Based on a review of the clinical records, the immediate cause of death was judged to be tuberculosis in 29 out of 47 patients (10 pulmonary tuberculosis and 19 extrapulmonary) and was a contributory cause of death in a further 11 patients. The reviewing internist and attending doctor differed over seven patients, because the former did not include tuberculosis as a cause of death, but at autopsy five of the seven were found to have tuberculosis (Table 3); one had interstitial lung disease and one salmonella septicaemia as their cause of death. In 25 patients (53%), the immediate cause of death judged from clinical review was identical to that of the autopsy. There was an overlap between the immediate or contributory causes of death in the remainder, except for one with insufficient premortem information. Premortem investigations confirmed tuberculosis in 23 patients (49%). Thirteen were sputum smear positive, nine were tuberculosis culture positive (six blood, two pleural fluid and one sputum), and one was diagnosed by cytological assessment of a fine needle aspiration biopsy. The antibiotics that patients had received at least one dose of before death were: prophylactic cotrimoxazole (18), high-dose cotrimoxazole (four), amoxicillin/clavulanate (22) and ciprofloxacin (seven).
The record review of five patients suggested Addisons's disease, of whom three had autopsy-confirmed adrenal pathology. Significant pathology found at autopsy but not ascertained by record review was a perforated duodenal ulcer, a perforated gastric ulcer, cerebral tuberculoma, tuberculous meningitis, disseminated Kaposi's sarcoma, fulminant hepatitis, pulmonary thromboembolism, a subdural hemorrhage, and two pelvic abscesses. In addition, seven out of seven cadavers with autopsy-determined cytomegalovirus pneumonitis, four out of five with Pneumocystis pneumonia, one out of two with disseminated cryptococcosis and one with cerebral toxoplasmosis were not diagnosed premortem.
The leading autopsy-determined causes of death were pulmonary tuberculosis, disseminated tuberculosis and bacterial pneumonia (Table 4). Two cadavers had disseminated cryptococcosis. Additional autopsy-determined immediate causes of death in one cadaver each were toxoplasma meningoencephalitis, gastroenteritis, salmonella septicemia, subdural hematoma, perforated duodenal ulcer, severe pulmonary edema on a background of milliary tuberculosis in the lung, and a large empyema from which Streptococcus pneumoniae was cultured postmortem. Autopsy confirmation of tuberculosis was found in 37 cadavers (79%) of whom 34 had disseminated tuberculosis (Table 3). Of 33 with pathological evidence of tuberculosis in their lungs, 17 had milliary tuberculosis of the lungs, 10 had tuberculous bronchopneumonia, five had cavitations and there was one with a Gohn focus. Acid-fast bacilli were not found in the lung sections of nine of these cadavers, whereas four had scanty, 11 moderate, and nine abundant acid-fast bacilli. Co-existing pulmonary pathology in those with autopsy-confirmed pulmonary tuberculosis was bacterial pneumonia (three), Pneumocystis pneumonia (two), cytomegalovirus pneumonitis (two), pulmonary thromboembolic disease (two), interstital lung disease (two) and bronchiolitis oblitererans pneumonia (one) and a massive empyema. Extensive bilateral, multilobe pulmonary pathology was a prominent finding irrespective of whether pulmonary tuberculosis was present or not. Ten cadavers had no autopsy evidence of tuberculosis, and none had premortem bacteriological confirmation of tuberculosis. Immediate and contributory causes of death in these 10 included: bronchopneumonia (six), cytomegalovirus pneumonitis (five), toxoplasmosis (one), disseminated Kaposi's sarcoma (two), cryptococcosis (one), hepatic necrosis (one) and interstitial pneumonitis (one). Six of the 10 had been started on tuberculosis treatment at the time of death. Two of the cadavers with cytomegalovirus pneumonitis had no other pulmonary pathology.
Thirty-one cadavers had pleural effusions at autopsy of which 25 were bilateral. Six of the 31 did not have tuberculosis at autopsy of whom two had Pneumocystis pneumonia (one with co-existent cytomegalovirus pneumonitis), two had cytomegalovirus pneumonitis alone and one each had bronchopneumonia or interstitial lung disease.
A range of kidney diseases was found at autopsy. Fourteen had renal tuberculosis and 27 had other renal pathology: interstitial nephritis (nine), acute or chronic pyelonephritis (seven), acute tubular necrosis (five), cryptococcal infection (two), simple renal cysts (two), HIV-associated nephropathy (one) and dystrophic calcification (one). Thirteen cadavers had adrenal pathology at autopsy, of whom eight had adrenal tuberculosis. Cryptoccocal and cytomegalovirus infection was found in the adrenal glands of one patient each. The remainder had non-specific adrenal cortical atrophy or lipid depletion. Eighteen cadavers had hepatic tuberculosis, 12 had steatosis and six had portal triaditis. Three had severe hepatic necrosis, which was probably caused by tuberculosis therapy. One cadaver had hepatic cryptococcal infection. No hepatic cytomegalovirus infections were found. Only two had increased liver iron. Twenty-five cadavers had autopsy-confirmed tuberculous lymphadenitis and both of those with disseminated cryptococcosis had lymph node involvement. Tuberculosis was found in the spleens of 22 cadavers. Examination of the hearts showed one with myocardial cryptococcal infection, two with mitral valve disease, and seven with subendocardial pallor detected on nitrobluetetrazolium enzyme analysis, suggesting subendocardial ischemia. Twelve cadavers had pericarditis, eight were non-specific, three were adhesive and one was tuberculous. Two brains were not examined microscopically but four of the remaining 45 had evidence of tuberculosis. Nineteen brains had HIV-associated meningoencephalitis, one of whom had a B-cell lymphoma. Cryptococcal infection was found in two brains as part of disseminated disease. Cerebral toxoplasmosis and an infarct in the left basal ganglia was found in a 29-year-old man treated for tuberculosis for 2 months before death in whom pulmonary tuberculosis, but not pulmonary toxoplasmosis, was confirmed at autopsy. One patient died of a subdural hemorrhage. In summary, therefore, lung, lymph nodes, spleen, liver and kidney were the leading sites of tuberculosis found at autopsy.
In our sample of 47 HIV-infected adults with a premortem diagnosis of tuberculosis one fifth did not have tuberculosis at autopsy. Most of those with tuberculosis had disseminated disease. Moreover, other severe co-morbid conditions were frequent but were often clinically unrecognized before death, and almost half had a severe bacterial infection and one fifth had cytomegalovirus implicated in their death.
Studies of autopsy-determined causes of death in HIV/tuberculosis co-infected adults are infrequent, retrospective and give scant detail on other causes of death [23–25]. Similar findings to ours were reported from Abidjan, where 14 HIV-infected adults who died while on, or shortly after, tuberculosis treatment had a variety of ‘primary’ and ‘associated’ causes of death . Two studies reported causes of death in HIV-infected cadavers with autopsy-confirmed tuberculosis [11,24]. First, in a subgroup of HIV-infected adults from Botswana , tuberculosis was the primary cause of death in 38 out of 42, and cerebral hemorrhage and Kaposi's sarcoma the primary cause in the remaining four. Second, a study from Sao Paulo  reported the leading co-morbid conditions in 100 cadavers to be Pneumocystis pneumonia, cerebral toxoplasmosis, wasting syndrome, pyogenic pneumonia and cytomegalovirus. In the pre-ART era, researchers from Milan  reported that 36% of 78 HIV-infected patients were diagnosed clinically but not at autopsy with mycobacterial disease; however, almost half their cases were non-tuberculous mycobacteria. The clinical ‘overdiagnosis’ of tuberculosis may reflect not only differences in study inclusion criteria but also physicians' experience, increasing demand for hospital care by HIV-infected patients and referral, admission and treatment policies. In addition, many of our sample had been on tuberculosis treatment before admission, which may alter the classic clinical or autopsy presentation of tuberculosis. Moreover, our study suggests that in those with advanced HIV disease, physicians' ability to discriminate between pulmonary and extrapulmonary disease premortem is poor, and that in those who die most have extrapulmonary tuberculosis, confirming the findings of previous autopsy studies in HIV-infected adults [10–15]. This has obvious implications for World Health Organization clinical staging of HIV disease, and by extension, timing of the initiation of ART.
The seroprevalence of cytomegalovirus in adults and children in sub-Saharan Africa is high irrespective of HIV status [36,37], and cytomegalovirus DNA was detected in two-thirds of the lungs of our sample of whom seven had histological evidence of cytomegalovirus. This is similar to other autopsy studies in HIV-infected patients from sub-Saharan Africa and to the prevalence reported in cadavers with autopsy-confirmed tuberculosis from Sao Paulo, Brazil , and in the pre-HAART era, 69 and 66% of postmortem samples from Mexico City  and London  had evidence of cytomegalovirus pathology and infection, respectively. Differences in the prevalence of cytomegalovirus-related pathology are probably related to eligibility criteria, methods of determining the presence of cytomegalovirus and the local epidemiology of cytomegalovirus. The clinical relevance of the pathological findings of pulmonary cytomegalovirus is, however, unclear, as it is diagnosed only if there is no other histological explanation. Difficulties in pathological diagnosis are compounded by the absence of both accurate premortem diagnostic tests and inexpensive, effective therapy for cytomegalovirus pneumonitis, with the result that there are a paucity of epidemiological data on this condition, from HIV-infected adults in sub-Saharan Africa. On the other hand, Pneumocystis and tuberculosis are recognized co-pathogens in HIV-infected adults [40–42]. M. avium complex was cultured from the spleens of two patients and is not uncommon in our setting . Low hemoglobin levels were found in most of our patients, in keeping with studies of the predictors of mortality in HIV-infected adults [44,45] and adults with tuberculosis , but the autopsy is a poor tool in assessing the contribution of anemia to mortality.
The earlier diagnosis and treatment of both HIV and tuberculosis may prevent the advanced stage of disease we found in this group. Adults who are HIV infected should receive care that includes active case finding  and preventive treatment for tuberculosis . Cotrimoxazole preventive treatment for other opportunistic infections , especially for those receiving tuberculosis treatment , coupled with ongoing immune monitoring and the prompt initiation of ART may reduce the high tuberculosis-associated mortality seen in sub-Saharan Africa.
The major limitations of this study are its small sample size and bias inherent in an autopsy study, exacerbated by the very low consent rate. There are limited data available from the two hospitals that describes the epidemiology of deaths in patients with a premortem diagnosis of tuberculosis. The median ages of 747 women and 813 men (male to female ratio 1.1: 1) who died at CHBH in 2006 with a premortem diagnosis of tuberculosis (including 32 children less than 15 years of age) was 38 [interquartile range (IQR) 30–45] and 40 (IQR 33–48) years, respectively, (M.E. Edginton, unpublished data) and at TBC the average age of 592 men and 411 women (male to female ratio 1.4: 1) dying in 2001–2003 was 42 and 37 years, respectively (D.E. Douglas, unpublished data). This is similar to our study. The median age of all 50 cadavers was 35 years (IQR 30–40) and the male to female ratio was 1.1: 1. Further generalization of the results of this study are limited because recruitment was not restricted to adults with bacteriologically confirmed tuberculosis and we only recruited patients from two hospitals in Soweto, reflecting disease patterns endemic to this locale. Finally, categorizing causes of death from lists of serious pathologies is subjective and may vary between internists and pathologists.
We recommend that prospective studies be undertaken to define risk factors for death in adult tuberculosis patients, and that the epidemiology and clinical significance of cytomegalovirus pneumonitis in sub-Saharan African countries should be better understood and its premortem diagnosis improved. Furthermore, trials of potent antibiotics to prevent and treat NTS and other severe bacterial infections should be considered. Deaths from tuberculosis will continue in high HIV prevalence settings until a combination of population HIV and tuberculosis preventive measures in conjunction with improved diagnosis and early treatment of tuberculosis and other common life-threatening opportunistic infections are implemented.
The authors would like to thank the families of the deceased for providing consent for autopsy, and Mr Thomas Mdingi for assisting in obtaining consents.
1. Dye C, Maher D, Weil D, Espinal M, Raviglione M. Targets for global tuberculosis control. Int J Tuberc Lung Dis 2006; 10:460–462.
2. Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, Raviglione MC, Dye C. The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Arch Intern Med 2003; 163:1009–1021.
3. Dye C, Watt CJ, Bleed DM, Hosseini SM, Raviglione MC. Evolution of tuberculosis control and prospects for reducing tuberculosis incidence, prevalence, and deaths globally. JAMA 2005; 293:2767–2775.
4. Alvarez GG, Thembela BL, Muller FJ, Clinch J, Singhal N, Cameron DW. Tuberculosis at Edendale Hospital in Pietermaritzburg, Kwazulu Natal, South Africa. Int J Tuberc Lung Dis 2004; 8:1472–1478.
5. Edginton ME, Wong ML, Phofa R, Mahlaba D, Hodkinson HJ. Tuberculosis at Chris Hani Baragwanath Hospital: numbers of patients diagnosed and outcomes of referrals to district clinics. Int J Tuberc Lung Dis 2005; 9:398–402.
6. Harries AD, Hargreaves NJ, Gausi F, Kwanjana JH, Salaniponi FM. High early death rate in tuberculosis patients in Malawi. Int J Tuberc Lung Dis 2001; 5:1000–1005.
7. Harries AD, Nyangulu DS, Kang'ombe C, Ndalama D, Glynn JR, Banda H, et al
. Treatment outcome of an unselected cohort of tuberculosis patients in relation to human immunodeficiency virus serostatus in Zomba Hospital, Malawi. Trans R Soc Trop Med Hyg 1998; 92:343–347.
8. WHO. Global tuberculosis control: surveillance, planning, financing. WHO report, 2006
. Geneva, Switzerland: World Health Organization; 2006.
9. UNAIDS/WHO. Progress on global access to HIV antiretroviral therapy: an update on “3 by 5”
. Geneva, Switzerland: World Health Organization; 2005.
10. Agyei A, Lartey M. Spectrum of opportunistic infections causing death in patients with AIDS in Ghana: correlation of clinical diagnoses with autopsy results
. 11th Conference on Retroviruses and Opportunstic Infections
. San Francisco, CA, 8–11 February 2004 [Abstract 574].
11. Ansari NA, Kombe AH, Kenyon TA, Hone NM, Tappero JW, Nyirenda ST, et al
. Pathology and causes of death in a group of 128 predominantly HIV-positive patients in Botswana, 1997–1998. Int J Tuberc Lung Dis 2002; 6:55–63.
12. Domoua K, N'Dhatz M, Coulibaly G, Traore F, Konan JB, Lucas S, et al
. Autopsy findings in 70 AIDS patients who died in a department of pneumology in Ivory Coast: impact of tuberculosis. Med Trop (Mars) 1995; 55:252–254.
13. Lucas SB, Hounnou A, Peacock C, Beaumel A, Djomand G, N'Gbichi JM, et al
. The mortality and pathology of HIV infection in a west African city. AIDS 1993; 7:1569–1579.
14. Rana F, Hawken MP, Meme HK, Chakaya JM, Githui WA, Odhiambo JA, et al
. Autopsy findings in HIV-1-infected adults in Kenya. J Acquir Immune Defic Syndr Hum Retrovirol 1997; 14:83–85.
15. Rana FS, Hawken MP, Mwachari C, Bhatt SM, Abdullah F, Ng'ang'a LW, et al
. Autopsy study of HIV-1-positive and HIV-1-negative adult medical patients in Nairobi, Kenya. J Acquir Immune Defic Syndr 2000; 24:23–29.
17. World Health Organization. International statistical classification of diseases and related health problems
, 10th revision. Geneva, Switzerland: World Health Organization; 1992.
18. Popoff MY, Bockemuhl J, Gheesling LL. Supplement 2001 (no. 45) to the Kauffmann–White scheme. Res Microbiol 2003; 154:173–174.
19. Larsen HH, Masur H, Kovacs JA, Gill VJ, Silcott VA, Kogulan P, et al
. Development and evaluation of a quantitative, touch-down, real-time PCR assay for diagnosing Pneumocystis carinii
pneumonia. J Clin Microbiol 2002; 40:490–494.
20. Talkington DF, Thacker WL, Keller DW, Jensen JS. Diagnosis of Mycoplasma pneumoniae
infection in autopsy and open-lung biopsy tissues by nested PCR. J Clin Microbiol 1998; 36:1151–1153.
21. Weinberg A, Spiers D, Cai GY, Long CM, Sun R, Tevere V. Evaluation of a commercial PCR kit for diagnosis of cytomegalovirus infection of the central nervous system. J Clin Microbiol 1998; 36:3382–3384.
22. Lingappa JR, Lawrence W, West-Keefe S, Gautom R, Cookson BT. Diagnosis of community-acquired pertussis infection: comparison of both culture and fluorescent-antibody assays with PCR detection using electrophoresis or dot blot hybridization. J Clin Microbiol 2002; 40:2908–2912.
23. Seleye-Fubara D, Etebu EN. Tuberculosis in Rivers state: autopsy and surgical pathology study in the University of Port Harcourt Teaching Hospital (UPTH), Port Harcourt (1990–2002). Niger J Med 2004; 13:388–392.
24. Gutierrez EB, Zanetta DM, Saldiva PH, Capelozzi VL. Autopsy-proven determinants of death in HIV-infected patients treated for pulmonary tuberculosis in Sao Paulo, Brazil. Pathol Res Pract 2002; 198:339–346.
25. Greenberg AE, Lucas S, Tossou O, Coulibaly IM, Coulibaly D, Kassim S, et al
. Autopsy-proven causes of death in HIV-infected patients treated for tuberculosis in Abidjan, Cote d'Ivoire. AIDS 1995; 9:1251–1254.
26. D'Arminio Monforte A, Vago L, Gori A, Antinori S, Franzetti F, Antonacci CM, et al
. Clinical diagnosis of mycobacterial diseases versus autopsy findings in 350 patients with AIDS. Eur J Clin Microbiol Infect Dis 1996; 15:453–458.
27. Anglaret X, Messou E, Ouassa T, Toure S, Dakoury-Dogbo N, Combe P, et al
. Pattern of bacterial diseases in a cohort of HIV-1 infected adults receiving cotrimoxazole prophylaxis in Abidjan, Cote d'Ivoire. AIDS 2003; 17:575–584.
28. Arthur G, Nduba VN, Kariuki SM, Kimari J, Bhatt SM, Gilks CF. Trends in bloodstream infections among human immunodeficiency virus-infected adults admitted to a hospital in Nairobi, Kenya, during the last decade. Clin Infect Dis 2001; 33:248–256.
29. Gilks CF, Brindle RJ, Otieno LS, Simani PM, Newnham RS, Bhatt SM, et al
. Life-threatening bacteraemia in HIV-1 seropositive adults admitted to hospital in Nairobi, Kenya. Lancet 1990; 336:545–549.
30. Ssali FN, Kamya MR, Wabwire-Mangen F, Kasasa S, Joloba M, Williams D, et al
. A prospective study of community-acquired bloodstream infections among febrile adults admitted to Mulago Hospital in Kampala, Uganda. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 19:484–489.
31. Gordon MA, Banda HT, Gondwe M, Gordon SB, Boeree MJ, Walsh AL, et al
. Nontyphoidal salmonella bacteraemia among HIV-infected Malawian adults: high mortality and frequent recrudescence. AIDS 2002; 16:1633–1641.
32. Gordon MA, Walsh AL, Chaponda M, Soko D, Mbvwinji M, Molyneux ME, Gordon SB. Bacteraemia and mortality among adult medical admissions in Malawi – predominance of nontyphi salmonellae and Streptococcus pneumoniae
. J Infect 2001; 42:44–49.
33. Wilson SJ, Wilson ML, Reller LB. Diagnostic utility of postmortem blood cultures. Arch Pathol Lab Med 1993; 117:986–988.
34. Wilson WR, Dolan CT, Washington JA II, Brown AL Jr, Ritts RE Jr. Clinical significance of postmortem cultures. Arch Pathol 1972; 94:244–249.
35. Brindle RJ, Nunn PP, Batchelor BI, Gathua SN, Kimari JN, Newnham RS, Waiyaki PG. Infection and morbidity in patients with tuberculosis in Nairobi, Kenya. AIDS 1993; 7:1469–1474.
36. Stroffolini T, Ngatchu T, Chiaramonte M, Giammanco A, Maggio M, Sarzana A, Taormina S. Prevalence of cytomegalovirus seropositivity in an urban childhood population in Cameroon. New Microbiol 1993; 16:83–85.
37. Ledru E, Diagbouga S, Ledru S, Cauchoix B, Yameogo M, Chami D, et al
. A study of Toxoplasma and Cytomegalovirus serology in tuberculosis and in HIV-infected patients in Burkina Faso. Acta Trop 1995; 59:149–154.
38. Mohar A, Romo J, Salido F, Jessurun J, Ponce de Leon S, Reyes E, et al
. The spectrum of clinical and pathological manifestations of AIDS in a consecutive series of autopsied patients in Mexico. AIDS 1992; 6:467–473.
39. Pillay D, Lipman MC, Lee CA, Johnson MA, Griffiths PD, McLaughlin JE. A clinico-pathological audit of opportunistic viral infections in HIV-infected patients. AIDS 1993; 7:969–974.
40. Malin AS, Gwanzura LK, Klein S, Robertson VJ, Musvaire P, Mason PR. Pneumocystis carinii
pneumonia in Zimbabwe. Lancet 1995; 346:1258–1261.
41. Karstaedt AS, Grannum S. Pneumocystis carinii
pneumonia in patients with AIDS in South Africa. Trans R Soc Trop Med Hyg 2001; 95:40–41.
42. Hargreaves NJ, Kadzakumanja O, Phiri S, Lee CH, Tang X, Salaniponi FM, et al
. Pneumocystis carinii
pneumonia in patients being registered for smear-negative pulmonary tuberculosis in Malawi. Trans R Soc Trop Med Hyg 2001; 95:402–408.
43. Pettipher CA, Karstaedt AS, Hopley M. Prevalence and clinical manifestations of disseminated Mycobacterium avium
complex infection in South Africans with acquired immunodeficiency syndrome. Clin Infect Dis 2001; 33:2068–2071.
44. O'Brien ME, Kupka R, Msamanga GI, Saathoff E, Hunter DJ, Fawzi WW. Anemia is an independent predictor of mortality and immunologic progression of disease among women with HIV in Tanzania. J Acquir Immune Defic Syndr 2005; 40:219–225.
45. Berhane K, Karim R, Cohen MH, Masri-Lavine L, Young M, Anastos K, et al
. Impact of highly active antiretroviral therapy on anemia and relationship between anemia and survival in a large cohort of HIV-infected women: Women's Interagency HIV Study. J Acquir Immune Defic Syndr 2004; 37:1245–1252.
46. Sacks LV, Pendle S. Factors related to in-hospital deaths in patients with tuberculosis. Arch Intern Med 1998; 158:1916–1922.
47. Golub JE, Mohan CI, Comstock GW, Chaisson RE. Active case finding of tuberculosis: historical perspective and future prospects. Int J Tuberc Lung Dis 2005; 9:1183–1203.
48. Wilkinson D, Squire SB, Garner P. Effect of preventive treatment for tuberculosis in adults infected with HIV: systematic review of randomised placebo controlled trials. BMJ 1998; 317:625–629.
49. Anglaret X, Chene G, Attia A, Toure S, Lafont S, Combe P, et al
. Early chemoprophylaxis with trimethoprim-sulphamethoxazole for HIV-1-infected adults in Abidjan, Cote d'Ivoire: a randomised trial. Cotrimo–CI Study Group. Lancet 1999; 353:1463–1468.
50. Wiktor SZ, Sassan-Morokro M, Grant AD, Abouya L, Karon JM, Maurice C, et al
. Efficacy of trimethoprim–sulphamethoxazole prophylaxis to decrease morbidity and mortality in HIV-1-infected patients with tuberculosis in Abidjan, Cote d'Ivoire: a randomised controlled trial. Lancet 1999; 353:1469–1475.