Final Program and Oral Abstracts [International Conference on AIDS (8th: 1992: Amsterdam, Netherlands)]

TRACK B: CLINICAL SCIENCE AND CARE

TuB 0533-TuB 0538

TuB 0533 MRALITY IN HIV-ASSOCIATED TUBEROJIS TREAED
S WITH TOI DIFERET DRUG REGIMENS. Nunn P.*.**,***, Odhiao Jose*, Githui W*,
Gathua S**, Wasunna K*, Eorris J***,EMAdam K***; *Kenya Medicl Research
Institute, Nairobi, Kenya; **Infectous Diseases Hospital, Nairobi, Kenya;
***lIndon School of Hygiene & Tropical Medicine, London, U.K;
Objective: To determine and compare mortality rates and causes among HIV-1 infected
TB patients treated with either standard or short course drug regimens.
Methods: 107 HIV+ and 174 HIV- TB patients were enrolled in a longitudinal cohort
study and followed monthly in Nairobi, Kenya. Treatment was with either the standard
regimen (Thiacetazone and Isoniazid daily for 12 months supplemented by Streptomycin
in the first month only - STH) or short course regimen (Rifampicin, Pyrazinamide,
Isoniazid daily for 2 months with Streptomycin in the first month, followed by
Thiacetazone & Isoniazid for 6 months (SHRTZ). Causes of mortality were based on
standard approach to morbid events.
Results: 35 HIV+ compared to 13 HIV- patients died before the end of therapy
(Unadjusted relative risk, RR=4.4, 95% confidence interval, CI 2.4 - 7.9). There was
an almost four-fold excess in mortality among HIV+ compared to HIV- patients within
6 months of start of anti-tuberculous therapy (Rate ratio = 3.8, 95% confidence
interval 1.7-8.1, P<0.001, adjusted for age, sex and education). Probability of
survival at one year after start of treatment was 66% and 92% for HIV+ and HIVpatients respectively (log rank test p< 10-6). Among HIV+ patients only significantly fewer deaths (p=0.003, adjusted for age, sex and education) occured with the
SHRTZ regimen in the first 6 months but this advantage had disappeared by 1 year.
Most of these deaths were due to non-tuberculous non-AIDS defining bacteraemias
due to Salmonella typhimurium and Staphylococcus aureus. Conclusions: Mortality
during anti-TB therapy is significantly higher among HIV+ compared to HIV- patients
but does not appear to be due to failure of anti-TB therapy. The excess mortality
among HIV+ patients is due to causes other than TB. Rifampicin containing regimen
(SHRTZ) is protective initially, probably by preventing other bacterial infections.
DR. JOSEPH ODHIAMBO, P.O. BOX 47855, NAIROBI-KENYA
TEL. [phone]
FAX. [phone]
TuB 0535 MULTI-DRUG RESISTANT TUBERCULOSIS
(MDR-TB) IN PATIENTS INFECTED WITH HIV
Busillo, C.*; Mullen, Michael,*; Soumakis, S.*; Lessnau,
K.*; Sanjana, V.*; Davidson, M.*; Talavera, W.*
*Cabrini Medical Center, New York, N.Y. 10003
Outbreaks of MDR-TB in patients infected with HIV have recently been described. We report an increased rate of MDR-TB in
patients hospitalized at our medical center with 24 of 72 patients with
TB between 12/90 and 5/91 as compared with 8 of 132 patients from 1982 -1987. Nineteen cases of ii DR-TB were identified in patients with HIV
infection; 13 of whom were previously diagnosed with AIDS. All were in
high risk groups, 10 homosexual men and 9 IVDU's. Ten had pulmonary involvement, 4 had extra pulmonary disease and 5 patients had both. TB
cultures from 10 out of 19 patients were resistant to 3 or more drugs.
Two of the 19 patients had secondary resistance. Fifteen patients
expired, although 9 of these received 4 or more drugs for a mean time
of 7 weeks. Significant markers of MDR-TB were continued fever,
worsening lung infiltrates, continued positive cultures and extrapulmonary spread while on therapy with a 4 drug regimen. Analysis of
restriction-fragment-length polymorphisms revealed that 14 of 15
strains were identical, suggestive of nosocomial spread. In conclusion
we have seen a significant increase of MDR-TB in HIV-infected patients
admitted to our hospital during this 17 month period. Multiple factors
probably contributed to the increased mortality in these patients;
including rapid progression of clinical TB in the immuno-compromised,
inadequate antimicrobial coverage when a standard drug regimen was
used and the time required for organism identification. New strategies
for diagnosis and empirical therapy of patients with suspected MDR-TB
need to be developed.
Michael Mullen, M.D.
227 East 19th Street
New York, N.Y. 10003 [phone]
TuB 0537 Tuberculosis in infants born to HIV-infected
mothers: a survey in a French obstetrical unit
Firtion. Ghislaine*; Gendrel, D.**; Badoual, J.**; Krivine, A.***;
Henrion, R.*; Mandelbrot, L.*; B6douet, J.****. Paris, France.
Objectives: A retrospective study was performed to evaluate risk factors for tuberculosis in
families with HIV-infected mothers.
Methods: Between July 1, 1986 and December 31, 1991, 162 infants born tol 50 HIV-infected
mothers were followed at the pediatric unit of our hospital.
Results: Tuberculosis was diagnosed (clinical findings, X.Ray, microbiology) in 8 families (5
mothers, 3 fathers) concerning 11 infants. Tuberculosis was observed in 3 children, 2 of whom
were HIV-infected (aged 21 and 29 months) and one uninfected (1.5 months). Geographic origin
of these families was Central Africa (6/8), North Africa (1/8), Caribean (1/8). This
distribution differed strongly from that of our overall HIV-infected population (12.5% Central
Africa, 11.2% North Africa, 3.3% Caribean). Among 3 cases of maternal tuberculosis at the
time of delivery, one newborn was infected.
Conclusions: Our data show that infants born to HIV-infected mothers are at risk for
tuberculosis, in particular in families from Africa. Therefore, screening for tuberculosis
should be systematic among such parents and infants. Mycobacterium tuberculosis stains should
be tested for resistance. We suggest prevention by BCG as early as possible for HIV-uninfected
exposed infants.
*Maternite Port Royal, Cochin; **Departement de Pediatrie, St. Vincent de Paul;
***Laboratoire de Microbiologie, St. Vincent de Paul; ****D6partement de M6decine Interne,
Cochin
Dr. FIRTION, matemrnite Port-Royal 123 bd Port-Royal 75014 PARIS
FRANCE - Te1.: 42-34-12-27 Fax: [phone]

TuB 0534 OUTBREAK OF MULTIPLE DRUG RESISTANT TUBERCULOSIS (MDR-TB) AMONG
PATIENTS WITH HIV INFECTION. Fischl. Margaret; Uttamchandani, R;
Daikos, G; Poblete R; Moreno J; Lai S. Univ. of Miami, Miami, USA.
Objectives: To evaluate an outbreak of MDR-TB in patients with HIV infection.
Methods: Retrospective cohort study of TB among patients with HIV infection since
1988 and a nested case control study to evaluate risk factors for MDR-TB.
Results: Sixty-two cases with MDR-TB and 55 controls with susceptible-TB were
identified from January 1988 to December 1990. Forty seven cases had AIDS (75.8%);
9, ARC (14.5%); and 6, no symptoms (9.7%). Sixteen isolates were resistant to 2
drugs; 22, to 3 drugs; 22, to 4 drugs; and 2, to 5 drugs. Sixteen controls had AIDS
(29.1%); 17, ARC (30.9%); and 22, no symptoms (40%). The median time interval
between AIDS and TB was 224.3 days for cases and 33.9 days for controls (P-0.007).
Controls were more likely to be black (P-0.03) or Haitian (P-0.005). Cases were more
likely to be homosexual men (P-0.008), to have had AIDS (P-0.008), to have been
admitted to an HIV ward (P-0.002), to have been seen in an HIV clinic (P<0.0001), or
to have received inhaled pentamidine (P-0.002) or intravenous therapy in an HIV
clinic (P<0.0001). Multivariate logistic regression analysis showed that AIDS
(P-0.0002) and the HIV clinic visits (P-0.002) were independently associated with
MDR-TB. Cases were also more likely to have a dry cough (P-0.01), a shorter interval
between cough and presentation (P-0.006), a chest radiograph that showed alveolar
infiltrates (P-0.03) and cavities (P-0.04), and were less likely to have
lymphadenopathy (P-0.03), resolution of fever (P-0.007) and interstitial infiltrates
(P-0.01). Acid fast smears and cultures were intermittently or persistently positive
among cases. The median survival after a diagnosis of TB was 1.6 months for cases
and 17.9 months for controls (P-0.0001). Treatment regimens with 5 or more drugs
were associated with better survival (0.0003).
Conclusions: Patients with AIDS are susceptible to the acquisition of drug resistant
M. tuberculosis with rapid progression to disease and death. Crowded conditions,
such as waiting rooms in HIV clinics, may facilitate nosocomial transmission.
Margaret A. Fischl, M.D.; University of Miami School of
Medicine Department of Medicine R-60A, POB 016960 Miami,
Fl., 33161. Phone: [phone]. Fax: [phone]
TuB 0536 PREVENTIVE TUBERCULOSIS CHEMOTHERAPY WITH ISONIAZIDE AMONG PERSONS
INFECTED WITH HIV-1. Wadhawan Devinder*,Hira,S*/**,Mwansa,N*,
Perine,P**. *University Teaching Hospital,Zambia.**USUHS,Bethesda.
Objectives. 1. To determine the rate of development of active tuberculosis (TB) among
HIV-infected persons; 2. To assess efficacy of isoniazide (INH) chemoprophylaxis in
preventing development of active TB.
Methods. A randomized, single-blinded, controlled study is underway at the University
Teaching Hospital in Lusaka since September 1988. Nonpregnant adults with western
blot confirmed HIV-1 infection and with absence of active tuberculosis as determined
by symptoms, signs and chest radiographs were recruited. Group 1. 352 patients
received INH 300mg by mouth daily for 6months. Group 2. 297 patients received B Co
tablets daily. Patients in both groups were examined quarterly and chest radiographs
were done every 6 months. Active TB was subsequently diagnosed by radiograph, sputum
culture or tissue biopsy.
Results. Patients in both groups were comparable by age, sex and staging of HIV
disease as done by Walter Reed Stages. Group 1. 298 patients were followed for 413
person years (p-y) and 7 developed active TB. Group 2. 246 patients were followed for
362 p-y and 23 developed active TB. All those who developed active TB were in WRIII
or WRIV stages at the time of recruitment. Hence, stratifying patients in WRIII to
WRIV, there were 193 in group 1 and 167 patients in group 2. The annual incidence of
TB in stratified group of patients in WRIII and WRIV was 2.6/100 p-y (7/268 p-y) in
group 1 (INH) and 11.3/100 p-y (23/203 p-y) in group 2 (control) (p <.001).
Conclusion. INH prophylaxis significantly reduced the incidence of active TB. However,
the rate of active TB increased with the post-prophylaxis interval. The activation
of mycobacterial infection was seen primarily among patients in WRIII and WRIV as
compared with those in WRI and WRII.
Wadhawan, Devinder. Department of Medicine, University Teaching
Hospital, P. 0. Box 50001, Lusaka, Zambia, Africa.Fax[phone]
TuB 0538 HIV-I INFECTION AND RECURRENCE OF TUBERCULOSIS, NAIROBI, KENYA.
Hawken Mark*, Nunn P***, Gathua S**, Godfrey-Faussett, P***, Brindle R**** Githui W*,
Odhiambo J*, Gilks C, Morris J***, McAdam K***. *Kenya Medical Research Institute,
Nairobi, Kenya. "Infectious Diseases Hospital, Nairobi, Kenya. ***London School of Hygiene
and Tropical Medicine, London, UK. ***"Public Health Laboratory, Oxford, UK.
Objective: To compare the recurrence rate of tuberculosis among HIV-1 positive and negative patients.
Methods: 63 HIV-1 positive and 136 HIV-1 negative patients completed treatment with either thiacetazone and
isoniazid daily for 12 months supplemented by streptomycin for the first month only, or daily rifampicin,
isoniazid and pyrazinamide, for 2 months with streptomycin in the first month, followed by thiacetazone and
isoniazid for 6 months. Al patients were actively followed. DNA finger printing was performed on initial and
recurrent isolates from 2 patients.
Results: Recurrence, defined as a positive culture for M. tuberculosis on at least 2 occasions, occurred in 8/63
(16/100 PYO) HIV-1 positive and 1/136 (0.7/100 PYO) HIV-1 negative patients (RR 8.7 95% CI 1.9-40). 7/8
recurrences among the HIV-1 positive group were associated with substitution of ethambutol for thiaceazone
because of a cutaneous hypersensitivity reaction to thiacetazone (p<l5). Risk factors for relapse included
younger age, lower initial haemoglobin level and total white cell count, but did not include initial resistance,
initial treatment regimen, WHO definition of AIDS on entry, or poor compliance. During the observation period
12/63 (19%) of the HIV-1 positive group died compared to 1/136 HIV-1 (1%) negatives. Loss to folblow up was
8/63 (13%) and 19/136 (14%) respectively. One patient had an identical DNA finger print pattern in her initial
and recurrent isolates; one patient had a different pattern.
Conclusion: Recurrence of tuberculosis is associated with HIV-1 infection. It could be due either to relapse of
the original infection or a second infection. DNA finger print analysis has shown both mechanisms are possible.
The relative importance of each is not yet known. Recurrence appears to be associated with interruption or
insufficient duration of therapy, or the use of ethambutol, or a combination of these.
Mark Hawken, Kenya Medical Research Institute, PO Box 20778, Nairobi, Kenya.
Tel: [phone]. Fax [phone].

Tu33