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

12th World AIDS Conference

Abstracts 13296-13299 145

Methods: Men enrolled in the San Francisco Men's Health Study, a population-based sample of 1034 unmarried men ages 25 to 54 residing in a high
HIV incidence area of San Francisco, were studied at baseline in 1984-85.
Number of total male intercourse partners in the preceding 2 years, and number of male partners in the preceding 2 years with whom subjects practiced 1
of 6 sexual practices: either insertive or receptive oral-genital (O-G), oral-anal
(O-A), or anal-genital (A-G) sex were measured by interview. HHV-8 seropositivity
was measured by an indirect immunofluorescent assay for antibodies to HHV-8
latency-associated nuclear antigens (anti-LANA).
Results: Anti-LANA antibodies were present in 261 out of 1015 (26%) available samples. In unadjusted analyses, seropositivity was directly correlated with
total number of preceding 2-year male intercourse partners (p < 0.001) and with
number of partners with whom the subject practiced either insertive or receptive
O-G, O-A, or A-G sex (p - 0.001 for all 6 acts). To estimate the independent
effect of O-G sex, only those who denied any form of O-A or A-G sex in the
prior 2 years were considered: 6 of 34 (18%) who practiced any form of O-G sex
were anti-LANA seropositive compared to 1 of 212 (0.5%) who did not practice
O-G sex (p - 0.001). Tight collinearity, however, between sexual practices precluded estimation of the independent effects of the other acts by either analogous
restricted analyses, stratification or modeling.
Conclusions: Oral-genital intercourse was associated with HHV-8 infection in
this cross-sectional study. Although this finding may be explained by other sexual
acts practiced prior to 2 years before entry into the study, it is consistent with the
ability to detect HHV-8 in saliva. Close correlations between the types of sexual
acts men practice may make simple unadjusted analyses misleading. Longitudinal studies of HHV-8 seroconversion will be needed to confirm the hypothesis of
oral transmission and define its relative importance.
13296 Hepatitis C prevalence and incidence in Vancouver IDUs
during an outbreak of HIV infection
DavidMichael Patrick, P.G.A. Cornelisse, C.H. Sherlock, M.L. Rekart,
S.A. Strathdee, M.T. Schechter, M.V. O'Shaugnessy. Vancouver, B.C. V5Z4R4
BCCDCS, STD/AIDS Control, 655 W. 12th Ave., Canda
Objectives: To determine predictors of Hepatitis C Virus (HCV) prevalence and
HCV incidence in a cohort of injection drug users (IDUs) during a documented
outbreak of HIV infection.
Design: Cohort study.
Methods: IDUs recruited to the Vancouver IDU Study completed intervieweradministered questionnaires and testing for HCV and HIV at baseline and semiannually. Incidence density (ID) was calculated. Logistic regression was used to
identify independent predictors of HCV seropositivity at baseline.
Results: As of Sept 97, 1,080 had been recruited. Baseline HCV prevalence
was 85% (95%CI:83.9-87.1). Of 172 IDUs who were HCV negative at baseline,
there were 23 documented seroconversions after a median 9.8 months of follow-up [ID = 26.2 per 100 person years (95%CI: 15.5, 40.0)]. HIV was associated
with baseline HCV infection [OR = 6.0;(95%CI: 3.3-11.2)] but small cell size
(HCV /HIV+) precluded its inclusion in a multivariate model. Independent factors
associated with HCV prevalence at baseline were female gender [AOR = 2.2;
(95%CI: 1.4-3.6)], ever being incarcerated [AOR = 2.4; (1.5-3.7)], ever attending
a needle exchange program (NEP) [AOR = 2.6; (1.4-4.6)], ever borrowing syringes [AOR = 1.7; (1.1-2.6)], age [AOR = 1.1 per year; (1.08-1.16)] and duration
of injection drug use [AOR = 1.4 per year; (1.2-1.6)]. There was a significant
interaction between age and duration of injection drug use (p = 0.0001). The
strength of association between HCV infection and duration of injection drug use
increased with age. Risk factors for seroconversion will be discussed.
Conclusions: HCV infection approaches saturation in IDUs and is associated
with duration of injection drug use, female gender, incarceration and past NEP
use. As NEP attracts high risk IDUs, the latter association is unlikely to be causal.
Harm reduction in prison, protection of women and primary prevention of injection
drug use all require continued emphasis.
13297 The association between Plasmodium falciparum malaria
illness and HIV-1 RNA viral burden
Irving Hoffman1, T.E. Taylor2, C. Jere3, J.J. Wirima3, M.E. Molyneux4,
J.R. Dyer5, M.S. Cohen6, P. Munthali3, N. Kumwenda7, S.A. Fiscus6,
H. Chakraborty6, TE. Taha7, J.J. Eron6. 'UNC-Chapel Hill Div of ID CB#7030,
Chapel Hill, NC 27599; 2Michigan State University, Lansing, MI; 6Univ. of North
Carolina, NC; 7Johns Hopkins University, Baltimore, MD, USA; 3Univ. of Malawi
College of Medicine, Blantyre, 4 Well Come Trust Centre, Blantyre, Malawi;
5 Towns Ville General Hospital, Queensland, Australia
Objectives: To determine the association between uncomplicated P falciparum
malaria illness in adult Malawians and blood plasma HIV-1 RNA viral burden.
Methods: Malawian adults with P falciparum parasitemia and symptoms consistent with malaria illness (the malaria group) and asymptomatic, aparasitemic
blood donors (the control group) were identified to be HIV infected by repeat
ELISA. Subjects with historic or physical evidence of TB, STD, AIDS, or other

acute infections were excluded. The malaria group received appropriate antimalarial chemotherapy and were followed with sequential blood films. In both
groups, blood plasma HIV-1 RNA viral loads were determined by NucliSens
(Organon-Teknika, Durham, NC) at baseline, and again at 1, 2, and 4 weeks.
Results: Baseline and follow-up data are currently available for 20 malaria
subjects and 24 controls. The median CD4+ counts in the malaria group (240
cells/ul) and the control group (195 cells/ul) were comparable (p =.39). At base
line, the median HIV-1 RNA levels in the malaria and control groups were 155,000
and 24,500 copies/ml, respectively (p <.0003) and at last follow-up (medians 25
and 28 days) the levels were 140,000 and 29,000 copies/mI. (p =.02). Following
treatment, individuals in the malaria group showed a significant decrease in viral
load (median =.20 logio, p =.04). By comparison, no significant changes were
noted at follow-up among individuals in the control group (median = +.08 loglo, p
=.69).
Conclusions: This study suggests that either malarial infection increases HIV
burden or alternatively, a high viral burden increases susceptibility to malarial
infection. Because of the modest response of HIV burden to successful malarial
therapy, a prospective, longitudinal study will be required to resolve this question.
If malarial infection increases HIV viral burden, it would be expected to accelerate
HIV disease progression and facilitate HIV transmission and help to explain the
dynamics of the HIV epidemic in sub-Saharan Africa.
13298 HIV and malaria overlap and do interact in sub-Saharan
African pregnant women
Richard Sreketee1, Bernar D. Nahlen2, J. Ayisi3, A. Van Eijk3, A. Misore3. 11600
Clifton Road N.E. E46 (CDC) Atlanta Georgia 30333; 2CDC DHAP-SE Atlanta
GA, USA; 3Kenya Medical Research Institute Kisumu, Kenya
Background: Documented interaction between HIV and malaria has been limited
to a link between malarial anemia, blood transfusion, and HIV transmission. One
study in Malawi showed higher malaria rates in multigravid HIV(+) women. We
evaluated the HIV and malaria in pregnant women in Kenya.
Methods: Between July 1996 and September 1997, pregnant women attending
an antenatal clinic or delivery unit in Kisumu, western Kenya were offered counseling and voluntary HIV testing and were examined for peripheral and placental
blood parasitemia.
Results: Among 411 (25.6%) HIV(+) and 1196 (74.4%) HIV(-) women, the
prevalence of peripheral parasitemia at enrollment and delivery and placental
parasitemia was higher in HIV(+) women in all gravidities; parasite densities
were higher in HIV(+) women, p <.01. Parasite prevalence decreased between
enrollment and delivery overall (20.5% to 15.8%, respectively) within gravidity
groups.

Gravidity
G1
G2
-G3
Total

HIV(+)
% placental para (n)
34.0 (156)
23.0 (122)
21.1 (133)
26.5 (411)

HIV( -)
% placental para (n)
20.5 (546)
13.9 (288)
6.4 (362)
14.6(1196)

Total
% (n)
23.5 (702)
16.6 (410)
10.3 (495)
17.7 (1607)

OR
(95%Cl)
1.7 (1.3-2.2)
1.5 (1.1-2.1)
2.3 (1.7-3.1)
17(1.4-2.0)

'10"
0.02
- 10 5
- 10-5

Conclusions: Compared with HIV(-) pregnant women, HIV(+) pregnant
women in all gravidities had more and higher density malaria infections. HIV
infection appears to reduce a pregnant woman's ability to control malaria infection. A review of the literature suggests that cross-sectional studies in other
populations (children and adults) may have biases making them unlikely to
demonstrate an HIV-malaria interaction, Further studies are warranted because
of the vast infection overlap in sub-Saharan Africa.
S13299 Effect of HIV infection on Schistosoma mansoni infection
among sugar estate residents in Ethiopia
Arnaud Fontanet1, T. Woldemichael2, T. Sahlu1, T. Rinke De Wit1, H. Yeneneht,
R.A. Coutinho3, L. Van Lieshout4. Ethiopian Netherlands AIDS Research
Project/ENARP Addis Abeba; 2EHNRI, Addis Abeba, Ethiopia, 3Municipal
Health Services, Amsterdam; 4University of Leiden, Leiden, Netherlands
Objective: to study the effect of HIV infection on S.mansoni infection intensity in
co-infected individuals.
Design: Community-based cross-sectional survey.
Methods: 1,239 randomly selected individuals aged 15-54 years living in a
sugar estate of central Ethiopia were enrolled in a survey of HIV infection and
parasitic diseases from November 1995 to August 1996. Stool was examined for
S.mansoni eggs (Kato smear) and serum was tested for HIV antibodies (ELISA
and Western blot). Urine from 287 study participants was tested for the concentration of the S.mansoni gut-associated circulating cathodic antigens (CCA).
which are known to be correlated with S.mansoni worm load.
Results: S.mansoni and HIV prevalence were 31.4% and 3.1% respectively.
The two epidemics tended to cluster into different populations of the estate: the
schistosomiasis epidemic was predominant in the camps, and primarily affected
people working in the field, whereas the HIV epidemic was found in the main
village. In participants infected with S.mansoni, the geometric mean egg count
was lower in HIV-positives when compared to HIV-negatives (118 vs 159 eggs/g.
p = 0.05), and HIV status remained a predictor of low egg count after controlling
for the other predictors of egg count (age, residence, and duration of work) in a
linear regression model. On the other hand, urine CCA levels did not differ between HIV-positive and HIV-negative individuals, suggesting that the decreased
egg output in HIV-positive might be related to reduced egg excretion (a CD4+
T-cells-dependent phenomenon).
Conclusion: In this study, S.mansoni egg counts, but not CCA concentrations,
were decreased in HIV-positive individuals when compared to HIV-negatives.
The potential liver damages associated with egg retention in HIV-positives, and
the effect of schistosomiasis on HIV infection progression will be studied in a
prospective study of co-infected individuals.