Tuberculous Pericardial Effusion

clinical Study
Tuberculous Pericardial Effusion
GS SAINANI, RAJESH G SAINANI
Abstract
Tuberculous pericardial effusion is common in Afro-Asian countries. Since, the introduction of HIV infection, the incidence
of tuberculous pericardial effusion has increased not only in Afro-Asian countries but also the world over. It presents with
the usual features of tuberculous infection (low grade fever, loss of appetite, loss of weight) along with features of pericardial
effusion (dyspnea, cough and enlarged heart). The salient features of pericardial effusion are low volume pulse or even
pulsus paradoxus, raised jugular venous pressure Kussmaul’s sign, congestive hepatomegaly, ascites and edema over legs. In
massive pericardial effusion, patient may go into cardiac tamponade when patient is breathless, restless with poor volume
pulse (typical paradoxus), engorged neck veins, sinus tachycardia, fall in blood pressure. Urgent pericardial paracentesis is
warranted to reverse the hemodynamic changes with improvement in symptoms and signs. Laboratory tests reveal raised
absolute lymphocyte count, raised ESR, cardiomegaly on X-ray chest, low voltage and sinus tachycardia on ECG, Echo-free
space seen between two pericardial layers on 2D-echo with heart floating in pericardial sac. Diagnostic pericardial paracentesis
shows that pericardial fluid is lymphocytic exudate, with elevated ADA and IFN-g levels. Tubercle bacilli may be isolated
on culture, guinea pig inoculation and nowadays by PCR technique. For management of tuberculous pericardial effusion,
antituberculous treatment with four standard drugs is started. Pericardial paracentesis with needle or even open drainage is
useful in relieving symptoms and rapid recovery. Adjunctive corticosteroids are useful for rapid recovery and for prevention
of development of constrictive pericarditis.
Keywords: Nasopharynx, foreign bodies, sewing needle
T
his is a common cause of chronic pericardial
effusion particularly in Afro-Asian countries.
It has become more common since occurrence
of HIV infection. It is relatively less common in welldeveloped countries. (USA, UK, Canada, Europe).
Infection is either blood-borne or spread from
neighboring structures like lymph nodes, lungs and
bronchi. The patient presents with general features of
tuberculosis such as low-grade fever, loss of appetite,
loss of weight and fatigability, cough along with
features of pericardial effusion.1
Clinical features of pericardial effusion
Symptoms vary according to the amount of pericardial
fluid which may vary from 150 ml to 1,000 ml. Dyspnea
is very common symptom and its severity will vary
according to the quantity of fluid. The dyspnea is due to
compression of the adjacent lungs and bronchi. Patient
feels worse in lying down position and feels better in
sitting and leaning forward as that posture reduces the
compression of lungs and bronchi. Cough is another
Dept. of Medicine, Jaslok Hospital and Research Centre, Mumbai
Address for correspondence
Dr GS Sainani
201, Buena Vista
Gen. Jagannath Bhosle Road, Mumbai - 400 021
E-mail: drsainani@vsnl.com
common symptom and is attributable to compression
of bronchi. Rarely, one may encounter hoarseness of
voice or difficulty in swallowing due to pressure on
recurrent laryngeal nerve or esophagus (seen in massive
pericardial effusion).
On physical examination, patient looks, thin, emaciated,
febrile, dyspneic and usually in propped up position.
Pulse is small in volume or may have typical pulsus
paradoxus (poor volume in inspiration and slightly
better volume in expiration). Blood pressure is on
lower side. There may be distension of neck veins with
presence of Kussmaul’s sign (Jugular venous pressure
rising during inspiration and lower during expiration).
There may be tender hepatomegaly and in some chronic
cases there is free fluid in abdomen (ascites). There is
invariably pitting edema over both legs.2
Precordial examination may show slight bulging
(fullness of intercostal spaces), cardiac impulse and
other pulsations usually not visible. Apex beat may be
just palpable. On percussion heart borders are enlarged.
On left side, one may get dull percussion note beyond
apex beat; on right side, one gets dull percussion note on
right side of the sternum particularly in 5th intercostal
space (Rotch’s sign). One may be able to demonstrate
shifting dullness in left second space which is dull in
recumbent position and becomes less dull in sitting
position due to shifting of fluid down. In some cases
Indian Journal of Clinical Practice, Vol. 22, No. 8, January 2012
371
clinical study
of massive pericardial effusion, one can demonstrate
Ewart’s sign (Dull note along with bronchial breathing
and egophony over left infrascapular region due to
compression of the lower lobe of left lung by pericardial
fluid). On auscultation, heart sounds are distant. Rarely,
one may hear a pericardial rub.2
In massive pericardial effusion, patient may drift into
cardiac tamponade, which should be recognized and
managed, on emergent basis. The clinical features of
tamponade are due to impaired diastolic filling leading
to low cardiac output due to increased intrapericardial
pressure. It should be stressed that presentation of
cardiac tamponade is due to rapidity with which the
fluid collects rather than total quantity of fluid. If fluid
starts collecting gradually over a long period, the heart
adopts and cardiac tamponade does not develop. But
if fluid collects rapidly, ventricular filling is impaired
which leads to increase in diastolic pressure resulting
in rise in atrial pressure, fall in stroke volume and
cardiac output. That is reflected in compensatory sinus
tachycardia and fall in systolic blood pressure. Increase
in atrial pressures leads to rise in jugular venous
pressure. Patient gets severe breathlessness in recumbent
position (orthopnea) and he finds some comfort in
sitting position and leaning forward. He may develop
cyanosis. Heart sounds are distant and one may hear III
heart sound. Pericardial paracentesis (preferably under
echocardiographic guidance) is urgently warranted to
reverse the hemodynamic alterations.1,2
Investigations



372
Blood: Total leukocyte count is usually normal with
lymphocyte predominance. Absoute lymphocyte
count is raised, erythrocyte sedimentation rate
(ESR) is raised.
X-ray: Chest shows enlarged cardiac shadow. In
massive effusion, heart looks funnel shaped. The
contours due to chambers and blood vessels are
lost. One can assess the thickness of pericardium
by injecting air in pericardial cavity after aspirating
some fluid (which results in hydropneumopericardium).
MRI: Hayashi et al3 reported a patient in when
gadolinium enhanced magnetic resonance imaging
(MRI) provided useful information. The incidence
of tuberculous pericarditis has decreased but
early diagnosis and treatment are critical because
constrictive pericarditis is a serious complication
which has a poor prognosis.4 MRI has recently
been used to evaluate suspected pericardial heart
Indian Journal of Clinical Practice, Vol. 22, No. 8, January 2012
disease5 and few reports on MRI findings in
tuberculous pericarditis are available.6,7




ECG: There is low voltage and sinus tachycardia,
T waves are usually inverted.
Echocardiography: This is most useful and
confirmatory as one sees echo-free space between
two pericardial layers. In massive pericardial
effusions, one would see heart floating in pericardial
sac.2
Cardiac catheterization: Before the advent of
2D-echo, cardiac catheterization and ventricular
angiography used to be done to confirm and
evaluate the severity of pericardial effusion.2
Diagnostitic pericardial paracentesis: Pericardial
paracentesis for diagnositic and therapeutic
purposes are carried out under echocardiographic
scanning. On aspiration, the fluid is straw colored
or sometimes blood tinged. The fluid is sent for
cytology, biochemistry, Ziehl-Neelson stain and
culture for tubercle bacilli. Guinea pig inoculation
may confirm tuberculous etiology. The fluid is
lymphocytic exudate. (Proteins are increased
and there is predominance of lymphocytes). For
therapeutic purposes, it is carried out in cardiac
tamponade to relieve dyspnea, palpitation.2
Gibbs et al8 in their survey of 10 years experience of
patients of pericardial effusion (who were managed by
percutaneous paracentesis) had 46 patients of pericardial
effusion. Amongst these, they documented 12 cases of
tuberculous pericarditis, nine were Indo-Asian and
three were of Afro-carribean origin. Fever, night sweats
and weight loss were common among these patients.
Pulsus paradoxus was the most important sign (100%)
for the presence of echocardiographic features of
tamponade. According to them pericardiocentesis
remains an effective measure for immediate relief of
symptoms of cardiac tamponade.
Reuter et al9 have stressed on laboratory tests for
diagnosing
tuberculous
pericarditis.
Definitive
diagnosis of tuberculous pericarditis requires isolation
of the tubercle bacillus from pericardial fluid, but
isolating the organism is difficult. The authors studied
233 consecutive cases of pericardial effusion who
underwent a predetermined diagnostic work-up. This
included (i) Clinical examination (ii) pericardial fluid
tests: Biochemistry; microbiology, cytology, differential
white blood cell (WBC) counts, gamma-interferon
(IFN-g), adenosine deaminase (ADA) levels, polymerase
chain reaction (PCR) testing for mycobacterium
tuberculosis (iii) HIV (iv) sputum smear and culture (v)
blood biochemistry and (vi) differential WBC count.
clinical study
These results showed that fever, night sweets, weight
loss, serum globulin (>40 g/l) and peripheral blood
leukocyte count (<10 x 109/l) were independently
predictive. Pericardial fluid IFN-g ≥50 pg/ml
concentration, had 92% sensitivity, 100% specificity and
a positive predictive value (PPV) of 100% for diagnosis
of tuberculous pericarditis. Pericardial fluid ADA
≥40 U/l had 87% sensitivity and 89% specificity.
A diagnositic model including pericardial ADA,
lymphocytic/neurtophic ratio, peripheral leukocyte
count and HIV status had 96% sensitivity and 97%
specificity; substituting pericardial IFN-g for ADA
yielded 98% sensitivity and 100% specificity. They
concluded that pericardial IFN-g is the most useful
diagnostic test. Alternatively they propose a prediction
model that incorporates ADA and differential WBC
counts
(10 transudative ascites, 8 pyopericardium) effusions
evaluated PCR, ADA activity and absolute lymphocyte
count (ALC). Their results showed that fluid PCR for
M. tuberculosis was positive in 74% of tuberculous
effusions, where as it was falsely positive in 13% of
the nontuberculous group. The mean fluid ADA and
ALC values were significantly higher in tuberculous
effusions than in nontuberculous effusions (p < 0.001).
The sensitivity and specificity of PCR, ADA and ALC
were 74% and 88%, 81% and 75%, and 90% and 83%,
respectively in diagnosing tuberculous effusions. They
concluded that fluid PCR should not be relied on as a
single test; rather, combined analysis with ADA or ALC
could be more useful in the diagnosis of tuberculous
effusions in children.
Burgess et al10 have stressed that traditional diagnostic
tests for tuberculous pericarditis are insensitive and
often require long culture periods. Hence, newer tools
such as pericardial ADA and IFN-a were investigated by
Burgess et al and they concluded that pericardial fluid
levels of ADA and IFN-a are useful in the diagnosis of
tuberculous pericarditis.
The diagnosis is confirmed by history of low-grade
fever, malaise, loss of appetite, loss of weight, cough,
dyspnea and clinical signs of pericardial effusion.
X-ray chest, EKG and 2D-echo confirm the diagnosis
of pericardial effusion. Examination of pericardial
fluid would differentiate the other etiological causes
of pericardial effusion. One should rule out viral,
rheumatic, purulent and collagen disorders as causes
of pericarditis. For confirmation of tuberculous
aetiology, one should do culture of M. tuberculosis, do
pericardial fluid PCR, estimate ADA and IFN-g levels in
pericardial fluid. Raised ESR, elevated absolute
lymphocyte count, lymphocytic exudative nature of
pericardial fluid, fluid PCR for M. tuberculosis, elevated
ADA and IFN-a or g help in confirming the tuberculous
etiology of pericardial fluid.
Mathur et al11 carried out study on 120 cases of serosal
effusion (50 pleural, 50 peritoneal and 20 cases of
pericardial effusion). ADA was found significantly
raised with a mean value of 100 U/l, 92 U/l and
90 U/l in tubercular pleural, peritoneal and pericardial
effusion respectively with overall 100% sensitivity and
94.6%, specificity and cut-off value of 40 U/l.
Cherian12 in his review article has suggested the
following criteria for diagnosis of tuberculous etiology
of pericardial effusion:







Culture of Mycobacterium
pericardial fluid or tissue
tuberculosis
from
Pericardial tuberculous granuloma with acid-fast
bacilli
Pericardial tuberculous granuloma + positive
tuberculin skin test
Tuberculous granuloma in scalene node or
peripheral lymph node or pleura with positive
tuberculin test.
Active tuberculosis elsewhere in the body
Mediastinal lymph nodes (matted with central
necrosis) on CT chest/abdomen
Response to specific antitubular therapy.
Mishra et al13 in their study of 31 tuberculous (20 pleural,
8 ascites and 3 pericardial) and 24 nontuberculous
Diagnosis
Treatment
Antituberculous treatment with primary four drugs
(rifampicin, isoniazid, pyrizinamide and ethambutol)
is given for first two months. Later pyrizinamide is
omitted and the remaining three drugs are given. Total
9-12 months of treatment is recommended with periodic
follow-up of symptoms and signs of pericardial effusion,
ESR, X-ray chest, ECG and 2D-echo till the complete
disappearance of fluid and other clinical features.2
Strang et al14 in their 10-year follow-up concluded that
in the absence of specific contraindication, corticosteroid
should be prescribed in addition to antituberculous
chemotherapy in the treatment of tuberculous pericardial
effusion. The ability of prednisolone to prevent
complication of constrictive pericarditis and thus
reduce the need for pericardiectomy is important.
Standard treatment for pericardial effusion is
Indian Journal of Clinical Practice, Vol. 22, No. 8, January 2012
373
clinical study
pericardiocentesis which is safe, especially under echo
guidance. However, open drainage on admission, in their
experience, proved to be more effective and it virtually
eliminated the need for repeat pericardiocentesis.
Strang,15
According to Wragg and
tuberculous
pericarditis is increasing in sub-Saharan Africa where
tuberculosis is the most common opportunistic infection
complicating HIV infection. Adjunctive prednisolone
is beneficial in HIV seronegative patients with TB
pericardial effusion, reducing the risk of recurrent
tamponade, death from pericarditis and constriction.16
Benefit from similar treatment in HIV seropositive
patient is encouraging, particularly where the absence of
antiretroviral drugs worsens the prognosis. Adjunctive
prednisolone in Hakim’s study17 resulted in significant
improvement in clinical condition and faster resolution
of physical signs of pericardial effusion.
Trautner and Darouiche18 reported their experience
of 10 cases of tuberculous pericarditis. They addressed
four specific topics: (1) The importance of tissue for
diagnosis (2) The optimal surgical management; (3)
The role of corticosteroids and (4) The impact of HIV
infection on the management of the disease. They
concluded from their experience of 10 cases and review
of literature that the optimal management should
include an open pericardial window with biopsy
both for the diagnosis and to prevent reaccumulation
of fluid. Corticosteroids offer benefit in preventing
fluid reaccumulation and preventing development of
constrictive pericarditis. Patients with HIV seropositive
also had an excellent response to open drainage,
corticosteroids and antituberculous therapy. They stress
that pericardial tissue specimens should be obtained to
provide the best chance of definitive diagnosis.
Mayosi et al19 searched MEDLINE (Jan 1966 to
May 2005) and the Cochrane Library to review and
summarize the literature on the pathogenesis, diagnosis
and management of tuberculous pericarditis. A
‘definite’ diagnosis of tuberculous pericarditis is based
on the demonstration of tubercle bacilli in pericardial
fluid or on histological section of the pericardium; a
‘probable’ diagnosis of tuberculous pericarditis is based
on the proof of tuberculosis elsewhere in a patient with
pericardial effusion, raised ADA levels with lymphocytic
pericardial fluid exudate and/or response to anti-Koch’s
treatment with four primary drugs.
According to their review, role of adjunctive
corticosteroid in preventing progression to constrictive
pericarditis or reducing mortality is still not clear. The
timing of pericardiectomy is not clear but majority
workers recommend a trial of medical therapy and
surgery in those who do not respond after six months
medical therapy.
Reuter et al20 in their study of 57 patients of tuberculous
pericarditis with adjunctive corticosteroids concluded
that intrapericardial and oral corticosteroids were welltolerated but did not improve the clinical outcome.
In their experience, standard 4-drug treatment for
six months was effective even in HIV-positive patients.
They could not demonstrate, in their 3-year study, the
significant benefits from adjunctive corticosteroids in
tuberculous pericarditis.
Conclusion
Tuberculosis is an important etiological cause of
pericardial effusion in Afro-Asian countries. Now with
spread of HIV infection, the incidence of tuberculous
pericardial effusion has increased not only in AfroAsian countries but also all over the world. Improved
techniques for recovery of M. tuberculosis, the use of
PCR technology, ADA levels, pericardial IFN-g, absolute
lymphocyte count, detection of mediastinal lymphnodes
on CT and more clearly defined observations on
2D-echo have improved the percentage of proper
diagnosis. For management, apart from antibuberculous
regime, pericardial parocentosis or open drainage
and corticosteroids (oral and/or intrapericardial) have
improved the morbidity, mortality and reduced the
sequalae of constrictive pericarditis.
References
1. Fowler
NO.
Tuberculous
1991;266(1):99-103.
pericardiatis.
JAMA
2. Sainani GS. Pericarditis with effusion. API Text Book of
Medicine. Sainani GS, (Ed.), published by Association
of Physicians of India Mumbai 2001:430-4.
3. Hayashi H, Kawamata K, Machida M, Kumuzaki T.
Tuberculous pericarditis: MRI features with contrast
enhancement. Br J Radiol 1998;71(846):680-2.
4. Desai HN. Tuberculous pericarditis. A review of 100 cases.
S Afr Med J 1979;55(22):877-80.
5. Stark DD, Higgins CB, Lanzer P, Lipton MJ, Schiller N,
Crooks LE, et al. Magnetic resonance imaging of the
pericardium: normal and pathological findings. Radiology
1984;150(2):469-74.
6. D’Silva SA, Nalladaru M, Dalvi BV, Kale PA,
Tendolkar AG. MRI as guide to surgical approach in
tuberculous pericardial abscess. Case report. Scand J
Thorac Cardiovasc Surg 1992;26(3):229-31.
7. Clifford CP, Davies GJ, Scott J, Shaunak S, Sarvill J, Schofield
JB. Tuberculous pericarditis with rapid progression to
Cont’d on page 415...
374
Indian Journal of Clinical Practice, Vol. 22, No. 8, January 2012
