Tuberculous pericarditis

Article written and reviewed by Dr Luca Moretti

The tuberculous pericarditis is an extrapulmonary localization of infection by mycobacteria belonging to the Mycobacterium tuberculosis complex. Pericardial involvement may produce effusion, tamponade, effusive-constrictive physiology, or constriction, with manifestations that depend on spread of infection, immune response, and the degree of tissue organization. Overt pulmonary tuberculosis need not be present, and the absence of typical radiographic lesions does not exclude cardiac involvement.

The disease is particularly important in regions with a high incidence of tuberculosis and in people with HIV infection, but it can also occur in low-prevalence settings. In the latter, migration, exposures, immunosuppression, and a history of tuberculosis modify the diagnostic probability. Recognition can be difficult because the fluid is often paucibacillary and initial symptoms may be nonspecific. Diagnosis requires distinguishing microbiologic evidence, compatible findings, and simple prior exposure to the mycobacterium.

Antimycobacterial treatment is essential for causal control but does not exhaust management. Cardiac compression requires drainage when indicated, whereas persistent fibrosis may make pericardiectomy necessary even after an infectious disease response. The role of adjunctive corticosteroids is more selective than an undifferentiated reading of older recommendations might suggest: benefits for some complications, uncertainty regarding overall outcome, and HIV status must be considered together.

Epidemiology, etiology, and granulomatous pathogenesis

The epidemiologic importance depends on tuberculosis transmission and on the prevalence of conditions that favor progression or reactivation. In endemic areas, tuberculosis is a major cause of large pericardial effusion and constriction; in low-incidence countries it accounts for a smaller proportion of cases. Percentages observed in referral centers do not necessarily describe the general population because drainage and hospitalization select more severe disease. Geographic setting guides probability but cannot replace evaluation of the individual patient.

The pericardium is reached mainly through lymphatic spread from infected thoracic structures or through hematogenous dissemination; extension from contiguous foci is another possibility. Pericardial disease may accompany disseminated tuberculosis or emerge without an obvious pulmonary focus. An abnormal mediastinal lymph node may provide both a clue and an alternative sampling site. Demonstration of tuberculosis in another organ strongly increases plausibility, but attribution of the effusion still requires correlation with the pericardial picture.

The granulomatous response involves macrophages, lymphocytes, and mediators that seek to contain infection while also potentially damaging tissue. Granulomas, caseous necrosis, and inflammatory infiltrates are characteristic findings when present, but their development depends on host immunity. In advanced immunosuppression the histologic pattern may be less organized and the microbial burden different. Thus, the absence of granulomas in a limited specimen does not exclude infection, whereas an isolated granuloma also requires consideration of other causes.

Traditionally, an initial fibrinous phase, an effusive phase, resorption with organization, and possible fibrotic evolution are described. This sequence helps explain remodeling but is neither mandatory nor fully observable in every patient. The disease may be diagnosed during an effusion, after constriction has developed, or in a mixed presentation. Treatment modifies the course, and not all patients pass through every phase or develop permanent damage.

During the effusive phase, altered microvascular permeability and lymphatic drainage promote an exudate that is often lymphocyte-rich and sometimes hemorrhagic. Fibrin creates adhesions and septa; organization involves both pericardial layers and may limit their sliding. Fibroblast activity and collagen deposition convert part of the lesion into a mechanical constraint. Hemorrhagic content and lymphocytic predominance are suggestive in a coherent setting but are not specific compared with malignancy and other inflammatory diseases.

The predisposing factors include HIV, immunosuppressive drugs, transplantation, malnutrition, and other conditions that impair infection control. Previous tuberculosis may indicate a risk of reactivation or reinfection, but it also makes it necessary to consider resistance and prior treatments. A history of exposure to antimycobacterial drugs, adherence, and origin from areas with drug resistance modify treatment planning. It is not enough to know that a patient has previously taken isoniazid or rifampin without reconstructing the regimen and its duration.

Pathophysiology, clinical manifestations, and physical findings

Onset is often subacute, with fever, night sweats, asthenia, anorexia, and weight loss. Chest pain and a friction rub may be present but do not necessarily dominate the picture. Progressive dyspnea may reflect effusion, compression, pleuropulmonary involvement, or reduced cardiac output. In immunocompromised patients, classic signs may be attenuated; an apparently minimally inflammatory course does not exclude clinically important disease. The duration of symptoms should be reconstructed together with any antibiotic or corticosteroid treatment that may have altered the presentation.

The effusion has variable consequences according to the rate of accumulation, volume, and compliance. Slow distension may allow large collections before instability occurs, whereas loculations and adhesions produce regional compression. Echocardiographic size does not equal hemodynamic severity. Tachycardia, jugular venous distention, hypotension, pulsus paradoxus, and signs of hypoperfusion suggest tamponade, but their expression also depends on volume status and pre-existing cardiac pressures.

The effusive-constrictive physiology combines pressure from fluid with impaired filling due to tissue restriction, often involving the visceral layer. After effective drainage, congestion may persist because fluid removal does not eliminate the constraint. A dilated vena cava alone is not sufficient to demonstrate this physiology: coherent functional findings and exclusion of residual collection, right ventricular dysfunction, and valvular disease are required. Observation before and after the procedure helps determine which component dominates the presentation.

In constriction, poor distensibility impairs cardiac expansion and accentuates ventricular competition during respiration. The patient may present with peripheral edema, ascites, hepatomegaly, early satiety, and reduced exercise tolerance, even with preserved ejection fraction. Elevated jugular venous pressure and Kussmaul sign are useful but not exclusive findings. The picture can be confused with liver disease or right-sided heart failure from another cause, particularly when infectious symptoms are no longer prominent.

The systemic physical examination looks for lymph nodes, pulmonary signs, pleural effusions, weight loss, and extrapulmonary sites. History-taking includes contacts, previous diagnoses, antiretroviral therapy, and other medications. Neurologic, osteoarticular, or abdominal symptoms suggest extension that may alter the treatment regimen and duration. It is incorrect to interpret every organ abnormality as a consequence of pericardial congestion when it may represent a separate tuberculous site.

The clinical severity results from the interaction among infection, hemodynamics, and general condition. A patient with tamponade requires urgent correction regardless of microbiologic confirmation; a stable patient with high suspicion still requires a rapid diagnostic pathway, because waiting for weeks without a plan may permit progression. Functional class, nutritional status, renal and hepatic function, and disease extent provide the baseline against which response and prognosis are subsequently judged.

Etiologic diagnosis and interpretation of pericardial fluid tests

The echocardiography defines effusion, fibrin, loculations, and filling physiology but does not demonstrate the mycobacterium. ECG, complete blood count, C-reactive protein, organ function, and thoracic imaging complete the initial assessment. CT may identify lymphadenopathy and disease sites useful for sampling; cardiac magnetic resonance characterizes pericardial activity, myocardium, and functional consequences. Results must be linked to epidemiologic probability and the possibility of obtaining microbiologic proof from the pericardium or other sites.

Level of evidence for tuberculous attribution

• Microbiologically documented diagnosis: identification of the Mycobacterium tuberculosis complex in pericardial fluid or tissue by culture or a reliable molecular test.
• Probable diagnosis: pericarditis compatible with tuberculosis demonstrated elsewhere or with a convincing combination of epidemiology, exudate characteristics, and local biomarkers after evaluation of alternatives.
• Supportive findings: granulomatous or caseating histology, imaging, and response to treatment, interpreted with their limitations.
• Simple evidence of tuberculosis infection: positive tuberculin skin test or interferon-γ release assay, which do not demonstrate active disease or pericardial localization.

When fluid is collected for an appropriate indication, it should be sent for mycobacterial culture and molecular testing in addition to relevant differential studies. Direct microscopy has limited sensitivity in paucibacillary forms. Culture allows identification and susceptibility testing but takes time and may be negative. A positive molecular test provides faster confirmation and may indicate rifampin resistance, without necessarily providing all the information of a complete susceptibility profile.

The sensitivity of Xpert MTB/RIF on pericardial fluid is not sufficient to exclude disease on its own. In the prospective study by Pandie and colleagues, conducted in a high-prevalence setting, sensitivity was 63.8% and specificity 100% against the microbiologic or histologic reference standard used. These values describe that population and method, not every laboratory. Xpert Ultra increases diagnostic opportunities but retains limitations; contemporary comparison with unstimulated interferon-γ confirms that a negative result still requires clinical interpretation.

The adenosine deaminase, commonly referred to as ADA, is an indicator of the local cellular immune response. A high value in a lymphocytic exudate supports suspicion, especially where pre-test probability is high. A threshold around 40 U/L is frequently used, but method, laboratory, and population must be considered. ADA may be elevated in other conditions, including pyogenic infections and some hematologic malignancies; it does not identify the mycobacterium and does not replace culture. A low value should not negate strong suspicion in an atypical or immunocompromised presentation.

The unstimulated interferon-γ in pericardial fluid has shown high accuracy in studies conducted in endemic areas and may be more informative than ADA in appropriate settings. It is not the same test as blood interferon-γ release assays: those measure a response to antigen stimulation and document immunologically recognized infection. Availability, assay platform, and fluid thresholds are not uniform. The result must be integrated with microbiology, cytology, and the clinical picture, especially when local prevalence is low.

The pericardial biopsy is considered when the result may change management or during an already indicated procedure. Separate specimens should allow histology, culture, and molecular analysis according to laboratory protocol. Necrotizing granulomas are strongly suggestive in a coherent context but do not eliminate every alternative diagnosis; small samples may miss focal lesions. Empirical response to drugs contributes to assessment but, by itself, is not proof of causation because symptoms and inflammation may improve for different reasons.

Disease extent, drug resistance, and pretreatment assessment

Investigation for associated sites increases diagnostic yield and identifies conditions that modify treatment. Respiratory testing on appropriate material, thoracic imaging, and sampling of lymph nodes or other lesions may provide confirmation even when pericardial fluid is negative. The choice of site should balance accessibility, risk, and the likelihood of a useful result. Pulmonary disease also has implications for respiratory precautions and contact management, whereas isolated pericardial involvement alone does not define airborne contagiousness.

The HIV test is relevant because it modifies etiologic probability, risk of dissemination, and therapeutic management. If positive, immune status, antiretroviral therapy, and opportunistic infections must be assessed. Initiation or optimization of antiretroviral therapy should be coordinated with tuberculosis treatment, considering rifampin interactions and the risk of immune reconstitution inflammatory syndrome. It is inappropriate to apply an identical timetable to every patient without considering other sites, particularly possible meningeal involvement.

The assessment for drug resistance includes rifampin and other drugs according to available tests and history. Previous incomplete treatment, contact with drug-resistant tuberculosis, or origin from settings with high resistance heighten concern, but susceptibility should not simply be presumed on the basis of nationality. A molecular resistance signal requires the confirmatory and management pathway defined by the tuberculosis program. A single drug should not be added to a failing regimen because this may select further resistance.

Before treatment, hepatic and renal function, complete blood count, weight, and conditions that affect dosing and tolerability are documented. For ethambutol, visual assessment and the ability to report changes in acuity or color perception are important. Isoniazid requires attention to neuropathy risk and the indication for pyridoxine; rifampin interacts with numerous drugs, including anticoagulants and antiretrovirals. Malnutrition, pregnancy, organ failure, and polypharmacy require specialist adjustments rather than empirical modification of the standard regimen.

The assessment of the pericardial substrate distinguishes treatable activity, compression, and fibrosis. Cardiac magnetic resonance may show edema and enhancement, but persistent late signal alone does not demonstrate microbiologic failure. CT defines calcifications and anatomy; echocardiography remains central for physiology. In uncertain cases of constriction, simultaneous catheterization may clarify respiratory interdependence. None of these tests replaces etiologic proof, but together they guide the need for drainage, therapeutic observation, or surgery.

The care plan must make months of treatment feasible. Access to medication, understanding of the regimen, adherence support, and toxicity surveillance influence outcomes as much as the initial prescription. Coordination with tuberculosis services permits contact management when necessary and continuity of follow-up. Response indicators should be defined from the outset: general condition, weight, fever, symptoms, effusion size, physiology, and available microbiologic data.

Antimycobacterial therapy and treatment of hemodynamic consequences

For disease caused by a susceptible strain, the standard regimen includes an intensive phase of two months with isoniazid, rifampin, pyrazinamide, and ethambutol, usually followed by four months of isoniazid and rifampin. A total duration of six months remains the standard pathway for pericardial disease uncomplicated by problems requiring adaptation. Shortened regimens studied for specific pulmonary forms should not be transferred automatically to pericardial tuberculosis, for which equivalent validation is lacking.

The daily doses should be calculated by weight and formulated according to tuberculosis program recommendations, with adjustments for organ function. In adults, isoniazid around 5 mg/kg, generally up to 300 mg, and rifampin around 10 mg/kg, generally up to 600 mg, are usual reference doses; pyrazinamide and ethambutol are likewise dosed by weight. Fixed-dose combinations may simplify the regimen when appropriate. These references do not replace assessment of pregnancy, pediatric age, marked obesity, or renal or hepatic impairment.

The empirical therapy may be necessary when probability is high and waiting entails risk, after obtaining possible specimens without dangerous delay. The choice should be accompanied by reassessment of the hypothesis and subsequent results. Lack of response requires consideration of an alternative diagnosis, adherence, absorption, resistance, anatomic complications, and paradoxical reaction. It is incorrect to continue an ineffective regimen indefinitely without re-examining these elements. Treatment of resistant forms should be defined in an expert center using an appropriate combination of active drugs.

The tolerability is monitored throughout treatment. Isoniazid, rifampin, and pyrazinamide can contribute to hepatotoxicity; ethambutol requires attention to visual function; neuropathy and drug interactions may compromise adherence and safety. Jaundice, persistent nausea, visual disturbances, or paresthesias require prompt assessment, without random drug discontinuations or substitutions that leave an incomplete regimen. Laboratory monitoring is tailored to baseline conditions, drugs, symptoms, and comorbidities. Pyridoxine is particularly important in people at risk of isoniazid-induced neuropathy.

The drainage of tamponade should not await the effect of antimycobacterial drugs. Image-guided pericardiocentesis or a surgical approach is selected according to anatomy and clinical condition. The material obtained should be used for diagnosis and susceptibility testing. Loculated collections, reaccumulation, or biopsy requirements may favor surgery. A stable effusion without compression is instead managed according to symptoms, size, diagnostic needs, and response, avoiding invasive procedures performed solely because the cause is tuberculosis.

The constrictive congestion may benefit from cautious diuretics as supportive treatment, but excessive preload reduction can worsen cardiac output. When a recent inflammatory component predominates, causal treatment may allow recovery and justify a period of observation in stable patients. Persistent major symptoms, organ damage, and nonreversible constrictive physiology require evaluation for pericardiectomy. Surgery corrects the mechanical constraint, whereas antimycobacterial treatment remains necessary for infection; the two interventions are not alternatives.

The HIV coinfection requires integration of therapies and management of interactions, especially with rifamycins. Immune reconstitution may make an inflammatory process more apparent and cause paradoxical worsening, but this diagnosis requires exclusion of resistance, poor adherence, additional infections, and tamponade. Worsening does not automatically mandate interruption of effective drugs. The choice of any additional treatment depends on severity and should be shared with those managing HIV infection and tuberculosis.

Adjunctive corticosteroids and limitations of the available evidence

The corticosteroids have been studied to reduce the inflammatory response and the risk of constrictive organization, always in addition to an adequate antimycobacterial regimen. The biologic rationale is plausible, but clinical benefit is not equivalent to simple reduction in fever or inflammatory markers. Mortality, tamponade, hospitalization, and constriction must be distinguished because a therapy may influence some outcomes without improving the overall composite outcome. Infectious and oncologic risks may also vary with immune status.

The IMPI trial, conducted in 1,400 patients with definite or probable tuberculous pericarditis, did not show a significant benefit of prednisolone on the primary composite outcome of death, tamponade requiring drainage, or constriction. Reductions in constriction and hospitalization were observed, together with an increase in malignancies, particularly HIV-associated cancers. The population included a high proportion of people with HIV, an essential feature for interpreting the results and applying them to practice. Immunotherapy with Mycobacterium indicus pranii has not become routine treatment for the disease.

The recommendations are not uniform. The World Health Organization allows adjunctive steroid use in tuberculous pericarditis with a conditional recommendation; the 2016 ATS/CDC/IDSA guidelines suggest that they not be used routinely, leaving room for selected patients at high risk of inflammatory complications. The 2025 ESC guidelines consider their use reasonable in people without HIV and advise avoiding them in those with HIV based on the observed balance of benefits and harms. Presenting a single mandatory indication for everyone would erase these real differences.

The clinical selection takes into account substantial inflammatory activity, significant effusion, early signs of constriction, and individual risk, without turning these elements into a universally validated score. Before prescribing steroids, one should be reasonably certain that the antimycobacterial regimen is adequate and that there are no other uncontrolled infections. In patients with HIV, use for a different indication, such as severe immune reconstitution inflammatory syndrome, requires a separate risk-benefit assessment and does not contradict avoiding automatic prescription for every pericardial localization.

If they are used, dose and tapering should follow a specialist protocol consistent with evidence, weight, comorbidities, and interactions. Rifampin may alter corticosteroid exposure, making it inappropriate to transfer without adaptation regimens used for idiopathic pericarditis. A regimen studied in a trial is not a universal prescription for every patient. Glucose, blood pressure, infectious risk, and other consequences of therapy are monitored, while efficacy is judged by clinical status and physiology as well as biomarkers.

The prevention of constriction cannot be entrusted to steroids alone. Timely diagnosis, antimycobacterial treatment, management of fluid, and recognition of persistent constraint remain fundamental. A fall in C-reactive protein during corticosteroid therapy does not prove sterilization, and established fibrotic constriction is not expected to regress by indefinitely increasing immunosuppression. Reassessment must therefore keep infection control, inflammatory activity, and structural consequences separate, assigning each treatment a realistic goal.

Prognosis, complications, and surveillance after treatment

The prognosis is more serious than in uncomplicated idiopathic pericarditis and depends on early diagnosis, HIV status, dissemination, drug resistance, and hemodynamic compromise. Mortality estimates vary widely among settings and cannot be transferred without considering availability of care and population characteristics. Response to treatment may be good even with large effusions, but infectious cure does not guarantee full recovery of pericardial distensibility. Outcome should be assessed across several dimensions rather than only by completion of the drug course.

The tamponade may occur at presentation or during reaccumulation. After drainage, persistent dyspnea and congestion require assessment of residual fluid, an effusive-constrictive component, and concomitant cardiac dysfunction. An effusion that enlarges during treatment may represent a paradoxical reaction, but this explanation must not precede exclusion of dangerous conditions. An initially favorable response followed by worsening requires reassessment of adherence, susceptibility, and associated disease sites.

The constriction may emerge gradually during disease organization. Follow-up should therefore include signs of congestion and functional capacity, with Doppler echocardiography when indicated. Measurement of effusion thickness alone is insufficient: fluid may decrease while tissue constraint becomes more evident. Imaging signs of inflammation may support a reversible component, but a decision to wait must be compatible with stability and absence of progressive organ injury. Refractory congestion requires timely discussion with an experienced surgical center.

The paradoxical reactions may cause increased inflammation or new manifestations during microbiologically effective treatment, including in the setting of immune reconstitution. No single test always distinguishes them from treatment failure. Judgment is based on adherence, microbiologic results, susceptibility, evolution of other sites, and alternative diagnoses. Treatment of the complication is selected according to severity; drainage remains necessary when the collection impairs filling, regardless of the mechanism of deterioration.

The drug toxicity may interrupt an otherwise effective course and promote incomplete regimens if not managed correctly. Visual disturbances, hepatotoxicity, neuropathy, and interactions require a plan for substitutions and reintroduction of drugs while maintaining an adequate number of active agents. Weight and nutritional status should be followed because they influence dosing and recovery. Communication among specialists prevents a drug from being stopped by one service without the others being aware of the regimen change.

The final surveillance documents clinical response, evolution of the collection, physiology, and actual completion of the intended doses. The frequency of subsequent follow-up depends on residual damage and complication risk, not on an identical calendar for everyone. Recurrence of fever, weight loss, or congestion requires assessment for relapse, new infection, or a mechanical sequela. Management of respiratory exposures and contacts follows the presence of associated contagious disease, while cardiologic recovery continues until function and independence are stabilized.

References
1. Schulz-Menger J et al. 2025 ESC Guidelines for the management of myocarditis and pericarditis. European Heart Journal. 2025;46(40):3952-4041.
2. Klein AL et al. Pericardial Diseases: International Position Statement on New Concepts and Advances in Multimodality Cardiac Imaging. JACC: Cardiovascular Imaging. 2024;17(8):937-988.
3. Adler Y et al. 2015 ESC Guidelines for the diagnosis and management of pericardial diseases: The Task Force for the Diagnosis and Management of Pericardial Diseases of the European Society of Cardiology (ESC). Endorsed by: The European Association for Cardio-Thoracic Surgery (EACTS). European Heart Journal. 2015;36(42):2921-2964.
4. Mayosi BM et al. Tuberculous pericarditis. Circulation. 2005;112(23):3608-3616.
5. Mayosi BM et al. Prednisolone and Mycobacterium indicus pranii in tuberculous pericarditis. New England Journal of Medicine. 2014;371(12):1121-1130.
6. Pandie S et al. Diagnostic accuracy of quantitative PCR (Xpert MTB/RIF) for tuberculous pericarditis compared to adenosine deaminase and unstimulated interferon-γ in a high burden setting: a prospective study. BMC Medicine. 2014;12:101.
7. Randall P et al. GeneXpert MTB/RIF Ultra vs Unstimulated Interferon γ (IRISA-TB) for the Diagnosis of Tuberculous Pericarditis in a Tuberculosis-Endemic Setting. Open Forum Infectious Diseases. 2024;11(3):ofae021.
8. Nahid P et al. Official American Thoracic Society/Centers for Disease Control and Prevention/Infectious Diseases Society of America Clinical Practice Guidelines: Treatment of Drug-Susceptible Tuberculosis. Clinical Infectious Diseases. 2016;63(7):e147-e195.
9. World Health Organization. The use of adjuvant steroids in the treatment of TB meningitis and pericarditis. TB Knowledge Sharing Platform. 2025. Consolidated recommendations, Module 4. Accessed September 26, 2026. https://tbksp.who.int/en/node/2965.
10. World Health Organization. Treatment of extrapulmonary TB. TB Knowledge Sharing Platform. 2025. Operational handbook, Module 4. Accessed September 26, 2026. https://tbksp.who.int/en/node/3024.
11. Welch TD et al. Echocardiographic diagnosis of constrictive pericarditis: Mayo Clinic criteria. Circulation: Cardiovascular Imaging. 2014;7(3):526-534.
12. Imazio M et al. Risk of constrictive pericarditis after acute pericarditis. Circulation. 2011;124(11):1270-1275.
13. Sagristà-Sauleda J et al. Effusive-constrictive pericarditis. New England Journal of Medicine. 2004;350(5):469-475.
14. Spodick DH. Acute cardiac tamponade. New England Journal of Medicine. 2003;349(7):684-690.