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

12th World AIDS Conference

Abstracts 13264-13268 139

616*/ 13264 HIV infection and TB associated with injecting
drug use and resistant M. tuberculosis, Bangkok,
Thailand
Juree Punnotok1, U. Pumprug1, T. Naiwatanakul2, N. Shaffer3, P. Supannachat1,
C. Choochotthaworn1, P. Ponglertnapagorn. 1 Central Chest Hospital, Central
Chest Hospital, 39 Tivanon Road, Monthaburi 1 100;2HIV/AIDS Collaboration,
Nonthaburi,3CDC & HIV/AIDS Collaboration, Nonthaburi;, Thailand
Objectives: To determine HIV seroprevalence among patients with pulmonary
tuberculosis (PTB) and compare epidemiologic and clinical characteristics and
levels of antimicrobial resistance of M. tuberculosis strains between HIV-positive
and HIV-negative PTB patients.
Methods: From June 1995 through July 1996, adults (age >16 yrs) at Central Chest Hospital with bacteriologically-confirmed PTB were offered voluntary
HIV counseling and testing and enrollment into a descriptive study employing
questionnaires and clinical examinations. M. tuberculosis strains were tested for
antimicrobial susceptibilities using the absolute concentration method on Ogawa
medium. A sample of HIV-pos patients with newly diagnosed PTB was evaluated
with lymphocyte immunophenotyping.
Results: HIV seroprevalence among 1,091 PTB patients was 22%. History of
injecting drug use was more common among HIV-pos (20%) than HIV-neg (3%)
patients (P.: 0.001). On chest x-ray, HIV-pos patients were more likely to have
miliary disease (11% vs. 3%, P < 0.001) and less likely to have cavitation (35%
vs. 66%, P < 0.001). Antimicrobial resistance was more common among M.
tuberculosis strains from HIV-pos patients (n = 191) than from HIV-neg patients
(n = 685): isoniazid, 12% vs. 4%, P -< 0.001; rifampicin 10% vs. 3%, P < 0.001;
ethambutol, 3% vs. 0.1%, P = 0.002; and multidrug resistance, 6% vs. 0.4%; P <
0.001. Among 83 HIV-pos patients with newly diagnosed PTB, the median CD4
count was 123 cells/pl; had CD4 counts <200.
Conclusions: HIV infection is common among pulmonary TB patients in
Bangkok and is associated with injecting drug use and higher levels of M. tuberculosis antimicrobial resistance. These findings underscore an urgent need
to assure that tuberculosis patients in Bangkok are offered HIV counseling and
testing and treated with adequate and complete courses of antimicrobial agents.
S13265 Characteristics of HIV-lnfected Tuberculosis Patients
Aged 20-49 years diagnosed in correctional facilities, US,
1993-1996
Cindy Weinbaum1, E. McCray2. I CDC/DTBE 1600 Clifton Road N.E. M/S E-10
Atlanta Georgia 30333; 2Centers For Disease Control And Prevention Atlanta
GA, USA
Background: Since 1993, information on HIV status and correctional facility
residence at diagnosis of tuberculosis (TB) has been reported to the US national
TB surveillance system. We describe characteristics of HIV(+) TB patients aged
20-49 years reported after TB diagnosis in a correctional facility ("inmates").
Methods: Data from the national TB surveillance system from 1993-1996
were analyzed. HIV(+) and HIV( ) inmates in the 20-49 year age group were
compared. No HIV results were reported from Rhode Island or California.
Results: From 1993-1996, 47,368 (50% of reported US TB cases) were 20-49
years old; 3,348 (7%) were inmates at TB diagnosis. Among these inmates, 1,663
(50%) had HIV status reported: 938 (56%) were HIV(+), and 725 (44%) HIV(-).
The proportion of reported HIV results for inmates aged 20-49 years increased
from 48% in 1993 to 57% in 1996 (p < 0.01). Inmates of federal and state
prisons were more likely than inmates of local jails to have HIV results reported
(51% vs. 44%) and to be HIV(+) (61% vs. 50%). New York and Texas reported
47% of inmates with HIV data but 60% of HIV(+) inmates. Compared to HIV(-)
inmates, more HIV(+) inmates were black (66% vs. 57%), US born (93% vs.
86%), and had died before being diagnosed with TB (6% vs. 1%). HIV(+) inmates
with pulmonary TB more often had normal chest radiographs (17% vs. 7%) or
noncavitary disease (84% vs. 67%). Sputum smear positivity (46% vs. 45%) and
culture positivity (81% vs. 79%) among pulmonary cases was not significantly
higher for HIV(+) inmates. Antituberculous drug resistance to at least rifampin
(9% vs. 2%) and isoniazid plus rifampin (7% vs. 2%) were more common among
HIV(+) inmates.
Conclusions: Reporting of HIV status for inmates with TB has improved, but
continues to be incomplete. Among inmates with HIV results, over half had HIV
infection. HIV(+) inmates with pulmonary TB are more difficult to diagnose because chest radiographs lack findings consistent with TB. Improved HIV reporting
would aid in directing prevention and rapid diagnostic efforts.

Methods: Data were collected from five clinics at four hospitals in NYC for
the CDC Adult Spectrum of Disease Project starting in November 1991. Medical
charts are abstracted every six months, and patients followed until death or they
do not visit the clinic for three consecutive intervals. Data was censored at Nov.
1997. Confirmed cases of MTB were ascertained by a match with the NYC AIDS
surveillance database, which receives data from the NYC Bureau of Tuberculosis
Control. Persons were eligible once CD4+ lymphocytes were <300 cells/ml, and
excluded if they had previous MTB. Incidence was calculated as the number of
first episodes of MTB divided by the MTB-free person-years of follow up. The
probability of MTB and death were calculated using the Kaplan-Meier product-limit
method; survival calculations use all-cause mortality.
Results: 1,602 patients had CD4+ < 300 cells/ml, but 154 had previous MTB
and were excluded, leaving 1,448. Median follow up was 441 days (range 1 to
2,037). 53 (4%) developed MTB, 41 pulmonary and 12 extrapulmonary, an incidence of 2.0/100 person-years. Incidence decreased from a high of 4.3 cases/100
person-years in 1992 to 2.2 in 1996, a 51% decline. The probability of MTB was
2% at one year, 5% at two years, and 9% at three. Risk of MTB was not related
significantly to sex, risk, race, clinic, or age. 36 (67%) of the 53 MTB patients are
known dead. Mortality among MTB cases was 30% at one year after diagnosis,
and 54% at two.
Conclusions: Incidence of MTB in these immunocompromised HIV-infected
persons in care in NYC clinics has fallen 51% since 1992. mirroring the 46%
decrease in MTB cases in NYC. The impact of MTB on immunocompormised
HIV-infected persons remains great, however, with incidence among the highest
reported in the United States' cities and a high mortality rate.
13267 Risk factors for tuberculosis (TB) in a rural African
population cohort with known HIV status
Judith R. Glynn', D.K. Warndroff2, S.S. Malema2, P. Nkhosa2, L. Bliss1,
A.C. Turner1, P.E.M. Fine'. 1London School of Hygiene and Tropical Med.
London, England; 2Karonga Prevention Study, Chilumba, Malawi
Background: Studies in antenatal women and intravenous drug users show high
relative risks for TB among HIV positive individuals compared to HIV negative
individuals, but may not be typical, and case control studies are subject to selection bias in the choice of controls. We had the opportunity to conduct a general
population cohort study in a rural area of Northern Malawi.
Methods: As part of the long running Karonga Prevention Study, filter paper
blood specimens were collected from all individuals seen in 2 rural areas during
a total population survey conducted between 1987-89. Socio-demographic data
were available from this and an earlier survey in the same population. These
specimens have now been tested for HIV. TB and leprosy in the district are
diagnosed by enhanced passive surveillance, with project staff based at health
centres and the hospital.
Results: Preliminary results from those aged over 14 years show that
175/10975 were HIV positive at the time of the survey. By 1995 subsequent
confirmed TB had been diagnosed in 6 (3.4%) HIV positives and 58 (0.54%)
HIV negatives (odds ratio 6.6, 95% confidence interval 2.8-15.4). In a logistic
regression analysis, the odds ratio was little changed by adjusting for age, sex.
schooling, housing conditions or area. Age over 30, increased schooling, and
better housing were all independently associated with increased risks of TB after
adjusting for HIV status and the other factors. There were no new cases of leprosy
among HIV positive individuals.
Conclusions: The relative risk of 7 for the association of TB and HIV found in
this general population cohort study was similar to that found in selected cohorts
and previous case control studies in this population and elsewhere. After adjusting for baseline HIV status, TB was more common at older ages and in higher
socio-economic groups. There was no evidence for an increased risk of leprosy
associated with HIV seropositivity.
13268 Double trouble: HIV/TB profiles from the US and the UK
derived from multiple surveillance systems
Diane Bennett1, D. Howitt2, J.L. Jones3, M. Moore1, J. Herbert2, E. McCray1,
J.M. Watson2. 1CDC MS E-10 Div of TB Elimination 1600 Clifton RD., NE,
Atlanta, GA 30333; 3CDC MS E47 Div. of AIDS, USA; 2CDSC 61 Colindale Ave.
AIDS Division London NW95EQ, UK
Objectives: to use data from multiple surveillance systems in two industrialized
countries, the United Kingdom (UK) and the United States (US) to examine the
interaction between HIV and Tuberculosis (TB).
Methods: 1993-96 data from the national US TB surveillance system and
1993-95 data from the UK TB laboratory reporting system were used to compare HIV-seropositive TB cases with other TB cases. Crude TB rates for the
HIV-seropositive populations were calculated. National AIDS data were used to
compare individuals diagnosed with TB to those diagnosed with other conditions.
Results: TB/AIDS cases were 15% of US TB cases, but only 4% of culture
positive UK TB cases. TB accounted for 5% of US and 4% of UK AIDS diagnoses.
The US TB annual case rate among people with HIV infection was approximately

333/100,000 (42 times that of the general population): in the UK the rate was
640/100,000 (116 times that of the general population). A significantly higher proportion of UK TB/AIDS cases (49%) were foreign-born than in the TB/non-AIDS
group (21%); in the US the situation was reversed (15% vs 33%). In the UK,
White individuals were predominant in both groups whereas in the US there
were three times as many Black than White individuals in the TB/AIDS group
but nearly equal numbers in the TB/non-AIDS groups. In both countries, Asian

S132661

Impact of tuberculosis on immunocompromized
HIV-infected persons in selected HIV clinics in New York
City (NYC), 1991-97

Jeffrey McFarland1, J. Sackoff2, S. Su2, T. Singh2, E. Bryan2, S. Forlenza2.
'Centers for Disease Control & Prevention, Atlanta, GA; 2New York City
Department of Health, New York, NY USA
Objective: To determine the incidence of M. tuberculosis disease (MTB) in HIVinfected persons in care at selected hospital-based clinics in NYC and to identify
demographic factors influencing incidence. To describe mortality among these
persons with MTB.
Background: In NYC, an epidemic of MTB followed the emergence of HIV,
with many people having both diseases. The reported cases of MTB peaked in
1992 and fell 46% by 1996.