensen-Urstad M. Vascular function correlates
with risk factors for cardiovascular disease in a healthy population of 35-year-old
subjects. J Intern Med 1997;241:507–513.
11. Celermajer DS, Sorensen KE, Bull C, Robinson J, Deanfield JE. Endotheliumdependent dilation in the systemic arteries of asymptomatic subjects relates to
coronary risk factors and their interaction. J Am Coll Cardiol 1994;24:1468 –1474.
12. Hashimoto M, Kozaki K, Eto M, Akishita M, Ako J, Iijima K, Kim S, Toba
K, Yoshizumi M, Ouchi Y. Association of coronary risk factors and endotheliumdependent flow-mediated dilatation of the brachial artery. Hypertens Res 2000;
23:233–238.
BRIEF REPORTS
307
Usefulness of Real-Time Three-Dimensional
Echocardiography for Reliable Measurement of
Cardiac Output in Patients With Ischemic or Idiopathic
Dilated Cardiomyopathy
Sean M. Fleming,
MB, MD,
Barry Cumberledge, FSCST, Christoph Kiesewetter,
Gareth Parry, MB, and Antoinette Kenny, MB, MD
The determination of stroke volume (SV) is a potentially important application of real-time 3-dimensional echocardiography (RT3DE). SV measurements by thermodilution were compared with
values obtained using transthoracic RT3DE in a
sequential cohort of patients who underwent assessment for potential cardiac transplantation.
There was a strong correlation between echocardiographically derived SV and catheterization data
(r ⴝ 0.95, n ⴝ 14). On average, RT3DE appeared
to underestimate SV by 7.5 ml (SD ⴝ 5.8) or 17%
(SD ⴝ 12%). A role for RT3DE in the measurement
of SV in severe heart failure is suggested. 䊚2005
by Excerpta Medica Inc.
(Am J Cardiol 2005;95:308 –310)
hree-dimensional echocardiography provides accurate measurements of left ventricular (LV) volT
ume and function and has been shown to have particular applicability in distorted and remodeled
ventricles.1,2 The clinical application of 3-dimensional
echocardiography, however, has been limited by prolonged acquisition, reconstruction, and data analysis
times.3 Real-time 3-dimensional echocardiography
(RT3DE) is now commercially available. By eliminating acquisition and reconstruction delays, RT3DE offers the potential to bring 3-dimensional echocardiography into the clinical mainstream. We examined the
application of RT3DE in patients who underwent assessment for cardiac transplantation. We aimed to
compare stroke volume (SV) quantification by
RT3DE with our clinical standard method, bolus thermodilution.
•••
Seventeen sequential patients who underwent
right-sided cardiac catheterization at the Department
of Cardiac Transplantation at Freeman Hospital
(Newcastle upon Tyne, United Kingdom) were enrolled. Images suitable for 3-dimensional analysis
were obtained from 14, and these patients formed the
study cohort.
In this clinically based study, SV as estimated
using bolus thermodilution was taken as the clinical
From the Departments of Echocardiography and Cardiac Transplantation, Freeman Hospital, High Heaton, Newcastle Upon Tyne, United
Kingdom. Dr. Fleming’s address is: Department of Echocardiography,
Freeman Hospital, High Heaton, NE7 7DN, United Kingdom. E-mail:
sfleming@esatclear.ie. Manuscript received April 7, 2004; revised
manuscript received and accepted August 31, 2004.
308
©2005 by Excerpta Medica Inc. All rights reserved.
The American Journal of Cardiology Vol. 95 January 15, 2005
MB, MD,
reference method. In each patient, repeated injections
were performed until 3 measurements were obtained
within a 15% range. The average value was taken as
the cardiac output and divided by the heart rate at the
time of measurement to give SV. Each patient underwent right-sided heart catheterization as part of a
workup for potential cardiac transplantation. Threedimensional echocardiographic volumes were obtained using the Philips Sonos 7500 and x4 transducer
(Philips Medical Systems, Andover, Massachusetts).
This transducer has a fully sampled matrix array with
3,000 physical image-formation channels and a frequency range of 1.6 to 4 MHz. For “full-volume”
acquisition of the left ventricle, a series of 4 electrocardiographic-triggered subvolumes are first acquired
during suspended respiration. These are then synchronized with the R wave to render a time-aligned full
volume data set with a sector size of 60° by 60°. This
process is automated, giving a total acquisition time
equivalent to 4 cardiac cycles. Subsequent rendering
is almost instantaneous, allowing the immediate playback and inspection of the 3-dimensional volume. All
3-dimensional data sets were acquired in the cardiac
catheterization laboratory, immediately after thermodilution measurements were taken and with the
patient in the left lateral decubitus position. At the
time of data acquisition, each data set was visually
inspected to ensure that image quality was acceptable,
that there were no obvious breathing or movement
artifacts, and that as much as possible of the left
ventricle was contained within the image pyramid. In
total, a median of 5 data sets were recorded for each
patient. Data sets that were clearly inadequate were
not stored. Analysis was performed off-line using
semiautomated border detection software (TomTec
Imaging Systems, Munich, Germany). Three-dimensional volumes were analyzed using an 8-plane apical
rotational method.4 End-diastolic volume (EDV) and
end-systolic volume (ESV) were measured and SV
estimated by subtraction. All measurements were performed blinded to the cardiac catheterization data.
When dealing with a large and distorted ventricle,
it may be difficult to fit the entire ventricle within the
3-dimensional data set. Using a 16-segment model,
and in the case of each participant, the presence or
absence of each wall segment within the full-volume
pyramid at end-diastole was noted on a binary scale.
The total number of imaged segments for the cohort as
a whole and the median number for each participant
are reported.
0002-9149/05/$–see front matter
doi:10.1016/j.amjcard.2004.08.101
FIGURE 1. Absolute difference plot: SV estimated by thermodilution less SV by RT3DE is plotted against mean SV. The continuous line represents the average absolute difference between the
methods (in milliliters), and the dashed lines represent 95% confidence intervals for this average.
Statistical analysis was performed using Excel
(Microsoft Corporation, Redmond, Washington).
Descriptive data are reported as medians and interquartile ranges (IQRs). For comparison between
methods, the correlation coefficient is reported.
Bland-Altman plots were constructed, and absolute
and percentage differences ⫾2 SDs are reported.
All 3-dimensional measurements were performed
twice by the first investigator (SMF). These results
were averaged to provide the measurements of LV
volumes used in the study. Intraobserver error is
reported as the average and SE of the percentage
difference between each pair of measurements.
Analysis was repeated by another investigator (CK)
for a random 10 studies, and interobserver error is
reported in a similar fashion.
In total, 14 patients (10 men) participated (median
age 50 years, range 28 to 62). Seven patients had
ischemic cardiomyopathy and 7 had idiopathic dilated
cardiomyopathy. Seven patients were in New York
Heart Association symptom class IV, and 7 were in
class III. At standard transthoracic echocardiographic
scanning, the median end-diastolic dimension was 66
mm (IQR 60 to 75) and the median ejection fraction
was 12.5% (IQR 12.5% to 21%). Four patients had
significant valvular regurgitation (1 moderate mitral
regurgitation, 2 severe mitral regurgitation, 1 moderate to severe aortic regurgitation). No patient had
significant pulmonary regurgitation. The cardiac
rhythm was sinus for all but a single participant, who
had atrial fibrillation. At right-sided cardiac catheterization, the median pulmonary artery wedge pressure
was 30 mm Hg (IQR 24 to 32), the median mean
pulmonary artery pressure was 39.5 mm Hg (IQR 28.5
to 46), the median cardiac output was 3.25 L/min
(IQR 2.9 to 3.7), and the median SV was 41 ml (IQR
35 to 57).
At RT3DE, the median EDV was 218 ml (IQR 160
to 281), the median ESV was 191 ml (IQR 119 to
220), the median SV was 36 ml (IQR 29 to 50), and
the median LV ejection fraction was 16% (IQR 13%
to 23%). Intraobserver average percentage differences
were 2.1% (SE 0.9%) for EDV, 1.5% (SE 1.4%) for
ESV, and 2.3% (SE 4.2%) for SV (n ⫽ 14). Interobserver average percentage differences were 2.9% (SE
2.9%) for EDV, 4.8% (SE 4.1%) for ESV, and 1.6%
(SE 6.8%) for SV (n ⫽ 10). In total, using a 16segment model of the left ventricle, 90.2% of the
segments (202 of 224) were contained within the data
set at end-diastole. Of the missing segments, 15 of 22
were apical segments, with the most commonly
missed segment being the apical inferior segment (inadequately imaged in 9 patients). The median number
of segments contained within the pyramid was 14.5
per patient (IQR 13.25 to 15.75).
There was a strong correlation between SV by
RT3DE and SV by thermodilution (r ⫽ 0.95). On the
Bland-Altman plot (Figure 1), SV by RT3DE underestimated SV by thermodilution on average by 7.5 ml
(95% confidence interval 4 to 19). Expressed as a
percentage of the mean value, the average difference
was 17% (⫺5% to 39%). When patients with significant valvular regurgitation were excluded from the
analysis, a marginally tighter correlation between SV
by RT3DE and SV by thermodilution (r ⫽ 0.96, n ⫽
10) emerged. However, the average underestimate on
the Bland-Altman plot of 9 ml (95% confidence interval 2 to 20) was greater than that seen in the entire
cohort. Expressed as a percentage of the mean value,
the average difference in this subgroup was 21%
(⫺1% to 43%).
•••
To date, 3-dimensional echocardiography has
largely implied the reconstruction of 3-dimensional
data sets from multiple 2-dimensional planes. The
acquisition of such data sets is cumbersome, time
consuming, and prone to movement and breathing
artifacts.3 Furthermore, such artifacts may not be apparent at the time of acquisition, because off-line
reconstruction is required before a data set can be
viewed. This impracticality, rather than concerns over
accuracy, has limited the clinical application of 3-dimensional echocardiography. The assessment of LV
function derived from such reconstructed data sets
represents a clear advance on 2-dimensional imaging
and has been extensively validated against different
gold standards.5–10
In the development of 3-dimensional echocardiography, attention has more recently focused on developing transducer technology. This has culminated in
the introduction of so-called matrix arrays, which feature a third dimension to the scanning array, allowing
true real-time 3-dimensional scanning. Initial research
work with a 256-element “sparse-array” matrix has
established the clinical potential of this method.1,2,11–13 However, limitations, as set out by Sugeng
et al4 in their review of real-time imaging, such as
poor image quality, a relatively low frame rate, and
the inability to produce 3-dimensional images on-line,
have precluded its routine use. The recent release of a
commercial system for RT3DE, however, offers the
potential to bring 3-dimensional echocardiography
into the clinical mainstream. Now, 3-dimensional images of the left ventricle can be acquired and rendered
BRIEF REPORTS
309
on-line. In this study, we present what we believe to
be the first study examining a commercially available
RT3DE system in a clinically relevant patient group.
We show that images suitable for analysis were produced in 14 of 17 unselected patients (83%) with
severe heart failure. We demonstrate excellent reproducibility; in these 14 patients, inter- and intraobserver mean absolute errors were small (⬍5%) in a
blinded analysis. In addition, we demonstrate a close
correlation between RT3DE and our clinical reference
method (r ⫽ 0.95). Excluding participants with significant left-sided regurgitation resulted in a marginal
improvement in the observed correlation. However,
overall, a difference between the methods emerged,
with RT3DE appearing to underestimate SV as assessed by thermodilution. We chose thermodilution as
a reference method, because at our institution it is the
method of choice for assessing cardiac function in the
setting of potential cardiac transplantation, and we
wished to make a clinically based comparison. Our
choice of thermodilution as a reference method, however, is a potential limitation to this study. There is
evidence that thermodilution tends to overestimate SV
by approximately 20%.14 This overestimation may be
particularly evident in severe heart failure.15 Two previous studies reported SV estimation by 3-dimensional echocardiographic techniques in comparison
with thermodilution in humans. The 2 studies used a
transesophageal approach and off-line reconstruction
and analysis. In the 2 studies, a similar difference
between methods as found in our study was noted,
with 3-dimensional echocardiography seeming to underestimate SV by 8 ml in patients with coronary
artery disease (n ⫽ 12) and by 6.4 ml in an intensive
care setting (n ⫽ 20).9,10 Taking all 3 studies together, a remarkably consistent difference is seen
between 3-dimensional echocardiography and thermodilution, with the echocardiographic method appearing to underestimate the invasively derived result by approximately 20%. Although we do not
make a comparison with a gold standard reference
method, it is notable that our study is consistent
with previous transesophageal studies and that the
reported percentage difference in all three 3-dimen-
310 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 95
sional echocardiographic studies approximates the
known overestimation of thermodilution.
1. Qin JX, Jones M, Shiota T, Greenberg NL, Tsujino H, Firstenberg MS, Gupta
PC, Zetts AD, Xu Y, Ping Sun J, et al. Validation of real-time three-dimensional
echocardiography for quantifying left ventricular volumes in the presence of a left
ventricular aneurysm: in vitro and in vivo studies. J Am Coll Cardiol 2000;36:
900 –907.
2. Lee D, Fuisz AR, Fan PH, Hsu TL, Liu CP, Chiang HT. Real-time 3-dimensional
echocardiographic evaluation of left ventricular volume: correlation with magnetic
resonance imaging—a validation study. J Am Soc Echocardiogr 2001;14:1001–1009.
3. Siu SC, Rivera JM, Handschumacher MD, Weyman AE, Levine RA, Picard
MH. Three-dimensional echocardiography: the influence of number of component images on accuracy of left ventricular quantification. J Am Soc Echocardiogr
1996;9:147–155.
4. Sugeng L, Weinert L, Lang RM. Left ventricular assessment using real time
three dimensional echocardiography. Heart 2003;89(suppl):iii29 –iii36.
5. Siu SC, Rivera JM, Guerrero JL, Handschumacher MD, Lethor JP, Weyman
AE, Levine RA, Picard MH. Three-dimensional echocardiography in vivo validation for left ventricular volume and function. Circulation 1993;88:1715–1723.
6. Siu SC, Levine RA, Rivera JM, Xie SW, Lethor JP, Handschumacher MD,
Weyman AE, Picard MH. Three-dimensional echocardiography improves noninvasive assessment of left ventricular volume and performance. Am Heart J
1995;130:812– 822.
7. Jiang L, Vazquez de Prada JA, Handschumacher MD, Vuille C, Guererro JL,
Picard MH, Joziatis JT, Fallon JT, Weyman AE, Levine RA. Quantitative
three-dimensional reconstruction of aneurysmal left ventricles. In vitro and in
vivo validation. Circulation 1995;91:222–230.
8. Buck T, Hunold P, Wentz KU, Tkalec W, Nesser HJ, Erbel R. Tomographic
three-dimensional echocardiographic determination of chamber size and systolic
function in patients with left ventricular aneurysm. Circulation 1997;96:4286 – 4297.
9. Kuhl HP, Franke A, Janssens U, Merx M, Graf J, Krebs W, Reul H, Rau G,
Hoffmann R, Klues HG, et al. Three-dimensional echocardiographic determination of left ventricular volumes and function by multiplane transesophageal
transducer: dynamic in vitro validation and in vivo comparison with angiography
and thermodilution. J Am Soc Echocardiogr 1998;11:1113–1124.
10. Muller S, Bartel T, Baumann G, Erbel R. Preliminary report: evaluation of
three-dimensional echocardiographic volumetry by simultaneous thermal dilution
in coronary heart disease. Cardiology 1996;87:552–559.
11. Shiota T, McCarthy PM, White RD, Qin JX, Greenberg NL, Flamm SD,
Wong J, Thomas JD. Initial clinical experience of real-time three-dimensional
echocardiography in patients with ischaemic and idiopathic dilated cardiomyopathy. Am J Cardiol 1999;84:1068 –1073.
12. Schmidt MA, Freidlin RZ, Ohazama CJ, Jones M, Laurienzo JM, Brenneman
CL, Norman JE, von Ramm OT, Panza JA. Anatomic validation of a novel
method for left ventricular volume and mass measurements with use of real-time
3-dimensional echocardiography. J Am Soc Echocardiogr 2001;14:1–10.
13. Mondelli JA, Di Luzio S, Nagaraj A, Kane BJ, Smulevitz B, Nagaraj AV,
Greene R, McPherson DD, Rigolin VH. The validation of volumetric real-time
3-dimensional echocardiography for the determination of left ventricular function. J Am Soc Echocardiogr 2001;14:994 –1000.
14. Russell AE, Smith SA, West MJ, Aylward PE, McRitchie RJ, Hassam RM,
Minson RB, Wing LM, Chalmers JP. Automated non-invasive measurement of
cardiac output by the carbon dioxide rebreathing method: comparison with dye
dilution and thermodilution. Br Heart J 1990;63:195–199.
15. Espersen K, Jensen EW, Rosenborg D, Thomsen JK, Eliasen K, Olsen
NV, Kanstrup IL. Comparison of cardiac output techniques: thermodilution,
Doppler, CO2 rebreathing and the direct Fick method. Acta Anaesth Scand
1995;39:245–251.
JANUARY 15, 2005