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

12th World AIDS Conference

Abstracts 22175-22178 297

the primary risk factor for opportunistic infections with MAC, that are commonly
associated with a poor prognosis.
Methods: Reported here is a retrospective analysis of the cases of disseminated MAC infection (dMAC) diagnosed among the 53 HIV-infected pediatric
patients followed in Institute of Infectious Diseases-University of Pavia, including
the clinical characteristics, laboratory correlates and therapeutic strategies.
Results: dMCA infection was diagnosed in 5/53 children (4 M, 1F), who were
classified in C3 category, according to 1994 CDC criteria. In all of them the
CD4+ lymphocyte count resulted <50/mm3 and a decreased haemoglobin level
was detected (min. 6.2 g/dl, max. 9.5g/dl). With regard to clinical manifestations
of infection, fever was always present, associated with diarrhea and significant
weight loss in 3 cases. Mycobacteria were isolated from blood cultures in all
children and in one case also from stool specimens.
In two patients the diagnosis was too late and any antimycobacterial treatment
couldn't be started; in the other three cases the therapeutic strategies (clarithromycin + ethambutol + rifabutin and clarithromycin + ethambutol+ ciprofloxacin
+ amikacin) provided only a symptomatic relief and a temporary improvement in
microbiologic indicators. All children died, with a median survival of 6.8 months
after the first MAC isolation.
Conclusion: Because children with HIV infection survive longer despite serious immune impairment, it becomes increasingly important to identify children at
highest risk of MAC infection and to individuate the most efficacy regimens of
primary chemoprophylaxis.
22175 Is MAC prophylaxis necessary in HIV positive individuals
responding to antiretroviral therapy?
Jas Gill, G. Moyle, B.S. Gazzard, M.R. Nelson. Chelsea and Westminster
Hospital, 369 Fulman Road, London, England, UK
Aim: To study whether MAC prophylaxis should be based on nadir of CD4
count, or whether it is safe to stop preventative therapy in patients responding to
antiretroviral therapy (ART).
Method: MAC prophylaxis is routinely offered to patients with a CD4 < 50. MAC
prophylaxis was stopped in patients responding to ART with two consecutive CD4
counts >50, with fully informed consent.
Results: 13 patients discontinued MAC prophylaxis - 8 rifabutin and 5
azithromycin. Median duration of MAC prophylaxis was 420 days. CD4 count
at discontinuation of prophylaxis ranged between 51 and 208 cells/ml (median
126). After a median of 231 days follow up (range 131-883 d, mean 391 d),
no patient has developed clinical or microbiological MAC disease. Assuming the
13 patients had similar characteristics to the original rifabutin MAC prophylaxis
studies, one would have expected 2.5 episodes of MAC infection off prophylaxis
during this time period.
Conclusion: Our series suggests that it is reasonable to stop MAC prophylaxis
in patients increasing CD4 to above 50 on ART. Controlled trials of discontinuation
of chemoprophylaxis are needed.
22176 A placebo-controlled trial of rifabutin added to a regimen
of clarithromycin and ethambutol in the treatment of M.
avium complex (MAC) bacteremia
Fred Gordin', P. Sullam2, S. Shafran3, D. Cohn4, C.R. Horsburgh5. 1VA Medical
Ctr (151B), 50 Irving St NW Washington, DC; 2University of California, San
Francisco; 3 University of Edmonton, Alberta; 4 University of Colorado, Denver;
5 Emory University, Atlanta, USA
Objective: To evaluate the efficacy of clarithromycin (CLA) + ethambutol (EMB)
vs CLA + EMB + rifabutin (RBT) in the treatment of HIV-postive (HIV+) patients
(pts) with AIDS and MAC bacteremia.
Methods: HIV+ pts with MAC bacteremia and >2 associated symptoms (fever,
sweats, fatigue, diarrhea, abdominal pain, weight loss) were randomized to receive CLA (500/mg BID), EMB (-15 mg/kg/day) and either RBT (300 mg QD) or
placebo. Pts had 2 blood cultures at baseline and at weeks 8 and 16, and single
blood cultures at weeks 2, 4, 12, 20 and every 4 weeks thereafter. A bacteriologic
response (BR) was defined as either a >2 log decrease in MAC colony forming
units (cfu) or sterilization of blood on 2 consecutive cultures.
Results: From Dec '93 to Sept '96, 198 patients were enrolled from 47 sites in
the United States, Canada, and Mexico. The mean age was 36 years, 87% were
male and 69% Caucasian; mean CD4 cell count was 20.
Blood culture results:

S22177 Molecular and clinical evidence for Pneumocystis carinii
strains of low pathogenicity from Africa compared to the
US-Europe-Australia
Sten H. Vermund2, S. Lucas4, M.S. Bartlett3, X. Tang3, M.P. Bhatta1,
J.W. Smith3, C.H. Lee3. 1University of Alabama; 2Dept Epi UAB, Rm 203 Tidwell
Hall, 720 South 20th St., Birmingham, AL; 3Indiana University, Indianapolis, IN,
USA; 4St. Thomas Hospital, London, UK
Objectives: To investigate whether genetic characteristics of P carinii might
explain the marked differences in frequency of this infection.
Design: Cross-sectional; Pathology specimens; Molecular probes for P carinii.
Methods: Tissue blocks of P carinii-infected lung and bronchioalveolar lavage
specimens were tested by polymerase chain reaction (PCR) for unique genotypes
of internal transcribed spacer regions (ITS). Two loci were typed: ITS1 between
18 S and 5.8 S and the ITS2 between the 5.8 and the 26 S regional subunits of
ribosomal RNA. This typing system has permitted the identification of 59 different genotypes of P carinii with 15 different ITS1 and 14 different ITS2 genetic
patterns noted in over 200 specimens from 4 continents. Since P carinii disease
is reported rarely from tropical Africa and Asia, the organisms' genotypes were
compared to those derived from temperate Western locations where clinical P
carinii pneumonia is common in HIV-infected persons. Gomori methenamine-silver nitrate (GMS) aldehyde stains were used for morphologic characterization;
these typically stain cyst walls.
Results: P carinii-infected lung specimens from 20 HIV positive pediatric and
adult patients in Abidjan, C6te d'lvoire, were examined by PCR and nucleotide
sequencing for types of ITSs. P carinii organisms in 9 specimens were substantially different from over 200 Western samples examined previously, in that they
had large deletions of the ITS1 region. P carinii organisms in all but one of the
20 African specimens also were different morphologically in GMS stained lung
sections compared with the US. Large masses of darkly stained, network-like
structures were observed in infected lungs, and P carinii cysts were seen embedded in these masses. This morphology is normally seen only in P carinii-infected
mouse lungs.
Conclusions: The origin of different frequencies of P carinii incidence among
HIV-infected persons may be due to strain differences. This observation may give
insight into P carinii vaccine development, inspire novel diagnostic approaches,
and permit expanded molecular epidemiology studies of transmission.
22178 Atovaquone suspension (ATQ) for prophylaxis of
Pneumocystis carinii pneumonia (PCP): Effects of
baseline prophylaxis on safety and efficacy
Paul Caldwell', Robert Murphy2, C. Chan3, T. Yurik4, J. Scott', W. El-Sadr5.
'Glaxo Wellcome Inc., PO Box 13398, Research Triangle Park, NC;
2Northwestern Univ. Medical School, Chicago, IL; 4Univ Minn School of Public
Health, Minneapolis, MN; 5Harlem Hospital Center, New York, NY USA;
3Toronto Hospital, Toronto, ON, Canada
Background: The impact of prior drug experience on the outcome of clinical trials
with antiretroviral drugs is well known. The role of prior experience with drugs
for PCP prophylaxis is less well appreciated. We report results from two studies
of ATQ for PCP prophylaxis in which the effects of baseline use of dapsone
(D) or aerosolized pentamidine (AP) were evaluated. Atovaquone Suspension
(MEPRONW, WELLVONE'")
Methods: Study 1 (CPCRA034/ACTG277) compared ATQ, 1500 mg daily, with
D, 100 mg daily, in 1057 patients. Study 2 (GW213) compared two doses of ATQ,
750 mg or 1500 mg daily, with AP, 300 mg monthly, in 549 patients. Both were
randomized, open-label studies in HIV-infected patients with standard indications
for PCP prophylaxis and intolerance to cotrimoxazole. Follow-up was for 24 and
11 months, respectively. Data were analyzed using a Cox proportional hazards
model and the log rank test, both with design stratification variables.
Results: In study 1, baseline D or ATQ was used in 52% and 3% of patients,
respectively. Median months of randomized therapy was similar for all patients
(ATQ = 7.2, D = 7.4) but was longer for ATQ (7.5) than for D (3.9) for patients
taking neither ATQ nor D at baseline. There was no significant difference in rates
of PCP between treatment groups for all patients (RR = 0.85, p = 0.20), but among
patients taking neither ATQ nor D at baseline, the rate of PCP was lower in the
ATQ group (RR = 0.67, p = 0.05). In study 2, baseline AP or ATQ was used in 50%
and 2% of patients, respectively. Median months of randomized therapy (ATQ750
= 6.2, ATQ1500 = 6.0, AP = 7.8) was similar regardless of baseline prophylaxis.
In study 2, the rates of PCP were similar among treatment groups for all patients
and for those taking neither study drug at baseline.
Conclusions: In study 1 baseline use of dapsone pre-selected for patients who
are better able to tolerate its continued use. For patients taking neither dapsone
nor ATQ at baseline, ATQ was better tolerated and may be associated with better
efficacy for PCP prophylaxis. In study 2 baseline use of aerosolized pentamidine
had no significant effect on outcomes. For patients intolerant to cotrimoxazole,
ATQ has safety and efficacy similar to aerosolized pentamidine and may provide
advantages over the use of dapsone for some patients.

Baseline cfu (median)
Week 8 BR
Week 16 BR

CLA + EMB + RBT (n = 102) CLA + EMB (n = 96)
18 24
46 /80 (58%) 47 /81 (58%)
No growth: 38/46 No growth: 36/47
44 /70 (63%) 42 /69 (61%)
No growth: 40/44 No growth: 37/72

p
0.194
0.95
0.81

Bacteriologic relapses occurred in 3 patients in each group. Abatement of
clinical symptoms was similar in both groups, and there was no difference in
survival.
Conclusion: RBT did not add to the efficacy of this CLA + EMB regimen. This
may have been due to drug interactions that produced a lower level of CLA in
patients receiving RBT, or it may have been due to the dose of RBT chosen. RBT
may be of value in treatment of MAC in regimens containing other macrolides or
no macrolides.