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

122 Abstracts 13183-13186

12th World AIDS Conference

baseline to assess the (future) impact of newly available pre-AIDS combination
therapies.
13183 Combined effect of CCR2, CCR5 and SDF1 chemokine
variants on clinical and immunological disease
progression in 410 HIV infected seroconverters
Jean-Baptiste Hubert1, loannis Theodorov2, L. Debre2, N. Nagierowska2,
C. Rouzioux'3, J.F. Delfraissy1, L. Neyer2. 1Inserm V292 Hopital de Bicentre 82
Rue, Di Gal Leclerc 94276 Krenlin-Bicetre; 2Hopital Pitie Salpetriere URA CNRS
G25 Paris; 3Hopital Necker Paris, France
Background: The objective was to determine the influence of the CCR2-641 and
SDF1-3'A genotypes on HIV disease progression in a cohort of 410 Caucasian
seroconverters already genotyped for the CCR5 A32 deletion (Meyer, AIDS
1997).
Methods: 410 HIV-infected Caucasian subjects followed in the French SEROCO cohort between 1988 and 1996 were studied. These patients had a known
date of infection and were included in the cohort within 2 years after infection
(median 7 months). Median follow-up since infection was 74 months. Progression
to CD4+ cells <200/il (n = 133), clinical AIDS (1993 European def., n = 105) and
death (n = 73) was studied. Adjustment for known risk factors of progression and
early serum viral load (during the 6 to 24 month-period after infection) was made
using a Cox model.
Results: Respectively 65 and 4 patients were heterozygous (CCR2 wt/641) and
homozygous (CCR2 641/641) for the CCR2-641 mutation; 26 patients were homozygous for the SDF1-3'A polymorphism (SDF 3'A/3'A). The allele frequencies
were similar to those described in Caucasian North-American patients (Smith,
Science 1997, Michael, Nat Med 1997, Winkler, Science 1998). The CCR5 A32
deletion tended to be more frequent in SDF 3'A/3A patients than in the remaining
(30.8% vs 15.9%, p = 0.06). A protective effect of at least one CCR2-641 allele (vs
CCR2 wt/wt) was observed for progression to CD4 < 200 (logrank test p = 0.03,
crude Relative Risk (RR): 0.57; [95%Cl: 0.33-0.97]. This protective effect persisted after adjustment for early viral load, age at infection, symptomatic primary
infection, CCR5 and SDF genotypes. On the other hand, no protective effect was
observed for progression to clinical AIDS or death. Progression was similar in
the SDF1 wt/wt and SDF1 wt/3'A. The 26 homozygous patients SDF1 3'A/3'A
tended to progress less rapidly to CD4 < 200 and death than the others (RR
of respectively 0.50 [0.22-1.14], p = 0.09 and 0.52 [0.16-1.64], p = 0.26). This
protective effect was independent of CCR2 or CCR5 genotype, but was slightly
reduced after adjustment for early viral load. When combining the different genotypes, individuals with at least one "protective variant" progressed significantly
less rapidly to CD4 < 200 (RR = 0.58), clinical AIDS (RR = 0.61), and death (0.68)
than the other patients (CCR5 wt/wt CCR2 wt/wt SDF wt/wt or SDF wt/3'A).
Conclusion: These results confirm the importance of both CCR2-641 and SDF
3'A/3'A on HIV disease progression in a European population. The fact that the
protective effect on CD4 < 200 is not reduced after adjustment for early viral load
for the CCR2 mutants is surprising and suggests that the CCR2 genotype is an
important predictive factor which should also be considered besides early viral
load (as measured by PCR-RNA).
S13184 1 Protective effect of CCR-5A32 deletion on cerebral or
extracerebral toxoplasmosis occurrence in a cohort of
1489 HIV-infected patients
Laurence Meyer1, J.F. Delfraissy1, F. Sanson2, N. Nagierowska2, J.B. Hubert1,
I. Theodorou2. 1 Inserm U292 Hopital de Bicetre 82 Rue du Gal Leclerc 94276
Kremlin-Bicetre; 2Hopital Pitie-Salpetriere VRA CNRS 625 PAris, France
Background: The A32 deletion has been found to delay progression to immunodeficiency in HIV-infected patients. Here we examined whether the deletion
confers a specific protection against toxoplasmosis occurrence.
Methods: 1489 Caucasian HIV-infected subjects from the French HEMOCO
(haemophiliacs) and SEROCO (non-haemophiliac adults) cohorts were genotyped for the A32 deletion. Patients were AIDS-free at enrolment and were
followed every 6 months between 1988 and 1996 (median follow-up 64 months).
Progression from inclusion to toxoplasmosis occurrence (overall and as a first
AIDS-defining illness only) was compared between heterozygotes (n = 227) and
wild-type (wt) patients (n = 1262). Progression to toxoplasmosis as a secondary
AIDS-defining illness since AIDS onset was also examined. Adjustment for age
and CD4+ cell counts at inclusion and prescription of a specific prophylaxis (as a
time-dependent variable) was made using a Cox model.
Results: 116 cases of toxoplasmosis were diagnosed during follow-up, among
which 58 as a first AIDS-defining illness. The overall progression to toxoplasmosis
tended to be slower in heterozygotes than in wt patients (Relative Risk (RR):
0.66; [95% CI: 0.38-1.15], p = 0.14). Progression to toxoplasmosis as a first
AIDS-defining illness was also studied, patients having progressed to another
AIDS-defining illness being censored at the date of AIDS onset. Only 3 cases of
toxoplasmosis were observed in heterozygous patients versus 55 in wt patients
(logrank test p = 0.02, RR = 0.28 [0.09-0.89]). The shape of the Kaplan-Meier

curves was similar in the SEROCO and HEMOCO cohorts. Adjustment for CD4+
cell counts and age did not modify the results. Further adjustment for prescription
of a specific prophylaxis only slightly reduced the protective effect of the deletion
(adjusted RR 0.38 [0.11-1.22], p = 0.10). Moreover, CD4+ cell counts at the time
of this toxoplasmosis onset was lower in heterozygous patients than in others,
the difference being of borderline significance (median of 15 versus 40 CD4+
cells//il, p = 0.08). Finally, no protective effect of the deletion was observed when

the risk of progression to toxoplamosis as a subsequent illness since AIDS onset
was studied.
Conclusion: The A32 deletion delays toxoplasmosis occurrence. This transient protection is no longer observed when patients become severely immunodepressed, i.e. after AIDS onset.
S13185 Incidence and prevalence trends of HIV among injecting
drug users in northern Italy, 1990-1996 (NISDA study)
Alfredo Nicolosi1, A. Camisani2, S. Quaresima2, E. Elba3, R. Durello4,
B. Tinghino5, E. Duratorre6. 1Dept. Epidemiology -ITBA CNR Via F LLI Cervi 93 -20090 Seagrate (MI); 2ASL Provincia Di Cremona - Sert Cremona; 3ASL
Provincia Dibrescia - Sert Brescia; 4ASL Provincia Milano 1 - Sert Limbiate (MI);
5ASL Provincia Di Milano 3 - Sert Monza (MI); 6ASL Provincia Di Varese - Sert
Varese, Italy
Objective: To describe the time trends of the HIV epidemic among intravenous
drug users (IDUs) attending drug dependence treatment centers (DDTCs) in
Lombardy and Northern Italy.
Methods: Incidence rates were estimated by using data of the cohort of
seronegative IDUs attending 32 DDTCs from Lombardy and limiting areas enrolled in the NISDA study. Incidence rates (IR) and confidence intervals (CI)
were calculated by the person-years (PY) method and represented as moving
averages. Prevalence of HIV infection was calculated by using data on IDUs
attending all the 84 DDTCs operating in the region Lombardy between 1990 and
1996 (on the average, more than 10,000 subjects per year).
Results: Between 1990 and 1996, 101 new seroconversions were observed in
subjects followed for 6,655 PY, yielding an overall incidence rate of 1.5 per 100
PY (95% CI 1.2-1.9): incidence was 1.3 (95% CI 1.0-1.6) in males and 2.6 in
females (95% CI 1.8-3.8). The incidence rate per 100 PY was 3.5 in 1990, 4.5
in 1991, 1.7 in 1992, 1.4 in 1993, 1.2 in 1994, 1.2 in 1995 and 1.1 in 1996. In
males the rate decreased from 3.0 in 1990 to 0.9 in 1996, and in females from
5.8 in 1990 to 2.0 in 1996. Between 1993 and 1996, no seroconversions were
observed in individuals under 20 years of age; the incidence rate per 100 PY in
age group 20-24 was 1.4 in males and 3.7 in females; in age group 25-29 it was
1.0 in males and 2.3 in females; in age group 30+ it was 0.7 in males and 1.4 in
females. From 1990 to 1996, the prevalence of HIV among IDUs already under
treatment decreased in males from 52% to 33%, and in females from 51% to
37%. Among IDUs who entered treatment, HIV prevalence, from 1990 to 1996,
decreased from 31% to 14% in males, and from 30% to 21% in females.
Conclusions: The speed of the epidemic of HIV infection among intravenous
drug users in Northern Italy has been declining in males, while the decline is
smaller in females. This pattern is due to the reduction in the parenteral in favor
of the heterosexual mode of HIV transmission.
13186 Possible epidemic spread of HIV by syringe-based drug
sharing in Poland: Paradigm for the new AIDS frontier
Philip Alcabes1, J.P. Grund2, M. Beniowski3, E. Bozek3, A. Kaciuba4,
A. Zielinski5. 'New York Univ. School of Medicine, 341 E. 25th Street, 2nd fl. NY
NY 10010; 21HRD, Open Society Institute, New York, NY; 4Lublin Social Assn.;
5Sunny Sch. of Health Sciences, Albany NY USA; 3Katowice Centre for AIDS
Diagn. & Ther Chorzow, Poland
Background: Since appearing in Poland in 1988, HIV has infected >10,000
people; 69% of reported infections were in known injecting drug users (IDUs).
HIV prevalence and AIDS case counts rose extremely rapidly in some Polish
cities despite the absence of shooting galleries, injectable cocaine, or high-frequency changes of sex partners that underpinned rapid HIV spread elsewhere.
The Polish pattern has been repeated recently in many parts of the former Soviet
Union. We sought to determine the nature and causes of epidemic HIV spread in
Polish IDUs.
Methods: HIV-antibody prevalence was measured by standard methods among
IDUs entering detox or treatment programs in Poland. Drug-using, -sharing, and
-selling practices were directly observed in manyPolish cities.
Results: Examples of explosive HIV spread in Poland: (1) In the southeastern
city of Pufawy, HIV prevalence among treatment-program entrants was 0/25 in
1992, 3/32 (9%) in 1993, 31/40 (78%) in 1994; 98% is estimated today. (2) HIV
prevalence among detox entrants in Silesia:
Year 1988 1989 1990 1991-94 1995
No. (%) HIV+ 0 (0) 2 (0.9) 95 (35.4) 569 (27.0) 156 (47.2)
Throughout the epidemic period the Polish drug scene was dominated by homemade opiate sold only in liquid form ("kompot"), requiring syringes to be used
in drug sharing and selling, and most Polish kompot producer/sellers were also
users. Kompot stocks for sale could be HIV-contaminated by the seller/IDU himself or when the dealer's supply contacts the syringe of an already-HIV-infected
user/client through syringe-mediated sale; contaminated stocks could infect all
subsequent clients. When samples of kompot offered for sale in one Polish city

were examined in 1996, visible blood residues were found in 3 of 17 samples.
Conclusion: Syringe-based sharing and selling of liquid drugs could be the
vehicle for explosive HIV spread throughout the Eastern European kompot zone
even when IDUs do not share equipment. Appropriate epidemic-preventive interventions at both community and individual level must caution against contamination of the local drug supply, teach safe drug sharing methods, and support
immediate disposal of used syringes without reuse.