Tuberculosis and the Heart

Arthur K. Mutyaba, MBChB, MMed, FCP(SA), Mpiko Ntsekhe, MD, PhD*

KEYWORDS Tuberculosis Tuberculous pericarditis Tuberculous myopericarditis Tuberculous myocarditis Tuberculous aortitis

KEY POINTS Tuberculosis remains an important health problem of the developing world. Tuberculous heart disease is an important extrapulmonary manifestation of the disease. Tuberculous pericarditis occurs with higher frequency in individuals infected with the human immunodeficiency virus (HIV) and is characterized by differences in immune and clinical responses compared with HIV-negative individuals. A biomarker-based approach that uses pericardial fluid unstimulated interferon-g, adenosine deaminase, and polymerase chain reaction offers a fast and reliable way of establishing the diagnosis. Four-drug antituberculous therapy is the mainstay of management for tuberculous pericarditis. Adjunctive steroid therapy has been shown to reduce hospitalization and progression to constrictive pericarditis, but not mortality; in HIV-positive patients, adjunctive steroids are associated with a higher incidence of malignancy.

Video content accompanies this article at http://www.cardiology.theclinics.com.

INTRODUCTION (Fig. 1). Autopsy studies performed in the pre- HIV/AIDS era suggest that the heart is involved in Recognized since antiquity, tuberculosis (TB) still approximately 2% of cases of patients who died contributes significantly to the global burden of from TB.3 Similar studies of patients who died of disease, with an estimated 9.6 million new cases 1 TB and were coinfected with HIV demonstrate of the disease worldwide in 2014. This is espe- multisystemic dissemination in up to 80% of pa- cially true in the developing world where human tients.4,5 Of all the extrapulmonary manifestations immunodeficiency virus (HIV) infection and AIDS, of TB, involvement of the heart is second only to socioeconomic deprivation, and poor health sys- central nervous system TB in terms of its devas- tems infrastructure interact to make TB a signifi- 2 tating morbidity and mortality. cant public health problem. Although it remains It has been almost a decade since there was a primarily a disease of the lungs, the classic lesion comprehensive review of the main form of tuber- Mycobacterium tuberculosis of TB—the acid fast culous heart disease.6 That review focused on (Mtb) bacilli in a necrotic core bound by aggre- the pathogenesis, diagnosis, and management of gates of various inflammatory cells (granuloma)— tuberculous pericarditis and concluded by noting can be found in virtually any part of the body that there remained major gaps in our

The authors have nothing to disclose. Division of Cardiology, Department of Medicine, Groote Schuur Hospital, University of Cape Town, E17 Cardiac Clinic, New Groote Schuur Hospital, Anzio Road, Observatory 7925, Cape Town, South Africa * Corresponding author. E-mail address: [email protected]

Cardiol Clin 35 (2017) 135–144 http://dx.doi.org/10.1016/j.ccl.2016.08.007

0733-8651/17/Ó 2016 Elsevier Inc. All rights reserved. cardiology.theclinics.com Downloaded from ClinicalKey.com at SA Consortium - University of Cape Town February 17, 2017. For personal use only. No other uses without permission. Copyright ©2017. Elsevier Inc. All rights reserved. 136 Mutyaba & Ntsekhe

Fig. 1. Mycobacterium tuberculosis. Tuberculous granuloma (A) showing an aggregation of lymphocytes, mono- cytes, multinucleate giant cells with a central area of caseous necrosis. Ziehl-Neelsen stain (B) demonstrating acid-fast M tuberculosis (arrow). (Courtesy of Dr Craig Jamieson, formerly of Anatomical Pathology Department, University of Cape Town, Observatory, Cape Town, South Africa.)

understanding of the subject. The identified gaps The IMPI immunotherapy trial was a double- included the need for improved diagnostic tools, blind, randomised control trial of 1400 patients a better understanding of the impact of HIV, and with definite or probable TB pericarditis who determination of the effectiveness of adjuvant were randomised to receive adjunctive corticoste- therapy on clinical outcomes.6 Since then, much roids or Mycobacterium indicus pranii versus pla- data, predominantly from sub-Saharan Africa, cebo in a 2 Â 2 factorial design.19 These studies, have been generated to address some of these which were conducted between 2004 and 2015, gaps. An updated comprehensive overview of TB form the backbone of new insights gained since and the heart with a summary of the new insights 2005. is therefore timely and hopefully of value to the There are 3 predominant clinical manifestations general physician and cardiologist on the front of tuberculous heart disease. In descending order lines of patient care. of frequency, these include TB pericarditis, The search strategy for this review involved a myocardial TB with or without aneurysm forma- comprehensive search of MEDLINE, EMBASE tion, and TB aortitis with or without mycotic aneu- and the Cochrane library of systematic reviews rysms and pseudoaneurysms involving the aortic with the MeSH terms: “tuberculosis and the heart,” valve and/or sinuses of Valsalva. Important clin- “cardiac tuberculosis,” “cardiovascular tubercu- ical, diagnostic, and management aspects of losis,” “myopericarditis and tuberculosis,” “tuber- each are reviewed. culous pericarditis,” “tuberculous aortitis,” and “HIV and the heart’’ from January 2005 to TUBERCULOUS PERICARDITIS December 2015. The reference lists of selected articles were searched for articles deemed to be of Tubercle bacilli access the pericardium via 3 main relevance to the subject. Appropriate English lan- mechanisms. These include retrograde lymphatic guage studies were retrieved and reviewed. spread from mediastinal, paratracheal and peri- There have been a number of important ad- bronchial lymph nodes,20 hematogenous spread vances in our understanding of tuberculous peri- (dominant in immunocompromised hosts),21 and, carditis over the last decade. Most of the new rarely, direct contiguous spread from adjacent information has been generated from sub- structures such as the lungs, pleura, and spine.20 Saharan Africa and Asia, where TB is endemic, In the presence of a competent immune system, HIV is epidemic, and the majority of patients with tuberculous pericardial disease is usually localized TB are also coinfected with HIV.2 The Investigation to the pericardial space. It is typically a paucibacil- of the Management of Pericarditis in Africa (IMPI) lary condition; tubercle proteins trigger a vigorous registry was a prospective observational cohort cell-mediated hypersensitivity response with of consecutive patients with suspected TB peri- T-helper cell (subtype 1) predominant cytokine carditis across multiple sites in sub-Saharan Af- release, leading to an inflammatory exudative effu- rica. Important questions related to the sion and its hemodynamic sequelae.6 In patients immunopathogenesis, clinical manifestations, with dysfunctional immunity as occurs in HIV/ diagnosis, and outcomes of TB pericarditis in the AIDS, there is evidence that mycobacterial replica- HIV era were investigated in the registry.7–18 tion is active, bacillary loads are high, and the

Downloaded from ClinicalKey.com at SA Consortium - University of Cape Town February 17, 2017. For personal use only. No other uses without permission. Copyright ©2017. Elsevier Inc. All rights reserved. Tuberculosis and the Heart 137 clinical manifestations of tuberculous pericarditis Effusive pericarditis is the commonest form of are related to the impact of the infectious and viru- pericardial TB in patients with and without HIV.6 lent nature of the Mtb itself in addition to the hemo- In the developing world, 40% to 70% of large peri- dynamic sequelae.13,21 Tuberculous pericarditis cardial effusions are tuberculous in origin.18,23,24 typically presents as 1 of 4 clinical syndromes, This is in contrast with the developed world, where namely, acute pericarditis, effusive pericarditis less than 4% of cases are tuberculous.25 The clin- and its complications, myopericarditis, and ical presentation of this group of patients is deter- constrictive pericarditis. Although it is convenient mined by the rate of fluid accumulation, magnitude to review them as distinct clinical entities, it is of pericardial fluid-induced cardiac compression, important to understand that there is much overlap and the severity of the inflammation, edema, and in the clinical manifestations. loss of visceral pericardial compliance.14,26 The triad of severe pericarditic chest pain, a peri- Where cardiac compression is significant or rapid, cardial friction rub, widespread ST-segment and T cardiac filling and stroke volume can be compro- wave abnormalities and PR segment depression mised significantly. If hemodynamic compensa- typical of acute pericarditis is an uncommon clinical tory mechanisms are adequate, the symptoms presentation of tuberculous pericarditis, account- are usually those of congestive heart failure ing for only 3% to 8% of patients who present without hypotension, and the clinical signs those with tuberculous pericarditis.7,19 The syndrome is of a large pericardial effusion24,27 (Table 2). Where thought to occur soon after inoculation of tubercle pericardial fluid accumulation is relatively quick bacilli into the pericardium, and is characterized and compensatory mechanisms are inadequate, pathologically by polymorphonuclear leukocytosis patients present with evidence of hypotension with abundant bacilli and granuloma formation.3,20 and tachycardia typical of tamponade.26 A subset The diagnosis of a tuberculous etiology depends on of patients may have sizable effusions with little the presence of constitutional symptoms, lack of evidence of cardiac compression, such that the evidence of conventional bacterial infection, and constitutional symptoms of active TB predominate the demonstration of concurrent TB infection else- and the effusion may be an incidental finding where in the body; easily accessible pericardial (Videos 1 and 2). Where the visceral pericardium fluid is typically absent in acute pericarditis. The is rendered noncompliant by inflammation and resolution of symptoms with initiation of empiric postinfectious injury, patients present with a com- antituberculous therapy in TB endemic areas is bination of the compressive hemodynamics of also suggestive.6 A diagnostic index score that tamponade and the physiology of constrictive uses 6 clinical and laboratory variables has been pericarditis-a syndrome aptly termed effusive developed for use in endemic areas.22 These vari- constrictive pericarditis.14,28,29 Invasive hemody- ables and their weighted score are outlined in namic studies suggests that effusive constrictive Table 1. A summed score of 6 or more has a sensi- pericarditis is common in patients with TB tivity of 86% and specificity of 85% for the diagnosis of tuberculous pericarditis.22 Acute tuberculous pericarditis is managed in similar Table 2 fashion to effusive tuberculous pericarditis as dis- Clinical signs observed in 88 patients with cussed elsewhere in this article. effusive tuberculous pericarditis

Physical Sign Prevalence, n (%) Table 1 Sinus tachycardia 68 (77) Tuberculous pericarditis diagnostic index Pulsus paradoxus (>12 mm Hg) 32 (36) Raised central venous pressure 74 (88) Variable Score Palpable apical impulse 53 (60) Night sweats 1 Increased cardiac dullness 83 (94) Weight loss 1 Muffled heart sounds 69 (78) Fever >38 C2Pericardial friction rub 16 (18) Â 9 White cell count <10 10 3 Hepatomegaly 84 (95) Serum globulin > 40 g/L 3 Ascites 64 (73) Summed score of 6 or more is suggestive of tuberculous Peripheral edema 22 (25) etiology for pericardial effusion. Tamponade 3 (3) Data from Reuter H, Burgess L, van Vuuren W, et al. Diagnosing tuberculous pericarditis. QJM Data from Strang JIG. Tuberculous pericarditis in Transkei. 2006;99(12):827–39. Clin Cardiol 1984;7(12):667–70.

pericarditis occurring in up to 50% of cases.14 The genetic material in pericardial fluid. Pandie and implication of a diagnosis of effusive constrictive colleagues18 compared the accuracy of the Xpert pericarditis is not clear, although there are some Mtb/RIF quantitative polymerase chain reaction observational data to suggest that long-term out- assay to pericardial fluid adenosine deaminase comes may be worse.29,30 Chest radiography, an (ADA) and uIFN-g measurement in the diagnosis electrocardiogram, and echocardiography remain of tuberculous pericarditis. Using a cutoff value essential clinical aids in the evaluation and man- of 44 pg/mL, uIFN-g had a sensitivity and speci- agement of patients with suspected effusive ficity of 95.7% and 96.3%, respectively, for a diag- TB pericarditis. The characteristic findings are nosis of TB, making it superior to both Xpert Mtb/ shown in Box 1. Echocardiography is especially RIF (sensitivity, 63.8%; specificity, 100%) and ADA important to confirm the presence and determine at a cutoff value of 35 IU/L (sensitivity, 95.7%; the size of the effusion and its suitability for safe specificity, 84%).18 Despite the superiority of peri- diagnostic or therapeutic pericardiocentesis.31 cardial fluid uIFN-g over both ADA and Xpert Mtb/ The early and accurate diagnosis of a tubercu- RIF as a diagnostic biomarker of tuberculous peri- lous etiology of pericardial effusion has remained carditis, its widespread use in clinical practice is an important obstacle to optimal patient care. hindered by its high cost. Given the low yield Although empiric therapy in high TB prevalence from fluid culture and the proven usefulness of parts of the world has been advocated by some, modern diagnostic biomarkers, the modern diag- it is important to recognize that the price to pay nostic strategy in patients with suspected TB peri- for missing treatable alternative causes is high.32 carditis in TB endemic areas should be to (1) Traditionally, the definitive diagnosis of tubercu- exclude alternative deadly causes of an inflamma- lous pericarditis depends on the demonstration tory exudative effusion (eg, bacterial, malignant, of tubercle bacilli in the pericardial fluid or tissue and uremic pericarditis), (2) use biomarkers of TB by either direct examination or culture.22,27,33 such as pericardial fluid uIFN-g and ADA, (3) This approach yields a positive diagnosis in 57% confirm TB at sites other than the pericardium of patients in historical series and in less than (eg, sputum, lymph nodes, pleural or ascitic fluid), 20% of patients in contemporary series.19,27,34 and (4) have a low threshold to treat for TB where The most recent advance in the diagnosis of TB no obvious alternative is apparent and the clinical pericarditis relates to the usefulness of unstimu- picture fits.18,31,35 lated gamma interferon (uIFN-g, a pericardial fluid Although the safety and efficacy of a 6-month biomarker of TB infection) and polymerase chain course of 4-drug antituberculous chemotherapy reaction–based methods of identifying Mtb in the management of effusive tuberculous pericarditis has long been established, the role of adjunctive therapies in improving survival and pericarditis-related outcomes has been unclear 36–38 Box 1 until recently. The results of the adequately Clinical diagnostic aids and characteristic powered IMPI Immunotherapy trial have provided findings in tuberculous effusive pericarditis the best evidence to date that, although adjunctive steroids decrease the incidence of constrictive Chest radiograph pericarditis and subsequent rehospitalization by Cardiomegaly, pulmonary infiltrates, pleural close to 45%, they do not reduce mortality rates effusion, mediastinal lymphadenopathy compared with placebo.19 In patients who were infected with HIV, steroids were associated with Electrocardiogram an increase in incidence of HIV-related malig- Sinus tachycardia, diffuse ST-segment and T nancies, making firm recommendations about wave changes, QRS microvoltage, QRS alter- their use in this subset of patients difficult.19 Other nans, atrial fibrillation adjunctive therapies whose efficacy has been Echocardiogram tested include Mycobacterium indicus pranii, Pericardial effusion (with or without fibrin which was neutral in the IMPI immunotherapy 19 stranding), pericardial thickening, caval vessel trial ; intrapericardial corticosteroids, which are 39 distension with diminished inspiratory collap- neutral ; intrapericardial thrombolysis, which se,a paradoxic septal motion,a right atrial/ven- has shown promising results in a small case se- tricular diastolic collapse,a exaggerated ries40; and routine pericardial evacuation, which respiratory transvalvular Doppler flow,a expira- is recommended by most authorities.31,41 Given a tory diastolic hepatic vein flow reversal the very close association of tuberculous pericar- 2 a Indicative of cardiac tamponade. ditis and HIV, patients in whom the diagnosis is suspected should undergo HIV testing with the

Downloaded from ClinicalKey.com at SA Consortium - University of Cape Town February 17, 2017. For personal use only. No other uses without permission. Copyright ©2017. Elsevier Inc. All rights reserved. Tuberculosis and the Heart 139 intention of starting combination antiretroviral Table 3 therapy where HIV coinfection is confirmed. Clinical signs observed in 67 patients with Although direct comparisons are practically tuberculous constrictive pericarditis impossible to make, it is likely that the large differ- ence in the use of combination antiretroviral ther- Physical Sign Prevalence, n (%) apy in the IMPI immunotherapy trial versus the Sinus tachycardia 47 (70) IMPI registry (>70% vs <10%)7,19 had a significant Pulsus paradoxus 32 (48) bearing on the large differences in survival in those (>12 mm Hg) who were HIV infected in the 2 studies. The natural history of patients with effusive peri- Increased central 67 (100) venous pressure carditis is variable and the clinical course in the in- dividual patient unpredictable. Short-term Palpable apical impulse 39 (58) complications include tamponade; the main longer Increased cardiac dullness 17 (25) term complication is fibrotic fusion of the 2 layers of Pericardial knock 14 (21) the pericardium and the development of constric- Muffled heart sounds 51 (76) tive pericarditis. In the anti-TB therapy only arm of Sudden inspiratory S2a split 24 (36) the 1400 patient IMPI Immunotherapy trial where Third heart sound 30 (45) close to 60% of participants underwent pericardio- Hepatomegaly 67 (100) centesis, approximately 18% of patients were dead at 12 months, 8% progressed to develop constric- Ascites 60 (89) tive pericarditis, 4% had a reaccumulation of the Peripheral edema 63 (94) pericardial effusion with the development of tam- a S2 – Second heart sound. ponade, and 40% to 60% recovered fully with no Data from Strang JIG. Tuberculous pericarditis in Trans- long-term clinical sequelae.19 Data from the IMPI kei. Clin Cardiol 1984;7(12):667–70. registry suggests that the disease may present more aggressively in patients who are HIV infected as this group of patients tend to have more signifi- The only definitive treatment for tuberculous cant dyspnea (New York Heart Association func- constrictive pericarditis remains surgical pericar- tional classes III and IV), hemodynamic instability, diectomy, even though it is associated with a peri- and more evidence of myocardial involvement operative mortality of between 5% and 14%.15,43 compared with their HIV-uninfected age- and In some instances of tuberculous pericarditis, sex-matched controls.8 Finally, data from the the underlying myocardium can become inflamed IMPI registry suggest that the incidence of constric- and edematous as a complication of the disease tive pericarditis may be significantly lower in process.44 Evidence for this myopericarditis takes HIV-infected participants; an explanation for this the form of elevated biomarkers of myocardial observation remains to be elucidated.10 injury (troponins, creatinine kinase, etc), dynamic The diagnosis of posttuberculous constrictive electrocardiographic changes consistent with pericarditis requires a combination of a high index myocardial injury, and evidence of mild impair- of suspicion, meticulous clinical evaluation for ment of left ventricular systolic function by direct what are often subtle signs (Table 3), and integra- imaging.31,44 Where cardiac MRI is available, gad- tion of data from imaging (echocardiography, olinium enhancement of both the pericardium and computed tomography, MRI; Fig. 2), and, occa- myocardium is characteristic of tuberculous myo- sionally, invasive hemodynamic evaluation.31,42 pericarditis (Fig. 4).31 The sparse number of case Once the diagnosis of constrictive pericarditis reports in the literature suggests that myopericar- has been made, establishing a tuberculous etiol- ditis is a rare complication of tuberculous pericar- ogy is less straightforward (Fig. 3). This is because ditis among HIV-negative, immunocompetent a significant proportion of patients present for the patients.45–47 However, data from the IMPI registry first time with established constriction and have no suggests that, among HIV-infected patients with prior history of TB.15 Whether or not a course of advanced immunosuppression and CD4 counts anti-TB therapy is required in these patients who of less than 100 cells/mL, TB myopericarditis is do not have evidence of active TB is not known. much more common.9 In a cross-sectional anal- Where there are recent symptoms or signs of TB ysis of a subset of 81 patients in the registry, or microscopic evaluation of postoperative peri- 53% had evidence of myopericarditis.9 There are cardial tissue reveals active inflammation, an no data to guide the management of this specific empiric course of anti-TB therapy is reason- complication of effusive pericarditis, but the diag- able.6,38 However, where this evidence of active nosis does not seem to carry a significant adverse TB is absent, there is little evidence of any benefit. effect on survival.9,44

Fig. 2. Tuberculous constrictive pericarditis. A 31-year-old woman with right heart failure. Cardiac MR T1- weighted half Fourier acquisition single shot turbo spin echo image (A) showing large right pleural effusion, bilateral pulmonary interstitial changes, and fibrocavitatory lesions in the left lung apex and atelectasis of the right lower lobe. (B) Short T1 inversion recovery imaging with a normal myocardial signal intensity ratio, but a markedly thickened pericardium measuring 15 mm. (C) Enhancement after administration of gadolinium contrast. (Courtesy of Dr Ntobeko Ntusi, Division of Cardiology, Groote Schuur Hospital/University of Cape Town, Observatory, Cape Town, South Africa.)

TUBERCULOUS MYOCARDITIS mechanisms similar to those giving rise to TB pericarditis and aortitis. For reasons that are not clear, Tuberculous myocarditis (a separate entity from there is an apparent predilection to the right heart, myopericarditis) is a very rare manifestation of car- particularly the right atrium.49 Pathologically, it diovascular TB with an occurrence rate of 0.14% manifests as either nodular tuberculomas with in more than 13,000 autopsies performed over 48 central caseation, miliary tubercles of the heart, 27 years by Rose and colleagues. Myocardial or a diffuse infiltrative pattern associated with peri- spread of TB is presumed to occur via carditis.48 Case reports suggest that tuberculous

Fig. 3. Tuberculous constrictive pericarditis. (A) Postmortem macroscopic appearance showing pericardial thickening and adherence to myocardium. (B) Microscopic appearance of heart in A demonstrating a thickened and fibrotic pericardium with extension of fibrosis into the myocardium. (Courtesy of Dr Craig Jamieson, formerly of Anatomic Pathology Department, University of Cape Town, Observatory, Cape Town, South Africa.)

Fig. 4. A 16-year-old boy with disseminated tuberculosis (TB) with heart failure owing to TB myopericarditis. Car- diac MR steady state free precession images w (A, B) with short axis and horizontal long axis view showing small pericardial effusion; short T1 inversion recovery image (T2-weighted imaging) showing increased myocardial signal intensity ratio, in keeping with myocardial edema (C). Late gadolinium enhancement images showing mid- wall to subepicardial enhancement, in keeping with myocarditis (D–F); the pericardium also enhances. (Courtesy of Dr Ntobeko Ntusi, Division of Cardiology, Groote Schuur Hospital/University of Cape Town, Observatory, Cape Town, South Africa.) myocarditis has a high case fatality rate in part plaques.59,60 Tuberculous aortitis usually presents because the diagnosis is difficult to make ante with mycotic aneurysms of the aorta with both the mortem.50–54 thoracic and abdominal aorta affected equally.59 The clinical presentation of tuberculous Aneurysms may be true or false and within the myocarditis includes atrial and ventricular tachy- ascending aorta, can extend to the aortic root arrhythmias, conduction defects, ventricular and involve the sinus of Valsalva.59,61 Rarely, aneurysms and pseudoaneurysms, dilated car- tuberculous aortitis can present as stenotic lesions diomyopathy with heart failure, and sudden car- of the aorta causing acquired aortic coarctation diac death.53,55–57 The diagnosis requires a high and hypertension.62 index of suspicion, an appropriate imaging mo- The clinical manifestations of tuberculous dality (echocardiogram, computed tomography mycotic aneurysm relate to its mass effect on scan, or MRI) and demonstration of caseous adjacent organs or the complications of rupture. granulomatous inflammation with or without Mtb Common symptoms include chest or back pain, bacilli in myocardial tissue obtained at endomyo- hoarseness, and stridor. Acute aortic regurgitation cardial biopsy.53 To date, there is little to offer pa- and cardiac tamponade have also been reported tients beyond the standard 4 drug anti-TB in patients with aneurysms involving the aortic therapy.55,56 root and sinus of Valsalva.61,63 Owing to its exceeding rarity, a high index of suspicion is TUBERCULOUS AORTITIS required to make the diagnosis of tuberculous aortic mycotic aneurysm in patients from TB Tuberculous infection of the aorta is an exceed- endemic areas. Where available, computed to- ingly rare manifestation of TB with an occurrence mography angiography is a fast and sensitive mo- rate of 0.004% in 22,792 postmortems examina- dality to aid with diagnosis and is useful to tions over 50 years.58 Mtb bacilli can access the delineate the size and nature of aneurysm and its aortic wall via contiguous extension from an adja- relation to surrounding structures. Demonstration cent infective focus (eg, tuberculous mediastinal of contrast extravasation indicates aneurysm lymphadenitis), via the vasa vasorum as part rupture and is an indication for emergency sur- of systemic seeding of bacilli and direct implanta- gery.61 Surgery with in situ reconstruction with tion of bacilli on preexisting atheromatous prosthetic graft or extraanatomic bypass added

to standard 4-drug antituberculous therapy is the immunodeficiency virus infection. Medicine (Balti- current standard of care.59 Without either form of more) 1991;70(6):384–97. treatment, tuberculous aortitis is uniformly fatal.59 6. Mayosi BM, Burgess LJ, Doubell AF. Tuberculous Confirmation of a tuberculous etiology is usually pericarditis. Circulation 2005;112(23):3608–16. made on histology of the resected or biopsied 7. Mayosi BM, Wiysonge CS, Ntsekhe M, et al. Clinical aorta showing the typical granulomatous inflam- characteristics and initial management of patients mation with caseation with or without demonstra- with tuberculous pericarditis in the HIV era: the Inves- tion of Mtb bacilli. tigation of the Management of Pericarditis in Africa (IMPI Africa) registry. BMC Infect Dis 2006;6:2. SUMMARY 8. Mayosi BM, Wiysonge CS, Ntsekhe M, et al. Mortal- ity in patients treated for tuberculous pericarditis in TB remains a significant health problem in the sub-Saharan Africa. S Afr Med J 2008;98(1):36–40. developing world. Cardiovascular TB is a poten- 9. Syed FF, Ntsekhe M, Gumedze F, et al. Myopericar- tially devastating presentation of this ancient infec- ditis in tuberculous pericardial effusion: prevalence, tion. The last 2 decades have witnessed predictors and outcome. Heart 2014;100(2):135–9. tremendous progress in our understanding of the 10. Ntsekhe M, Wiysonge CS, Gumedze F, et al. HIV disease, including our ability to recognize and di- infection is associated with a lower incidence of agnose it, and our capacity to alter its natural his- constriction in presumed tuberculous pericarditis: tory and improve survival through the use various a prospective observational study. PLoS One 2008; therapeutic options. Most patients present with 3(6):e2253. well-recognized, stereotypical clinical syndromes, 11. Shenje J, Ifeoma Adimora-Nweke F, Ross IL, et al. and where there is timely diagnosis and prompt Poor penetration of antibiotics into pericardium in optimal treatment, survival has improved from a pericardial tuberculosis. EBioMedicine 2015;2(11): universally fatal condition before the use of 1640–9. 4-drug anti-TB therapy in the past century to 12. Matthews K, Deffur A, Ntsekhe M, et al. greater than 90% in the IMPI immunotherapy trial. A compartmentalized profibrotic immune response Despite this progress, challenges remain. Chief characterizes pericardial tuberculosis, irrespective among them is making the evidence based inter- of HIV-1 infection. Am J Respir Crit Care Med ventions discussed, which should be the standard 2015;192(12):1518–21. of care (such as echocardiography, pericardio- 13. Pasipanodya JG, Mubanga M, Ntsekhe M, et al. centesis kits, uIFN-g assays, MRI for complex Tuberculous pericarditis is multibacillary and bacte- cases, and surgical pericardiectomy) available rial burden drives high mortality. EBioMedicine and accessible to the poorest and most vulnerable 2015;2(11):1634–9. communities where cardiac and other forms of TB 14. Ntsekhe M, Matthews K, Syed FF, et al. Prevalence, are most prevalent. hemodynamics, and cytokine profile of effusive- constrictive pericarditis in patients with tuberculous SUPPLEMENTARY DATA pericardial effusion. PLoS One 2013;8(10):e77532. Supplementary data related to this article can be 15. Mutyaba AK, Balkaran S, Cloete R, et al. Constrictive found online at http://dx.doi.org/10.1016/j.ccl. pericarditis requiring pericardiectomy at Groote 2016.08.007. Schuur Hospital, Cape Town, South Africa: causes and perioperative outcomes in the HIV era (1990- REFERENCES 2012). J Thorac Cardiovasc Surg 2014;148(6): 3058–65. 1. World Health Organization (WHO). Global tubercu- 16. Matthews K, Ntsekhe M, Syed F, et al. HIV-1 infection losis report 2015. Available at: http://www.who.int/tb/ alters CD41 memory T-cell phenotype at the site of publications/global_report/en/. Accessed March 16, disease in extrapulmonary tuberculosis. Eur J Immu- 2016. nol 2012;42(1):147–57. 2. Dheda K, Barry CE, Maartens G. Tuberculosis. Lan- 17. Matthews K, Wilkinson KA, Kalsdorf B, et al. Pre- cet 2016;387(10024):1211–26. dominance of interleukin-22 over interleukin-17 at 3. Fowler NO. Tuberculous pericarditis. JAMA 1991; the site of disease in human tuberculosis. Tubercu- 266(1):99–103. losis (Edinb) 2011;91(6):587–93. 4. Rana FS, Hawken MP, Mwachari C, et al. Autopsy 18. Pandie S, Peter JG, Kerbelker ZS, et al. Diagnostic study of HIV-1-positive and HIV-1-negative adult accuracy of quantitative PCR (Xpert MTB/RIF) for medical patients in Nairobi, Kenya. J Acquir Immune tuberculous pericarditis compared to adenosine Defic Syndr 2000;24(1):23–9. deaminase and unstimulated interferon-g in a high 5. Shafer RW, Kim DS, Weiss JP, et al. Extrapulmonary burden setting: a prospective study. BMC Med tuberculosis in patients with human 2014;12:101.

19. Mayosi BM, Ntsekhe M, Bosch J, et al. Prednisolone tuberculosis. A twice-weekly, directly observed, and Mycobacterium indicus pranii in tuberculous and cost-effective regimen. Ann Intern Med 1990; pericarditis. N Engl J Med 2014;371(12):1121–30. 112(6):407–15. 20. Spodick DH. Tuberculous pericarditis. AMA Arch 37. Ntsekhe M, Wiysonge C, Volmink JA, et al. Adjuvant Intern Med 1956;98(6):737–49. corticosteroids for tuberculous pericarditis: prom- 21. Ntsekhe M, Mayosi BM. Tuberculous pericarditis ising, but not proven. QJM 2003;96(8):593–9. with and without HIV. Heart Fail Rev 2013;18(3): 38. Mayosi BM, Ntsekhe M, Volmink JA, et al. Interven- 367–73. tions for treating tuberculous pericarditis. Cochrane 22. Reuter H, Burgess L, van Vuuren W, et al. Diag- Database Syst Rev 2002;(4):CD000526. nosing tuberculous pericarditis. QJM 2006;99(12): 39. Reuter H, Burgess LJ, Louw VJ, et al. Experience with 827–39. adjunctive corticosteroids in managing tuberculous 23. Reuter H, Burgess LJ, Louw VJ, et al. The manage- pericarditis. Cardiovasc J S Afr 2006;17(5):233–8. ment of tuberculous pericardial effusion: experience 40. Cui H, Chen X, Cui C, et al. Prevention of pericardial in 233 consecutive patients. Cardiovasc J S Afr constriction by transcatheter intrapericardial fibrino- 2007;18(1):20–5. lysis with urokinase. Chin Med Sci J 2005;20(1):5–10. 24. Strang JIG. Tuberculous pericarditis in Transkei. Clin 41. Seferovic PM, Ristic AD, Maksimovic R, et al. Peri- Cardiol 1984;7(12):667–70. cardial syndromes: an update after the ESC guide- 25. Khandaker MH, Espinosa RE, Nishimura RA, et al. lines 2004. Heart Fail Rev 2013;18(3):255–66. Pericardial disease: diagnosis and management. 42. Klein Allan L, Abbara S, Agler DA, et al. American Mayo Clin Proc 2010;85(6):572–93. Society of Echocardiography clinical recommenda- 26. Shabetai R. The pathophysiology of cardiac tampo- tions for multimodality cardiovascular imaging of pa- nade. Cardiovasc Clin 1976;7(3):67–89. tients with pericardial disease: endorsed by the 27. Strang JI, Kakaza HH, Gibson DG, et al. Controlled Society for Cardiovascular Magnetic Resonance clinical trial of complete open surgical drainage and Society of Cardiovascular Computed Tomogra- and of prednisolone in treatment of tuberculous phy. J Am Soc Echocardiogr 2013;26(9):965–1012. pericardial effusion in Transkei. Lancet 1988; 43. Bashi V, John S, Ravikumar E, et al. Early and late re- 2(8614):759–64. sults of pericardiectomy in 118 cases of constrictive 28. Hancock EW. Subacute effusive-constrictive pericar- pericarditis. Thorax 1988;43(8):637–41. ditis. Circulation 1971;43(2):183–92. 44. Imazio M, Brucato A, Barbieri A, et al. Good prog- 29. Russell JB, Syed FF, Ntsekhe M, et al. Tuberculous nosis for pericarditis with and without myocardial effusive-constrictive pericarditis. Cardiovasc J Afr involvement: results from a multicenter, prospective 2008;19(4):200–1. cohort study. Circulation 2013;128(1):42–9. 30. Ntsekhe M, Shey Wiysonge C, Commerford PJ, et al. 45. Desai N, Desai S, Chaddha U, et al. Tuberculous The prevalence and outcome of effusive constrictive myopericarditis: a rare presentation in an immuno- pericarditis: a systematic review of the literature. competent host. BMJ Case Rep 2013. Cardiovasc J Afr 2012;23(5):281–5. 46. Schrire V. Experience with pericarditis at Groote 31. 2015 ESC Guidelines for the diagnosis and man- Schuur Hospital, Cape Town: an analysis of one hun- agement of pericardial diseases. Available at: http:// dred and sixty cases studied over a six-year period. eurheartj.oxfordjournals.org/content/ehj/36/42/2921. S Afr Med J 1959;33:810–7. full.pdf. Accessed May 2, 2016. 47. Rooney JJ, Crocco JA, Lyons HA. Tuberculous peri- 32. Mayosi BM, Volmink JA, Commerford PJ. Pericardial carditis. Ann Intern Med 1970;72(1):73–81. disease: an evidence-based approach to diagnosis 48. Rose AG. Cardiac tuberculosis. A study of 19 pa- and treatment. In: Yusuf S, Cairns JA, Camm JA, ed- tients. Arch Pathol Lab Med 1987;111(5):422–6. itors. Evidence-based cardiology. 2nd edition. United 49. Momtahen M, Givtaj N, Ojaghi Z, et al. Cardiac tu- Kingdom: Blackwell Science Ltd; 2003. p. 735–48. berculoma of the right atrium. J Card Surg 2011; 33. Strang G, Latouf S, Commerford P, et al. Bedside 26(4):367–9. culture to confirm tuberculous pericarditis. Lancet 50. Hitsumoto T, Ikeda S, Miyazaki S, et al. The case of 1991;338(8782–8783):1600–1. tuberculous myocarditis manifested by extensive 34. Strang JI, Kakaza HH, Gibson DG, et al. Controlled myocardial calcification. J Card Fail 2015;21(10S):170. trial of prednisolone as adjuvant in treatment of 51. Behr G, Palin HC, Temperly JM. Myocardial tubercu- tuberculous constrictive pericarditis in Transkei. losis. Br Med J 1977;1(6066):951. Lancet 1987;2(8573):1418–22. 52. Silingardi E, Rivasi F, Santunione AL, et al. Sudden 35. Syed FF, Mayosi BM. A modern approach to tuber- death from tubercular myocarditis. J Forensic Sci culous pericarditis. Prog Cardiovasc Dis 2007; 2006;51(3):667–9. 50(3):218–36. 53. Agarwal R, Malhotra P, Awasthi A, et al. Tuberculous 36. Cohn DL, Catlin BJ, Peterson KL, et al. A 62-dose, 6- dilated cardiomyopathy: an under-recognized en- month therapy for pulmonary and extrapulmonary tity? BMC Infect Dis 2005;5:29.

54. Njovane X. Intramyocardial tuberculosis–a rare under- medical and surgical experience. Chest 1999; diagnosed entity. S Afr Med J 2009;99(3):152–3. 115(2):522–31. 55. Michira BN, Alkizim FO,Matheka DM. Patterns andclin- 60. Gornik HL, Creager MA. Aortitis. Circulation 2008; ical manifestations of tuberculous myocarditis: a sys- 117(23):3039–51. tematic review of cases. Pan Afr Med J 2015;21:118. 61. Pathirana U, Kularatne S, Karunaratne S, et al. 56. Liu A, Hu Y, Coates A. Sudden cardiac death and Ascending aortic aneurysm caused by Mycobacte- tuberculosis - how much do we know? Tuberculosis rium tuberculosis. BMC Res Notes 2015;8:659. (Edinb) 2012;92(4):307–13. 62. Lin M-M, Cheng H-M. Images in cardiovascular 57. Kinare SG, Bhatia BI. Tuberculous coronary arteritis medicine: tuberculous aortitis. Intern Med 2012; with aneurysm of the ventricular septum. Chest 51(15):1983–5. Available at: http://www.ncbi.nlm. 1971;60(6):613–6. nih.gov/pubmed/22864122. Accessed June 21, 58. Parkhurst GF, Dekcer JP. Bacterial aortitis and 2016. mycotic aneurysm of the aorta; a report of twelve 63. Palaniswamy C, Kumar U, Selvaraj DR, et al. Tuber- cases. Am J Pathol 1955;31(5):821–35. culous mycotic aneurysm of aortic root: an unusual 59. Long R, Guzman R, Greenberg H, et al. Tuberculous cause of cardiac tamponade. Trop Doct 2009; mycotic aneurysm of the aorta: review of published 39(2):112–3.