Bridging the Gap: Conference Record [Abstract book, International Conference on AIDS (12th: 1998: Geneva, Switzerland)]

138 Abstracts 13259-13263

12th World AIDS Conference

613*/13259 The impact of HIV infection on tuberculosis
control: A prospective cohort study of African
patients
Pamela Sonnenberg', J. Murray2, S.C.P. Shearer3, P. Godfrey-Faussett4.
Epidemiology Research Unit, Johannesburg; 2National Centre for Occupational
Health, Johannesburg; 3Gold Fields of South Africa, Johannesburg, South
Africa; 4Zambart Project & LSHTM, Lusaka, Zambia
Background: The South African gold mining industry has high rates of pulmonary
tuberculosis (TB), and more recently, HIV infection in its migrant workforce. The
infrastructure and relatively stable population offered an opportunity to study the
outcome of a TB control programme in the age of HIV infection. Causes and
risk factors for treatment failure and death were investigated in this prospective
cohort.
Setting: Tuberculosis case finding and directly observed therapy are provided
to workers on four gold mines near Johannesburg, through primary health care
facilities at each mine hostel, while the hospital provides secondary care and
supervision of the TB control programme. Methods: All patients with culture
positive TB in 1995 were recruited, patients interviewed, HIV status and CD4
counts obtained for consenting subjects and outcome assessed after 6 months.
Autopsies were performed on 12/26 known HIV positive deaths.
Results: Of 376 cases of sputum positive TB (1318 per 100000), 49% were
HIV positive. No association was found between HIV status and previous tuberculosis. The overall cure rate was 82.2%, and if deaths are excluded, cure rates
were similar for HIV positive (89%) and negative (88%) patients. The proportion
of patients with recurrent disease (22%) at entry was high and treatment failure
common in this group. Risk factors for failure in new cases were taking leave and
irregular clinic attendance, but not drug resistance. Failure in recurrent cases was
associated with leave and single drug resistance at diagnosis. Total mortality was
0.5% in HIV negative and 13.7% in HIV positive patients, strongly related to CD4
lymphocyte depletion. Early mortality in HIV positives was due to TB, with late
deaths most commonly due to cryptococcal pneumonia.
Conclusion: We have shown that a well-run TB control programme can result
in acceptable cure rates in a population with a very high incidence of pulmonary tuberculosis and a high prevalence of HIV infection. Particular vigilance
is needed in such an environment for concurrent infections, which if undetected
and untreated, may contribute significantly to mortality during TB treatment in HIV
positive patients.
132601 Characteristics of HIV-infected tuberculosis patients and
the natural history of HIV and TB coinfection
Mark N. Lobato, J.L. Jones, J. Li, J.W. Ward, P.L. Fleming. Centers for Disease
Control 1600 Clifton Rd. Mail Stop E-47 Atlanta 30333 GA, USA
Objectives: To characterize HIV-infected persons with tuberculosis (TB) and to
describe the spectrum of disease and survival among HIV-infected persons that
have TB.
Methods: Six states that reported 62% of AIDS cases nationally in 1992-94
provided information on all cases of TB in persons with HIV infection for a mean
follow-up of 29 months. Cases were categorized by whether the initial AIDS
diagnosis was TB, TB and another opportunistic infection (01) diagnosed concurrently, or another 01 with TB diagnosed later. Data included demographics,
HIV exposure mode, CD4 count at AIDS diagnosis, subsequent Ols, and date of
death. Survival was compared by using a product-limit estimate.
Results: In the six high morbidity states, 2,594 HIV-infected persons were
reported with TB. Cases were predominantly male (84%), nonwhite (80%), 25 to
44 years of age (78%), bornin the United States (73%) and had injection drug
use as an HIV exposure (50%). More of those who had TB initially were injection
drug users (p < 0.001). The initial AIDS-diagnosis were TB (72%), TB and an 01
(14%), and an Of other than TB (14%). Of persons whose AIDS-defining condition
was TB first, 41% had at least one other 01 subsequently. The most common
subsequent Ols were Pneumocystis carinii pneumonia (22%), wasting (12%)
candida esophagitis (8%), and disseminated MAC (8%). Survival from AIDS was
longer for those who had TB only at the initial AIDS diagnosis (p < 0.001).
AIDS diagnosis Median CD4 Median survival Interquartile range
at AIDS (mos.) (mos.)
TB first 110 38 16-censored
TB and Ol 44 21 12-42
Other 01 57 19 8-51
Conclusions: Although persons who had TB at the initial AIDS diagnosis
survived longer than did those who had other diagnoses, there is considerable
subsequent and comorbidity associated with TB. Survival of persons whose AIDS
diagnosis was TB and an 01 was similar to those that had an 01 initially.
1449*/13261 Feasibility of preventive therapy of tuberculosis for
HIV-infected subjects - GISTA Study

Giuseppe Ippolito, Giorgio Antonucci. CTR. rif. AIDS IRCCS L.Spallanzani via
Portuense 292 00149 Rome, Italy
Objective: The aim of the present study is to evaluate the feasibility of an isoniazid
preventive therapy (IPT) program for HIV-infected persons in Italy.
Design: Observational, prospective study.

Methods: Data on screening for tuberculosis infection and initiation of IPT were
collected for 2066 subjects presenting for the first time with HIV-infection to 27
clinical units of infectious disease in Italy over a 18 months period.
Results: 1618 (78.3%) out of the subjects 2066 considered actually underwent
PPD skin test, and 1519 (73.5%) subjects returned for skin test reading. Proportion of subjects who completed PPD screening was significantly higher in subjects
born in Italy compared to foreign borne subjects, while significant differences were
not observed according to age, sex, and HIV-transmission category. Out of the
1519 subjects screened, 90 (5.9%) were PPD positive, of whom 9 had active TB.
Proportion of PPD positivity was significantly related to CD4+ cell count (8.6%
among subjects with CD4+ > 350/ml, 6.2% among those with CD4+ 200-350/ml,
and 2.0% among those CD4+ < 200/ml). 113 (5.5%) out of 2066 enrolled subjects
were potential candidates for IPT: 90 PPD positive, 7 with documented history of
PPD positivity, 8 with inactive TB lesions on chest X-ray, and 8 "close contacts".
Of these subjects, 39 subjects had contraindications to IPT (32 with chronic liver
disease) and 24 refused IPT. 52 (2.5%) subjects of those initially considered
started IPT.
Conclusions: The following factors may limit the effectivness of an IPT program: testing with PPD patients with advanced immunosuppression; high number
of IPT contraindication (especially chronic liver diseases) and low level of acceptance of IPT by the patients.
113262 Tuberculosis diagnosis in HIV+/AIDS patients in the City
of Rio de Janeiro
Guida Silva, D. Barreira, A.L.M. Bellizzi, S.M.E. Santo. Health Secretariat of the
City of Rio De Janeiro, rua Casuarina, Humaita, Rio De Janeiro, RJ, Brazil
Objective: To study the methods used in the Primary Health Care Units in the
City of Rio de Janeiro to diagnose tuberculosis (TB) in HIV+/AIDS patients and in
HIV negative patients.
Design: Cross-sectional study.
Methodology: 376 individuals from 18 Primary Care Units from July to December of 1996, with more than 14 years old, both sex, with diagnosis of pulmonary TB
were submitted after written consent to serologic testing for HIV, thorax X-rays and
sputum smear (Ziehl-Nielsen) and culture (Lowenstein-Jensen) for TB. The HIV
tests were done by ELISA method in the Central Laboratory, and the confirmation
with Western Blot. The X-rays were done in the Health Units.
Results: From 376 tested patients, 31 were MV positive and 345 negative.
Among the negatives 150 (43.8%) had positive smear for TB, 172 (49.8%) negative and 23 (6.4%) didn't get the sample. From the 172 with negative smear 14
(8.1%) had contaminated, 59 (34.3%) negative and 99 (57.6%) positive cultures.
From 31 HIV positive patients, only 26 (83.9%) got samples. From these 21
(80.8%) had negative smear and 5 (18.8%) positive. From the 21 negatives, 13
(61.9%) had negative and 8 (38.1%) positive cultures.
Conclusion: Comparing to the HIV negative group, the HIV positive patients
got sputum for diagnosis 2.5 times less, had 2.3 times less positivity to smear and
1.5 times less positively to culture. The diagnosis in these patients was confirmed
only in 58.1% of cases against 70.2% in the HIV negative TB patients. Other
means of TB diagnosis are necessary for these patients, and bronchoscopy shall
be available for them.
132631 BCG vaccine effectiveness in preventing tuberculosis
and its interaction with HIV infection
Maria Patricia Arbelaez1, K.E. Nelson2, A. Muhoz2. 1Facultad Nacional De
Salud Publica Apartado Aereo Medellin, Colombia; 2Johns Hopkins School of
Public Health, Baltimore, MD, USA
Background and Objective: The efficacy of Bacillus Calmette-Guerin (BCG)
vaccine in preventing tuberculosis (TB) has not been reported among HIV infected individuals. The purpose of this case-control study was to determine the
effect of HIV infection on the protection against TB conferred by BCG vaccination.
Methods: A matched case-control study, consisting of two groups of TB cases
differentiated by their HIV infection status, was conducted. Controls without TB
were also selected by HIV status. The study population consisted of 88 TB
cases and 88 controls among HIV positive individuals and 314 TB cases and
310 controls among HIV negative individuals. Matching by HIV status precluded
quantification of association between HIV and TB but did not preclude the investigations of interactions between BCG vaccination and HIV status on TB which
was the primary scientific aim of our study. Matching by HIV status made this
study feasible.
Results: The level of protection against all clinical forms of TB found was
22% among HIV positive individuals (Odds Ratio [OR]: 0.78 95% Confidence
Interval [95% CI]: 0.48-1.26) and 26% among HIV negatives (OR: 0.74, 95%
CI: 0.52-1.05). There was a significant difference (p-value = 0.002) in the level
of protection against extrapulmonar TB (ETB) between HIV negative (OR: 0.54,
95% CI: 0.32-0.93) and HIV positive individuals (OR: 1.36, 95% CI: 0.72-2.57).
Conclusions: BCG has a modest effect against all forms of TB independent
of HIV status, BCG confers protection against extrapulmonar TB among HIV

negative individuals, and HIV infection seems to abrogate the protective effect of
BCG against extrapulmonar TB. Our data support the public health importance
of BCG vaccine against extrapulmonar forms of TB among immunocompetent
individuals.