- Email: [email protected]
Clinical Features of HIV-Seropositive and HIV-Seronegative Patients With Tuberculous Pleural Effusion in Dar es Salaam, Tanzania
Clinical Features of HIV-Seropositive and HIV-Seronegative Patients With Tuberculous Pleural Effusion in Dar es Salaam, Tanzania* Clemens Richter, MD; Roos Perenboom, MD; Isaac Mtoni, MD; james Kitinya , MD; Hassani Chande, MD; Andrew B.M. Swai, MD; Ramadhani R. Kazema , MD; and Linus M. Chuwa , MD In a prospective study, we investigated whether human immunodeficiency virus (HIV) infection alters the clinical presentation in patients with tuberculous pleuritis. One hundred twelve of 118 patients who presented with pleural effusion suffered from tuberculosis (TB); 65 patients (58%) were HIV seropositive. Evidence of disseminated TB was found more often in HIV-positive than in HIV -negative patients (30.8 % vs 10.6%, p<0.02). Dyspnea, fever, night sweat, fatigue, and diarrhea, severe tachypnea, hepatomegaly, splenomegaly, and lymphadenopathy were significantly more common in HIV-infected than in HIV-negative patients with TB. The same applied to a negative Mantoux reaction, lower hemoglobin, higher .Bz-microglobulin values, and in pleural fluid, lower albumin and higher ')'-globulin levels. Among HIV-infected patients, PPD skin test anergy was significantly associated with relative low albumin
Jn 1990, 8 million new cases of tuberculosis (TB)
were r e ported worldwide ; 95% of them occurred in developing countries_! The World Health Organization estimated that by January 1992, 3 million people in sub-Saharan countries were dually infected with the tubercle bacillus and the immunodeficiency virus (HIV). Their cumulative life-time risk of progressing to active TB is estimated to be 30% or more. 2 The proportion of extrapulmonary TB is higher among HIV -infected patients compared with patients without HIV infection. 3 To examine the impact of HIV infection on clinical features , diagnosis, and outcome in patients with extrapulmonary TB manifestations, we performed a prospective study at Muhimbili Medical Centre (MMC, the teaching hospital of the University of Dares Salaam). This report describes the results of a study of the influence of HIV infection on the clinical, laboratory, and radiologic *From the Departments of Medicine (Drs. Richter, Perenboom, and Swai), Histopathology (Drs. Kitinya and Chande), Radiology (Dr. Kazema), Biochemistry (Dr. Chuwa), Muhimbili Medical Centre, and National Institute for Medical Research (Dr. Mtoni), Dar es Salaam, Tanzania. Supported by the Royal Netherlands Tuberculosis Association. Manuscript received Sept. 13, 1993; revision accepted May 5, 1994.
Reprint requests: Dr. Richter, Ruysdaellaan 21, 1213 er Hilversum, the Netherlands.
and )'-globulin levels of pleural fluid. However, the radiographic features did not differ with respect to HIV status; they were predominantly those of primary pleuritis (78% in each group). We conclude that coexisting HIV infection affects clinical and laboratory features, but not the radiographic presentation of patients with (Chest 1994; 106:1471-76) TB pleuritis in Tanzania. AFB=acid-fast bacilli; f:I2·M={:I 2-microglobulin; ELISA= enzyme-linked immunosorbent assay; HIV=human immunodeficiency virus; MMC=Muhimbili Medical Centre; PPD=purificd protein derivative; TB=tuberculosis Key words: clinical, laboratory, and radiologic features; disseminated tuberculosis; HIV infection; Tanzania; Tuberculous pleuritis
features among patients with tuberculous pleuritis , which is a commonly diagnosed form of extrapulmonary TB at our Centre 4 M ETHODS
Patient Selection
All patients with pleural effusion admitted to four medical wards in MMC between January and August 1991 were enrolled. Excluded were patients already receiving antituberculous therapy and those with obvious congestive heart failure as underlying cause of pleural effusion. Investigations
A questionnaire designed to collect data of history and clinical examination with special attention to possible HIV -related symptoms and signs were completed for each patient. The result of a purified protein derivative (PPD) skin test, using 0.1 mL of 2 TU of RT-23 tuberculin (Statens Seruminstitut, Copenhagen, Denmark), which is equivalent to 5 TU of PPD-S were noted. No anergy panel was used. According to Centers for Disease Control and Prevention criteria,5 the skin test was defined positive when the induration at 48 h was ;:::5 mm in HIV-infected patients or ;:::10 mm in HIV-seronegative patients. The radiologic findings of a chest radiograph were recorded. We applied radiologic criteria to classify our patients as ·suffering from primary or reactivation TB (primary: no parenchymal infiltrates at all or the presence of hilar adenopathy with or without noncavitary mid-lower field infiltrates; reactivation: infiltrates or cavities in upper field ). 6•7 After pretest counseling and obtained informed consent for HIV testing, a blood sample was taken for CHEST 1106 I 5 I NOVEMBER, 1994
1471
a complete blood cell count, erythrocyte sedimentation rate (ESR), iJ2-microglobulin ({12-M) levels (Enzygnost iJ2-M micro Behring), and HIV test (Organon ELISA). Samples that tested positive in the enzyme-linked immunosorbent assay (ELISA) test, were confirmed by Western Blot. All patients underwent a pleural fluid aspiration and a blind pleural biopsy (Abraham needle). Pleural fluid analysis consisted of biochemical, bacterial, and cytologic examination. Smears of centrifuged deposit of pleural fluid, pleural tissue, and huffy coat were performed for detection of acid-fast bacilli (AFB) by auramine stain-if positive confirmed by Ziehl-Neelsen stain-in all patients, of urine (in patients with leukocyturia or proteinuria), of lymph nodes (in patients with coexisting lymphadenopathy on physical examination), and whenever clinically indicated of sputum, ascites, and bone marrow. In addition, mycobacterial cultures of all mentioned specimens were carried out on Loewenstein-Jensen medium. Histologic examination was done of the pleural tissue and of lymph nodes.
Criteria for Diagnosis of TB The diagnosis was considered proved when AFB were identified in auramine/Ziehl-Neelsen stains, when mycobacteria were cultured, or when by histologic examination granulomas with caseous necrosis or AFB were seen. Disseminated TB was considered to be present when TB could be proved in at least two different organs. Diagnosis of probable TB was mainly based on a good response of the patient to antituberculous therapy (disappearance of pleural fluid and of TB-related complaints).8
Analysis Data were analyzed using the statistical package for the social sciences. Where appropriate we used the x2 test with Yates' correction or Fisher's Exact Test for 2X2 tables and the Student's t test.
Patients
RESULTS
There were 133 patients with pleural effusion. Fifteen of these patients could not be included in the final analysis of the study because of death (3 patients), hospital discharge (4 patients), or fast improvement with antibiotic therapy (3 patients) before the diagnostic procedures could be performed, or because of failure to get pleural tissue and follow-up data (5 patients). All remaining 118 patients had exudates (pleural protein levels >30 g/ L) . Tuberculosis was diagnosed in 112 patients (94.9%), adenocarcinoma in 1, bacterial infection in 2, and aspecific inflammation in 3 patients. Only three patients with TB had been treated for TB in the past (one HIV positive, two HIV negative); 40% of all patients with TB had a bacille Calmette-Guerin (BCG) vaccination scar. The HIV positivity rate in the patients with TB pleuritis was 58% and three of six patients without TB were HIV seropositive.
Clinical Features The HIV -positive patients more often had complaints at presentation than the patients without HIV infection with significant differences for dyspnea, fever, night sweat, fatigue, and diarrhea (Table 1). The mean duration of complaints before presentation 1472
Table !-Clinical Features of 112 Patients With TB Pleuritis at Enrollment Related to HIV Status
Sex, male Age <40 yr Cough Chest pain Dyspnea Fever Night sweat Fatigue Weight loss (history) Diarrhea Respiratory rate >30/ min Hepatomegaly Splenomegaly Lymphadenopathy*
HIV (+) TB, % n=65
HIV (-) TB, % n=47
p Value
53.8 78.5 92.3 84.6 96.9 78.5 83.1 90.8 89.2 23.1 50.8 41.5 40.0 35.4
55.3 78.7 83.3 83.0 80.9 61.7 63.8 74.5 83.0 4.3 25.5 21.3 14.9 12.8
NSf NS NS NS <0.007 <0.05 <0.02 <0.02 NS <0.007 <0.007 <0.03 <0.004 <0.007
*Found on physical examination. fNS=not significant.
was 1 to 3 weeks longer for the HIV -positive patients than for the HIV -negative patients; for chest pain, cough, and weight loss, the duration was 8 to 9 weeks (range, 1 to 40 weeks) . For the duration of fatigue, this difference was significant (7.8 weeks among HIV -positive patients and 4.9 weeks among HIVnegative patients (p<0.03).
Laboratory Investigations A negative PPD skin test was present in 25 of 53 (47.2%) HIV-positive patients, but in only 5 of 43 (11.6%) HIV -negative patients (p<0.002). Severity of anaemia and high levels of .82-M were also related to HIV infection (p<0.008 and <0.0005, respectively) (Table 2) . Significant differences of biochemical parameters of pleural fluid between patients with and without HIV infection were shown for albumin and ')'-globulin levels (Table 2). Predominance of lymphocytes in the differential cell count of WBC in pleural fluid was present in 84% of patients, irrespective of HIV Table 2-Hematologic and Biochemical Values of 112 Patients With TB Pleuritis Related to HIV Status
Blood Hemoglobin, g/dL iJ2-M levels, mg/ L Pleural fluid Glucose, mmol/L Cholesterol, mmol/L Protein, g/ L Albumin, g/ L -y-globulin, g/ L
HIV (+) TB n=65 Mean (SD)
HIV (-) TB n=47 Mean (SD)
p Value
9.67 (2.64) 6.5 (3. 78)
11.05 (2.59) 4.5 (2.20)
<0.008 <0.0005
3.82 (1.42) 2.24 (0.56) 60.09 (13.34) 22.27 (9.81) 25.81 (9.81)
4.26 (1.68) 2.35 (0.69) 55.67 (9.11) 26.19 (7.76) 14.97 (5.85)
NS* NS NS <0.008 <0.0001
*NS=not significant. HIV+ and HIV- Patients With Tuberculous Pleural Effusion (Richter eta/)
status. Among HIV -seropositive patients, those with skin test anergy to PPD had significantly less often protein levels of >70 g/L, albumin of >25 g/L, and ')'-globulin of >25 g/L than those with a positive PPD reaction (12% vs 39%, 20% vs 50%, and 24% vs 54%, respectively). High ')'-globulin levels of >25 g/L were also more common in HIV -infected patients with high ~2-M levels of >6 mg/L than in those with ~2-M levels of <6 mg/L (13/25 vs 8/36, p<0.02) .
Radiologic Investigations
patients with TB pleuritis was higher than the rate of 38% among patients with pulmonary TB.9 Similar differences of HIV seroprevalence rates between patients with extrapulmonary and pulmonary TB have been reported from other African countries. 10-12 Disseminated TB occurs more often when HIV infection is coexisting. 13·14 In Africa, evidence of disseminated TB in HIV -infected patients has been demonstrated before in clinical and autopsy studies.I5-17 In Zimbabwe, 16 extrathoracic lymphadenopathy in HIV -associated pulmonary TB was mainly caused by TB and in our series of HIV -associated TB pleuritis as well.
HIV -positive patients did not differ significantly from HIV -negative patients with regard to radiographic chest findings of mediastinal shift as sign of massive effusion (20% vs 26.7%), bilateral effusions (9.5% vs 8.9%), presence of hilar/mediastinal adenopathy (12.7% vs 11.1%), and presence of infiltrates (40% vs 35.6%). The radiologic patterns of primary and reactivation TB were equally distributed among HIV -positive and HIV -negative patients. In 78% of cases, there was evidence of primary TB and in 15% there was evidence of reactivation TB. In 7% of patients with infiltrates in all six lung fields, no classification was possible. The radiologic pattern of HIVinfected patients did not show a significant correlation with the results of PPD skin test or of ~2-M levels.
A longer duration of general symptoms like fever in patients with HIV -associated TB has been reported from Africa before.l 1 We also found that HIVseropositive patients with TB had more often constitutional and TB-related symptoms and, in concordance with the literature, also HIV -related features.ll.l 2 Immunosuppression by HIV infection may account for these differences as reflected by the nearly 50% of PPD skin test anergy among HIV -infected patients. This percentage is higher than reported among HIV -positive patients with TB from Zaire and Uganda who mainly had pulmonary
Diagnosis
Laboratory Presentation
Among the 112 patients with TB, the TB was proved by culture, smear, or histologic findings in 84 patients (75%) and considered probable in 28 patients. In HIV -infected patients, the diagnosis of TB was confirmed more often and disseminated TB appeared to be more common than in the HIV -negative patients (Table 3). Lymph nodes were the most frequent site of dissemination (in 26% of HIV-positive and 10.6% of HIV -negative patients).
The more pronounced anemia in HIV -infected patients may reflect the severity and dissemination of TB in immunosuppressed individuals and the presence of other HIV -related complications. The pleural fluid analysis revealed no differences between HIV-positive and HIV-negative patients with regard to mean glucose and cholesterol levels. Decreased pleural fluid glucose levels of <3.3 mmol/L were found in about 30% of patients with TB, which is not much different from the 20% found by others.l 9 Increased cholesterol level of > 1.43 mmol/L that recently has been proposed to differentiate exudates from transudates 20 was found in about 85% of our patients. To our knowledge, data of protein electrophoresis of pleural fluid in patients with TB have not been presented before. By serum electrophoresis, a decrease in total proteins and albumin with corresponding increase in ')'-globulin fraction was found in patients with pulmonary TB.21 We did not measure serum protein in our patients. It is known that during inflammation, capillary permeability increases, allowing high molecular proteins to leak into the pleural fluid . The pleural fluid total protein levels of HIV -seronegative patients were in the same range as found by others.6•22 With associated HIV infection, protein levels tended to be higher, mainly due to the
DISCUSSION TB Pleuritis and HIV Infection Tuberculosis is a common cause of pleural effusion at MMC. In MMC, the HIV infection rate of 58% in Table 3-Diagnosis of TB Pleuritis in 112 Patients Related to HN Status
(1) Proven TB total Pleuritis only Pleuritis and pulmonary TB Dissemination (2) Probable TB *NS=not significant.
HIV (+) TB n=65 n (%)
HIV (-) TB n=47 n (%)
p Value
54 (83.1) 30 (46.2) 4 (6.2)
30 (63.8) 21 (44.7) 4(8.5)
<0.02 NS* NS
20 (30.8) 11 (16.9)
5 (10.6) 17 (36.2)
<0.02 <0.02
Clinical Features
TB.ll,18
CHEST /106/5/ NOVEMBER, 1994
1473
increased ')'-globulin fraction .
Humoral and Cellular Immunity in HIV-Infect ed Patients Jacobson et aJ23 have shown that polyclonal activation of B lymphocytes as measured by hypergammaglobulinemia develops within l year after seraconversion , persists at high levels, and declines 8 to 16 months after diagnosis of AIDS , parallel to a decrease of CD4 cells. In our patients, a d celining B-cell response as reflected by low ')'-globulin levels in pleural fluid was associated with a decreased cellular immune response as expressed by a negativ~ Mantoux reaction. HIV -infected patients with relative high level of ')'-globulin in pleural fluid, however, may still have well preserved B-and T-cell responses and are probably in an earlier stage of HIV infection. The fact that high /1z-M levels in HIV -infected patients were associated with high ')'-globulin levels suggests that interpretation of /1z~M levels should be cautious when an opportunistic infection is coexisting. Tuberculosis can independently increase {1 2-M levels 24 and the effect of TB on raising the ')'-globulin levels may be greatest in an earlier stage of HIV infection when B-cell responses tend to be hyperreactive. TB Pleuritis- Primary or Reactivation Tuberculosis? In industrialized countries, a change of epidemiologic aspects of TB pleuritis has been noted independently of HIV infection. Prior to 1980, TB pleuritis was considered a late manifestation of primary TB in young adults and children. With the declining prevalence in the young population, TB has now become a disease of older adults with an increased proportion of cases of reactivation pleuritis.6•7 In sub-Saharan Africa, the TB epidemic started only a few generations ago and has just reached its natural peak. 25 This early stage of epidemic is reflected by the young age group that i s mostly affected and prone to early dissemination and primary disease. It is known that patients with coexisting HIV infection are at increased risk to develop either reactivation or primary TB. Our data show that the overall radiologic pattern of TB pleuritis remains unchanged in a high prevalence area of HIV and TB infection. This may be due to the fact that in HIV -infected patients, primary and reactivation pleuritis increase at the same rate. CoNcLusioNs In our series of patients with TB pleuritis, most had associated HIV infection. HIV -seropositive patients more often had disseminated TB and were generally more ill than seronegative patients. HIV infection and presence or absence of cellular immunity as re1474
fleeted by PPD skin test results in HIV -infected patients significantly altered the laboratory presentation, but not the chest radiographic features . ACKNOWLEDGMENTS: The authors thank the laboratory staff of the departments of hematology , biochemistry, histopathology, and the TB reference laboratory for their skilled services. The authors thank Prof. D.H. Mwakyusa and Prof. S. Y. Maselle for their useful advice, and Prof. A. de Geus and Dr. P. Speelman for their helpful suggestions about the manuscript. R EFERENCES
Sudre P, ten Dam G, Kochi A. Tuberculosis: a global overview of the situation today. WHO Bull OMS 1992; 70:149-59 2 Narain JP, Raviglione MC, Kochi A. HIV-associated tuberculosis in developing countries: epidemiology and strategies for prevention. Tuber Lung Dis 1992; 73:311-21 3 FitzGerald MJ, Grzybowski S, Ellen EA. The impact of human immunodeficiency virus infection on tuberculosis and its control. Chest 1991; 100:191-200 4 Richter C, Ndosi B, Mwam my AS, Mbwambo RK. Extrapulmonary tuberculosis-a simple diagnosis? a retrospective study at D ares Salaam , Tanzania. Trop Geogr Med 1991 ; 43:375-78 5 Centers for Disease Control. Tuberculosis and human immunodeficiency virus infection: recommendations of the Advisory Committee for the Elimination of Tuberculosis. MMWR 1989; 34:236-50 6 Epstein DM, Kline LR, Albelda SM, Miller WT. Tuberculous pleural effusions. Chest 1987; 91:106-09 7 Antoniskis D, Amin K, Barnes PF. Pleuritis as a manifestation of reactivation tuberculosis. Am J M ed 1990; 89:447-50 8 Richter C, Perenboom R, Swai ABM, Kitinya J, Mtoni I, Chande H, et al. Diagnosis of tuberculosis in patients with pleural effusion in an area of HIV-infection and limited diagnostic facilities. Trop Geogr Med 1994 (in press) 9 Pallangyo KJ, Hakansson A, Mtesa J, Lema L, Minjas J, Bredberg-Raden U, et al. HIV-infection and its influence on outcome among adult patients with malaria, meningitis, pneumonia, pyomyositis or tuberculosis in Dar es Salaam, Tanzania. Fifth International Conference on AIDS in Africa, Kinshasa, Zaire; October 10-12, 1990 (abstract FPB 41 ) 10 Nunn P, Gicheha C, Hayes R, Gathua S, Brindle R, Kibuga D , et al. Cross-sectional survey of HIV infection among patients with tuberculosis in Nairobi, Kenya. Tuber Lung Dis 1992; 73:45-51 11 Colebunders RL, Ryder RW, Nzilambi N, Dikilu K, Willame JC, Kaboto M, et al. HIV infection in patients with tuberculosis in Kinshasa, Zaire. Am Rev Respir Dis 1989; 139: 1082-85 12 Elliot AM, Luo N, Tembo G, Halwiindi B, Steenbergen G, Machiels L, et l.a Impact of HIV on tuberculosis in Zambia: a cross sectional study. BMJ 1990; 301:412-15 13 Soriano E , Mallolas J, Gatell JM, Latorre X,Miro JM, Pecchiar M, e t al. Characteristics of tuberculosis in HIV-infected patients: a case-control study. AIDS 1988; 2:429-32 14 Handwerger S, Mildvan D, Senie R, McKinley FW. Tuberculosis and the acquired immunodeficiency syndrome at a New York City hospital: 1978-1985. Chest 1987; 91:176-80 15 Gilks CF, Brindle RJ, Otieno LS, Bhatt SM, Newnham RS, Simani PM, et al. Extrapulmonary and disseminated tuberculosis in HIV-1-seropositive patients presenting to the acute medical services in Nairobi. AIDS 1990; 4:981-85 16 Pithie AD, Chicksen B.Fine-needle extrathoracic lymph node aspiration in HIV -associated sputum-negative tuberculosis. Lancet 1992; 340:1504-05 17 Lucas S, Hounnou A, Beaumel A, Diomande M, Peacock C, Kadio A, et l.a The pathology of adult HIV infection in Abidjan, Cote D'Ivoire. Eighth Inte rnational Conference on AIDS, HIV +
and HIV - Patients With Tuberculous Pleural Effusion (Richter eta/)
18
19 20
21
Amsterdam, the Netherlands: July 19-24, 1992 (abstract PoB 3751) Migliori GB, Borghesi A, Adriko C, Manfrin V, Okware S, Naamara W, et al. Tuberculosis and HIV infection association in a rural district of northern Uganda: epidemiological and clinical considerations. Tuber Lung Dis 1992; 73:285-90 Sahn SA. Pleural fluid analysis: narrowing the differential diagnosis. Semin Respir Med 1987; 9:22-9 Valdes L, Pose A, Suarez J, Gonzalez-Juanatey JR, Sarandeses A, Jose ES, et al. Cholesterol: a useful parameter for distinguishing between pleural exsudates and transsudates. Chest 1991; 99:1097-1102 Azis S, Agha F, Lodi TZ. Protein electrophoresis in tuberculosis. J Pak Med Assoc 1991; 41 :58-60
22 Seibert AF, Haynes Jr, Middleton R, Bass JB. Tuberculous pleural effusion: twenty-year experience. Chest 1991 ; 99:883-86 23 Jacobson DL, McCutchan JA, Spechko PL, Abramson I, Smith RS, Bartok A, et al. The evolution of lymphadenopathy and hypergammaglobulinaemia are evidence for early and sustained polyclonal B lymphocyte activation during human immunodeficiency virus infection. J Infect Dis 1991; 163:240-46 24 Wallis RS, Vjecha M, Amir-Tahmasseb M, Okwera A, Byekwaso F, Nyole S, et al. Influence of tuberculosis on human immunodeficiency virus (HIV-1): enhanced cytokine expression and elevated /12-microglobulin in HIV -!-associated tuberculosis. J Infect Dis 1993; 167:43-8 25 Bates JH , Stead WW. The history of tuberculosis as a global epidemic. Med Clin North Am 1993; 77:1205-17
CHEST I 106 I 5 I NOVEMBER, 1994
1475
Jn 1990, 8 million new cases of tuberculosis (TB)
were r e ported worldwide ; 95% of them occurred in developing countries_! The World Health Organization estimated that by January 1992, 3 million people in sub-Saharan countries were dually infected with the tubercle bacillus and the immunodeficiency virus (HIV). Their cumulative life-time risk of progressing to active TB is estimated to be 30% or more. 2 The proportion of extrapulmonary TB is higher among HIV -infected patients compared with patients without HIV infection. 3 To examine the impact of HIV infection on clinical features , diagnosis, and outcome in patients with extrapulmonary TB manifestations, we performed a prospective study at Muhimbili Medical Centre (MMC, the teaching hospital of the University of Dares Salaam). This report describes the results of a study of the influence of HIV infection on the clinical, laboratory, and radiologic *From the Departments of Medicine (Drs. Richter, Perenboom, and Swai), Histopathology (Drs. Kitinya and Chande), Radiology (Dr. Kazema), Biochemistry (Dr. Chuwa), Muhimbili Medical Centre, and National Institute for Medical Research (Dr. Mtoni), Dar es Salaam, Tanzania. Supported by the Royal Netherlands Tuberculosis Association. Manuscript received Sept. 13, 1993; revision accepted May 5, 1994.
Reprint requests: Dr. Richter, Ruysdaellaan 21, 1213 er Hilversum, the Netherlands.
and )'-globulin levels of pleural fluid. However, the radiographic features did not differ with respect to HIV status; they were predominantly those of primary pleuritis (78% in each group). We conclude that coexisting HIV infection affects clinical and laboratory features, but not the radiographic presentation of patients with (Chest 1994; 106:1471-76) TB pleuritis in Tanzania. AFB=acid-fast bacilli; f:I2·M={:I 2-microglobulin; ELISA= enzyme-linked immunosorbent assay; HIV=human immunodeficiency virus; MMC=Muhimbili Medical Centre; PPD=purificd protein derivative; TB=tuberculosis Key words: clinical, laboratory, and radiologic features; disseminated tuberculosis; HIV infection; Tanzania; Tuberculous pleuritis
features among patients with tuberculous pleuritis , which is a commonly diagnosed form of extrapulmonary TB at our Centre 4 M ETHODS
Patient Selection
All patients with pleural effusion admitted to four medical wards in MMC between January and August 1991 were enrolled. Excluded were patients already receiving antituberculous therapy and those with obvious congestive heart failure as underlying cause of pleural effusion. Investigations
A questionnaire designed to collect data of history and clinical examination with special attention to possible HIV -related symptoms and signs were completed for each patient. The result of a purified protein derivative (PPD) skin test, using 0.1 mL of 2 TU of RT-23 tuberculin (Statens Seruminstitut, Copenhagen, Denmark), which is equivalent to 5 TU of PPD-S were noted. No anergy panel was used. According to Centers for Disease Control and Prevention criteria,5 the skin test was defined positive when the induration at 48 h was ;:::5 mm in HIV-infected patients or ;:::10 mm in HIV-seronegative patients. The radiologic findings of a chest radiograph were recorded. We applied radiologic criteria to classify our patients as ·suffering from primary or reactivation TB (primary: no parenchymal infiltrates at all or the presence of hilar adenopathy with or without noncavitary mid-lower field infiltrates; reactivation: infiltrates or cavities in upper field ). 6•7 After pretest counseling and obtained informed consent for HIV testing, a blood sample was taken for CHEST 1106 I 5 I NOVEMBER, 1994
1471
a complete blood cell count, erythrocyte sedimentation rate (ESR), iJ2-microglobulin ({12-M) levels (Enzygnost iJ2-M micro Behring), and HIV test (Organon ELISA). Samples that tested positive in the enzyme-linked immunosorbent assay (ELISA) test, were confirmed by Western Blot. All patients underwent a pleural fluid aspiration and a blind pleural biopsy (Abraham needle). Pleural fluid analysis consisted of biochemical, bacterial, and cytologic examination. Smears of centrifuged deposit of pleural fluid, pleural tissue, and huffy coat were performed for detection of acid-fast bacilli (AFB) by auramine stain-if positive confirmed by Ziehl-Neelsen stain-in all patients, of urine (in patients with leukocyturia or proteinuria), of lymph nodes (in patients with coexisting lymphadenopathy on physical examination), and whenever clinically indicated of sputum, ascites, and bone marrow. In addition, mycobacterial cultures of all mentioned specimens were carried out on Loewenstein-Jensen medium. Histologic examination was done of the pleural tissue and of lymph nodes.
Criteria for Diagnosis of TB The diagnosis was considered proved when AFB were identified in auramine/Ziehl-Neelsen stains, when mycobacteria were cultured, or when by histologic examination granulomas with caseous necrosis or AFB were seen. Disseminated TB was considered to be present when TB could be proved in at least two different organs. Diagnosis of probable TB was mainly based on a good response of the patient to antituberculous therapy (disappearance of pleural fluid and of TB-related complaints).8
Analysis Data were analyzed using the statistical package for the social sciences. Where appropriate we used the x2 test with Yates' correction or Fisher's Exact Test for 2X2 tables and the Student's t test.
Patients
RESULTS
There were 133 patients with pleural effusion. Fifteen of these patients could not be included in the final analysis of the study because of death (3 patients), hospital discharge (4 patients), or fast improvement with antibiotic therapy (3 patients) before the diagnostic procedures could be performed, or because of failure to get pleural tissue and follow-up data (5 patients). All remaining 118 patients had exudates (pleural protein levels >30 g/ L) . Tuberculosis was diagnosed in 112 patients (94.9%), adenocarcinoma in 1, bacterial infection in 2, and aspecific inflammation in 3 patients. Only three patients with TB had been treated for TB in the past (one HIV positive, two HIV negative); 40% of all patients with TB had a bacille Calmette-Guerin (BCG) vaccination scar. The HIV positivity rate in the patients with TB pleuritis was 58% and three of six patients without TB were HIV seropositive.
Clinical Features The HIV -positive patients more often had complaints at presentation than the patients without HIV infection with significant differences for dyspnea, fever, night sweat, fatigue, and diarrhea (Table 1). The mean duration of complaints before presentation 1472
Table !-Clinical Features of 112 Patients With TB Pleuritis at Enrollment Related to HIV Status
Sex, male Age <40 yr Cough Chest pain Dyspnea Fever Night sweat Fatigue Weight loss (history) Diarrhea Respiratory rate >30/ min Hepatomegaly Splenomegaly Lymphadenopathy*
HIV (+) TB, % n=65
HIV (-) TB, % n=47
p Value
53.8 78.5 92.3 84.6 96.9 78.5 83.1 90.8 89.2 23.1 50.8 41.5 40.0 35.4
55.3 78.7 83.3 83.0 80.9 61.7 63.8 74.5 83.0 4.3 25.5 21.3 14.9 12.8
NSf NS NS NS <0.007 <0.05 <0.02 <0.02 NS <0.007 <0.007 <0.03 <0.004 <0.007
*Found on physical examination. fNS=not significant.
was 1 to 3 weeks longer for the HIV -positive patients than for the HIV -negative patients; for chest pain, cough, and weight loss, the duration was 8 to 9 weeks (range, 1 to 40 weeks) . For the duration of fatigue, this difference was significant (7.8 weeks among HIV -positive patients and 4.9 weeks among HIVnegative patients (p<0.03).
Laboratory Investigations A negative PPD skin test was present in 25 of 53 (47.2%) HIV-positive patients, but in only 5 of 43 (11.6%) HIV -negative patients (p<0.002). Severity of anaemia and high levels of .82-M were also related to HIV infection (p<0.008 and <0.0005, respectively) (Table 2) . Significant differences of biochemical parameters of pleural fluid between patients with and without HIV infection were shown for albumin and ')'-globulin levels (Table 2). Predominance of lymphocytes in the differential cell count of WBC in pleural fluid was present in 84% of patients, irrespective of HIV Table 2-Hematologic and Biochemical Values of 112 Patients With TB Pleuritis Related to HIV Status
Blood Hemoglobin, g/dL iJ2-M levels, mg/ L Pleural fluid Glucose, mmol/L Cholesterol, mmol/L Protein, g/ L Albumin, g/ L -y-globulin, g/ L
HIV (+) TB n=65 Mean (SD)
HIV (-) TB n=47 Mean (SD)
p Value
9.67 (2.64) 6.5 (3. 78)
11.05 (2.59) 4.5 (2.20)
<0.008 <0.0005
3.82 (1.42) 2.24 (0.56) 60.09 (13.34) 22.27 (9.81) 25.81 (9.81)
4.26 (1.68) 2.35 (0.69) 55.67 (9.11) 26.19 (7.76) 14.97 (5.85)
NS* NS NS <0.008 <0.0001
*NS=not significant. HIV+ and HIV- Patients With Tuberculous Pleural Effusion (Richter eta/)
status. Among HIV -seropositive patients, those with skin test anergy to PPD had significantly less often protein levels of >70 g/L, albumin of >25 g/L, and ')'-globulin of >25 g/L than those with a positive PPD reaction (12% vs 39%, 20% vs 50%, and 24% vs 54%, respectively). High ')'-globulin levels of >25 g/L were also more common in HIV -infected patients with high ~2-M levels of >6 mg/L than in those with ~2-M levels of <6 mg/L (13/25 vs 8/36, p<0.02) .
Radiologic Investigations
patients with TB pleuritis was higher than the rate of 38% among patients with pulmonary TB.9 Similar differences of HIV seroprevalence rates between patients with extrapulmonary and pulmonary TB have been reported from other African countries. 10-12 Disseminated TB occurs more often when HIV infection is coexisting. 13·14 In Africa, evidence of disseminated TB in HIV -infected patients has been demonstrated before in clinical and autopsy studies.I5-17 In Zimbabwe, 16 extrathoracic lymphadenopathy in HIV -associated pulmonary TB was mainly caused by TB and in our series of HIV -associated TB pleuritis as well.
HIV -positive patients did not differ significantly from HIV -negative patients with regard to radiographic chest findings of mediastinal shift as sign of massive effusion (20% vs 26.7%), bilateral effusions (9.5% vs 8.9%), presence of hilar/mediastinal adenopathy (12.7% vs 11.1%), and presence of infiltrates (40% vs 35.6%). The radiologic patterns of primary and reactivation TB were equally distributed among HIV -positive and HIV -negative patients. In 78% of cases, there was evidence of primary TB and in 15% there was evidence of reactivation TB. In 7% of patients with infiltrates in all six lung fields, no classification was possible. The radiologic pattern of HIVinfected patients did not show a significant correlation with the results of PPD skin test or of ~2-M levels.
A longer duration of general symptoms like fever in patients with HIV -associated TB has been reported from Africa before.l 1 We also found that HIVseropositive patients with TB had more often constitutional and TB-related symptoms and, in concordance with the literature, also HIV -related features.ll.l 2 Immunosuppression by HIV infection may account for these differences as reflected by the nearly 50% of PPD skin test anergy among HIV -infected patients. This percentage is higher than reported among HIV -positive patients with TB from Zaire and Uganda who mainly had pulmonary
Diagnosis
Laboratory Presentation
Among the 112 patients with TB, the TB was proved by culture, smear, or histologic findings in 84 patients (75%) and considered probable in 28 patients. In HIV -infected patients, the diagnosis of TB was confirmed more often and disseminated TB appeared to be more common than in the HIV -negative patients (Table 3). Lymph nodes were the most frequent site of dissemination (in 26% of HIV-positive and 10.6% of HIV -negative patients).
The more pronounced anemia in HIV -infected patients may reflect the severity and dissemination of TB in immunosuppressed individuals and the presence of other HIV -related complications. The pleural fluid analysis revealed no differences between HIV-positive and HIV-negative patients with regard to mean glucose and cholesterol levels. Decreased pleural fluid glucose levels of <3.3 mmol/L were found in about 30% of patients with TB, which is not much different from the 20% found by others.l 9 Increased cholesterol level of > 1.43 mmol/L that recently has been proposed to differentiate exudates from transudates 20 was found in about 85% of our patients. To our knowledge, data of protein electrophoresis of pleural fluid in patients with TB have not been presented before. By serum electrophoresis, a decrease in total proteins and albumin with corresponding increase in ')'-globulin fraction was found in patients with pulmonary TB.21 We did not measure serum protein in our patients. It is known that during inflammation, capillary permeability increases, allowing high molecular proteins to leak into the pleural fluid . The pleural fluid total protein levels of HIV -seronegative patients were in the same range as found by others.6•22 With associated HIV infection, protein levels tended to be higher, mainly due to the
DISCUSSION TB Pleuritis and HIV Infection Tuberculosis is a common cause of pleural effusion at MMC. In MMC, the HIV infection rate of 58% in Table 3-Diagnosis of TB Pleuritis in 112 Patients Related to HN Status
(1) Proven TB total Pleuritis only Pleuritis and pulmonary TB Dissemination (2) Probable TB *NS=not significant.
HIV (+) TB n=65 n (%)
HIV (-) TB n=47 n (%)
p Value
54 (83.1) 30 (46.2) 4 (6.2)
30 (63.8) 21 (44.7) 4(8.5)
<0.02 NS* NS
20 (30.8) 11 (16.9)
5 (10.6) 17 (36.2)
<0.02 <0.02
Clinical Features
TB.ll,18
CHEST /106/5/ NOVEMBER, 1994
1473
increased ')'-globulin fraction .
Humoral and Cellular Immunity in HIV-Infect ed Patients Jacobson et aJ23 have shown that polyclonal activation of B lymphocytes as measured by hypergammaglobulinemia develops within l year after seraconversion , persists at high levels, and declines 8 to 16 months after diagnosis of AIDS , parallel to a decrease of CD4 cells. In our patients, a d celining B-cell response as reflected by low ')'-globulin levels in pleural fluid was associated with a decreased cellular immune response as expressed by a negativ~ Mantoux reaction. HIV -infected patients with relative high level of ')'-globulin in pleural fluid, however, may still have well preserved B-and T-cell responses and are probably in an earlier stage of HIV infection. The fact that high /1z-M levels in HIV -infected patients were associated with high ')'-globulin levels suggests that interpretation of /1z~M levels should be cautious when an opportunistic infection is coexisting. Tuberculosis can independently increase {1 2-M levels 24 and the effect of TB on raising the ')'-globulin levels may be greatest in an earlier stage of HIV infection when B-cell responses tend to be hyperreactive. TB Pleuritis- Primary or Reactivation Tuberculosis? In industrialized countries, a change of epidemiologic aspects of TB pleuritis has been noted independently of HIV infection. Prior to 1980, TB pleuritis was considered a late manifestation of primary TB in young adults and children. With the declining prevalence in the young population, TB has now become a disease of older adults with an increased proportion of cases of reactivation pleuritis.6•7 In sub-Saharan Africa, the TB epidemic started only a few generations ago and has just reached its natural peak. 25 This early stage of epidemic is reflected by the young age group that i s mostly affected and prone to early dissemination and primary disease. It is known that patients with coexisting HIV infection are at increased risk to develop either reactivation or primary TB. Our data show that the overall radiologic pattern of TB pleuritis remains unchanged in a high prevalence area of HIV and TB infection. This may be due to the fact that in HIV -infected patients, primary and reactivation pleuritis increase at the same rate. CoNcLusioNs In our series of patients with TB pleuritis, most had associated HIV infection. HIV -seropositive patients more often had disseminated TB and were generally more ill than seronegative patients. HIV infection and presence or absence of cellular immunity as re1474
fleeted by PPD skin test results in HIV -infected patients significantly altered the laboratory presentation, but not the chest radiographic features . ACKNOWLEDGMENTS: The authors thank the laboratory staff of the departments of hematology , biochemistry, histopathology, and the TB reference laboratory for their skilled services. The authors thank Prof. D.H. Mwakyusa and Prof. S. Y. Maselle for their useful advice, and Prof. A. de Geus and Dr. P. Speelman for their helpful suggestions about the manuscript. R EFERENCES
Sudre P, ten Dam G, Kochi A. Tuberculosis: a global overview of the situation today. WHO Bull OMS 1992; 70:149-59 2 Narain JP, Raviglione MC, Kochi A. HIV-associated tuberculosis in developing countries: epidemiology and strategies for prevention. Tuber Lung Dis 1992; 73:311-21 3 FitzGerald MJ, Grzybowski S, Ellen EA. The impact of human immunodeficiency virus infection on tuberculosis and its control. Chest 1991; 100:191-200 4 Richter C, Ndosi B, Mwam my AS, Mbwambo RK. Extrapulmonary tuberculosis-a simple diagnosis? a retrospective study at D ares Salaam , Tanzania. Trop Geogr Med 1991 ; 43:375-78 5 Centers for Disease Control. Tuberculosis and human immunodeficiency virus infection: recommendations of the Advisory Committee for the Elimination of Tuberculosis. MMWR 1989; 34:236-50 6 Epstein DM, Kline LR, Albelda SM, Miller WT. Tuberculous pleural effusions. Chest 1987; 91:106-09 7 Antoniskis D, Amin K, Barnes PF. Pleuritis as a manifestation of reactivation tuberculosis. Am J M ed 1990; 89:447-50 8 Richter C, Perenboom R, Swai ABM, Kitinya J, Mtoni I, Chande H, et al. Diagnosis of tuberculosis in patients with pleural effusion in an area of HIV-infection and limited diagnostic facilities. Trop Geogr Med 1994 (in press) 9 Pallangyo KJ, Hakansson A, Mtesa J, Lema L, Minjas J, Bredberg-Raden U, et al. HIV-infection and its influence on outcome among adult patients with malaria, meningitis, pneumonia, pyomyositis or tuberculosis in Dar es Salaam, Tanzania. Fifth International Conference on AIDS in Africa, Kinshasa, Zaire; October 10-12, 1990 (abstract FPB 41 ) 10 Nunn P, Gicheha C, Hayes R, Gathua S, Brindle R, Kibuga D , et al. Cross-sectional survey of HIV infection among patients with tuberculosis in Nairobi, Kenya. Tuber Lung Dis 1992; 73:45-51 11 Colebunders RL, Ryder RW, Nzilambi N, Dikilu K, Willame JC, Kaboto M, et al. HIV infection in patients with tuberculosis in Kinshasa, Zaire. Am Rev Respir Dis 1989; 139: 1082-85 12 Elliot AM, Luo N, Tembo G, Halwiindi B, Steenbergen G, Machiels L, et l.a Impact of HIV on tuberculosis in Zambia: a cross sectional study. BMJ 1990; 301:412-15 13 Soriano E , Mallolas J, Gatell JM, Latorre X,Miro JM, Pecchiar M, e t al. Characteristics of tuberculosis in HIV-infected patients: a case-control study. AIDS 1988; 2:429-32 14 Handwerger S, Mildvan D, Senie R, McKinley FW. Tuberculosis and the acquired immunodeficiency syndrome at a New York City hospital: 1978-1985. Chest 1987; 91:176-80 15 Gilks CF, Brindle RJ, Otieno LS, Bhatt SM, Newnham RS, Simani PM, et al. Extrapulmonary and disseminated tuberculosis in HIV-1-seropositive patients presenting to the acute medical services in Nairobi. AIDS 1990; 4:981-85 16 Pithie AD, Chicksen B.Fine-needle extrathoracic lymph node aspiration in HIV -associated sputum-negative tuberculosis. Lancet 1992; 340:1504-05 17 Lucas S, Hounnou A, Beaumel A, Diomande M, Peacock C, Kadio A, et l.a The pathology of adult HIV infection in Abidjan, Cote D'Ivoire. Eighth Inte rnational Conference on AIDS, HIV +
and HIV - Patients With Tuberculous Pleural Effusion (Richter eta/)
18
19 20
21
Amsterdam, the Netherlands: July 19-24, 1992 (abstract PoB 3751) Migliori GB, Borghesi A, Adriko C, Manfrin V, Okware S, Naamara W, et al. Tuberculosis and HIV infection association in a rural district of northern Uganda: epidemiological and clinical considerations. Tuber Lung Dis 1992; 73:285-90 Sahn SA. Pleural fluid analysis: narrowing the differential diagnosis. Semin Respir Med 1987; 9:22-9 Valdes L, Pose A, Suarez J, Gonzalez-Juanatey JR, Sarandeses A, Jose ES, et al. Cholesterol: a useful parameter for distinguishing between pleural exsudates and transsudates. Chest 1991; 99:1097-1102 Azis S, Agha F, Lodi TZ. Protein electrophoresis in tuberculosis. J Pak Med Assoc 1991; 41 :58-60
22 Seibert AF, Haynes Jr, Middleton R, Bass JB. Tuberculous pleural effusion: twenty-year experience. Chest 1991 ; 99:883-86 23 Jacobson DL, McCutchan JA, Spechko PL, Abramson I, Smith RS, Bartok A, et al. The evolution of lymphadenopathy and hypergammaglobulinaemia are evidence for early and sustained polyclonal B lymphocyte activation during human immunodeficiency virus infection. J Infect Dis 1991; 163:240-46 24 Wallis RS, Vjecha M, Amir-Tahmasseb M, Okwera A, Byekwaso F, Nyole S, et al. Influence of tuberculosis on human immunodeficiency virus (HIV-1): enhanced cytokine expression and elevated /12-microglobulin in HIV -!-associated tuberculosis. J Infect Dis 1993; 167:43-8 25 Bates JH , Stead WW. The history of tuberculosis as a global epidemic. Med Clin North Am 1993; 77:1205-17
CHEST I 106 I 5 I NOVEMBER, 1994
1475