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

150 Abstracts 13320-13323

12th World AIDS Conference

Results: During the study period (1985-1997), HIV-1 seroincidence rate was
increasing (from 0 in 1985 to 3.9 per 100 PYO in 1996) whereas HIV-2 seroincidence rate was lower and stable at approximately 1 per 100 PYO.
Incidence rates of AIDS in seroincident women
HIV-1 HIV-2
Number of Women 81 50
Total PYO 230.34 210.53
AIDS Cases 13 2
AIDS Incidence Rates* [95% CI] 5.64 [3.3-9.7] 0.95 [0.2-3.8]
Rate Ratio [95% Cl] 6 [1.4-25.2]
HIV-1 infected women progressed to AIDS faster than their HIV-2 counterparts
(Kaplan Meier disease-free survival analysis, Wilcoxon-Gehan test: P < 0.01)
with the probability of AIDS free survival for HIV-1 of 72% [95% CI = 53-85%]
at 5 years. In contrast, no HIV-2 seroincident women developed AIDS during the
same time period. A Cox proportional hazards model showed a hazard ratio of 14
[95% CI = 2.5-74] when the AIDS incidence rates of HIV-1 and HIV-2 infection
were compared.
Conclusions: These findings document the disease incident rates for both
HIV-1 and HIV-2 in this region and continue to establish the distinct disease
incident rate for HIV-2 as compared to HIV-1.
113320 Effect of mannose-binding lectin variant alleles on HIV-1
disease progression in 426 patients with a known date of
infection
Nagdalena Nagierowcka', J.B. Hubert2, P. Dembre3, C. Rouzioux4,
J.F. Delfraissy5, L. Neyer2, I. Theodorn3. 1Laboratoire D'nnunologie Cellulaire
Et Tissulaire URA CHRS G25 Pitie-Salpetriere; 2/nsern U212 Hopital de Bicetre
Vrenlin-Bicetre; 3Hopital Pitie Salpetriere URA CNRS G25 Paris; 4Hopital
Necker Paris; 5Hopital de Bicetre Krenlin-Bicetre, France
Background: Mannose binding lectin (MBL) is a versatile macromolecule, embodying functional characteristics of IgM, IgG and Clq; it is believed to be an
important constituent of the innate immune system. The objective was to determine the influence of variant alleles (in codons 52 and 54) in the MBL gene on
disease progression from different stages of HIV-1 infection.
Methods: HIV-1 disease progression since infection to clinical AIDS (1993
European definition, n = 108), CD4+ cells <200//pL (n = 140) and death (n = 74)
was analysed according to the MBL genotype (determined by PCR RFLP) in 426
patients (331 men and 95 women) with a known date of seroconversion enrolled
in the SEROCO cohort (median follow-up 74 months). Late survival was studied
in patients whose CD4 fell below 200/ptL.
Results: At least one variant MBL allele was found in 112 (23.6%) individuals
(MBL+ group). MBL+ and MBL- patients did not differ according to Caucasian
origin, gender, sexual preference or age at infection. Early serum viral load (during the 6 to 24 month-period after infection) was similar in both groups (median:
3.9 versus 4.0 loglo copies/ml). Overall, Kaplan-Meier curves did not show any
difference in progression to clinical AIDS (logrank: p = 0.39) or death (p = 0.27) in
MBL+ compared to MBL- patients. There was a tendency (p = 0.09) to a slower
progression to CD4 < 200 in MBL+ patients (RR = 0.70; 95% Cl: 0.44-1.08).
The frequency of the different AIDS-defining illnesses did not differ according to
MBL status. Among the 140 patients who progressed to CD4+ cells <200/p/L,
31 (22.1%) where MBL+. In this subgroup, age, clinical status (AIDS or not) and
viral load at the CD4 threshold of 200 did not differ significantly between MBL+
and MBL- patients. Survival time since CD4+ < 200 was significantly shorter
in MBL+ compared to MBL- patients (median: 34 versus 50 months; logrank:
p = 0.04). After adjustment for risk factors of progression (fixed variables) and
treatments (antiretroviral and prophylaxis of O.1. as time-dependent variables) in
a Cox model, relative risk (RR) of death since CD4 < 200 was 2.2 [1.1-4.5] for
MBL+ compared to MBL- patients.
Conclusion: Our results are in keeping with a recent report (Garred et al,
Lancet 1997), confirming that variant MBL alleles affect survival of HIV-infected
patients after the adaptative immune system is impaired.
| 13321 | Amsterdam cohort study: Presence of the variant
mannose-binding lectin (MBL) alleles associated with
slower progression to AIDS?
Jaap Maas1, A.M. De Roda Husman2, M. Brouwer2, A. Krol1, R.A. Coutinho1,
I.P.M. Keet1, H. Schuitemaker2. 1 Nieuwe Achtergracht 100, 1018 WT, Municipal
Health Service; 2Central Laboratory Blood Transfusion Service, Amsterdam, The
Netherlands
Objective: To examine the association between Mannose Binding Lectin (MBL)
polymorphism and progression to AIDS and death in HIV-1 infection.
Methods: For 131 HIV-1 infected homosexual men with a known time of seroconversion, Kaplan Meier and Cox proportional hazards analyses were performed
to determine the association between MBL genotype and time from HIV-1 sero
conversion to AIDS and death, and time from AIDS to death. Conditional logistic
regression was used to study the MBL genotype frequency in a matched casecontrol study, comparing 16 seroprevalent non progressors (>9 years AIDS free)
individually matched with 2 seroprevalent progressors (AIDS within 2-7 years).
Results: Of the 131 seroconverters, of whom 61 developed AIDS, 76 were
typed as homozygous wild type and 55 as carriers of variant alleles (52 heterozygous and 3 homozygous variant alleles). In the survival analyses the HIV-1 infected

men with the variant alleles progressed somewhat slower to AIDS (relative hazard (RH) 0.61 (95%CI: 0.36-1.10)) and death (RH 0.73 (95%CI: 0.42-1.25)). No
difference could be found between survival time from AIDS to death. Interestingly,
CD4+ T cell count determined at moment of AIDS were found to be significant
lower among persons with the mutation (177 x 106/1 versus 212 x 106/1). In the
case-control study, 7 out of 16 non progressors as compared to 11 out of 32 progressors were carriers of the variant alleles (odds ratio 0.61 (95%CI: 0.15-2.42)).
Conclusion: Although an unexplained lower CD4+ T cell count was found at
moment of AIDS diagnosis, the results of our study are suggestive for a weak
pre-AIDS protective effect of variant MBL alleles.
113322 Patterns of longitudinal HIV-1 viral load after
seroconversion associated with HIV-1 disease
progression among injection drug users
Cynthia Lyles2, J. Astemborski2 3, T.C. Quinn', T.R. Sterling', T. Freedman2,
D. Vlahov2,3. School of Medicine, 2School of Public Health, 3Dept. of
Epidemiology - Johns Hopkins University, 615 N. Wolfe St., Baltimore, MD, USA
Objectives: To evaluate whether the level, change, or variability in longitudinal
HIV-1 viral load data after seroconversion are associated with progression to
AIDS in a cohort of injection drug users (IDU).
Methods: A nested case-control study was designed to select 2 clinically divergent groups of IDU with known dates of HIV seroconversion from a cohort
established in 1988-89. The cases, defined as having an AIDS diagnosis prior
to July 1, 1996, and controls, defined as being AIDS-free through July 1, 1996,
were required to provide at least 3 visits prior to AIDS or July 1, 1996, for viral
load quantification. Plasma levels of HIV-1 RNA were measured by RT-PCR assay (Roche). Subject-specific summary measures of longitudinal viral load were:
initial level, change per year and variability in viral load. Mixed effects regression
models were used to relate case status to initial level, subject slope and subject
variability of the viral load, in separate models, while accounting for the correlation
within casesets.
Results: 617 longitudinal viral loads were measured on 24 AIDS cases and 47
matched AIDS-free controls. Median levels of initial CD4' cell count and viral load
(CD4; vl) were (742; 63,347) for cases and (733; 43,316) for controls, respectively;
while, median levels throughout entire follow-up were (397; 75,406) for cases and
(502; 21,224) for controls, respectively. Cases had an estimated 0.46 (p <.05)
higher initial log viral load relative to controls, after adjusting for gender, CD4+ cell
count and time since seroconversion. The viral load increased by an estimated
39% per year for cases and only 2% per year for controls. After controlling for
gender and initial viral load level, viral load increased at a significantly faster
rate for cases than for controls (p <.01). The variability of viral load about the
trajectory did not differ by case status.
Conclusions: AIDS-free subjects appear to have fairly stable viral load levels
over time. Higher viral load level, as well as a steeper increase in viral load over
time, were indicative of faster progression to AIDS in a cohort of injection drug
users. The degree of variability of the viral load measures was not related to
progression. In addition to level, changes in viral load may need to be assessed
when monitoring the effectiveness of therapy.
13323 Temporal changes on the rate of progression to AIDS in
the HIV-ltalian seroconversion study (ISS)
Maria Dorrucci, P. Pezzotti, M.B. Alliegro, G. Rezza. The Italian Seroconversion
Study; Istituto Superiore di Sanita', Coa V Le Regina Elena 299, Roma, Italy
Objectives: To evaluate temporal changes on the incubation period from HIV
seroconversion to AIDS.
Design: Prospective cohort study of HIV-positive adults with known seroconversion (SC) dates.
Methods: Participants were followed-up from SC to death or 30-6-1997. All
patients lost to follow-up were cross-checked with the National AIDS registry.
Clinical AIDS was the end-point. SC dates were estimated as the midpoint
between the last HIV-negative and the first HIV-positive test. Multivariate Cox
models were applied to estimate relative hazards (RHs) of AIDS. Year of SC (as
a fixed categorical variable) and calendar time (as a time-dependent categorical
variable) were considered to evaluate cohort and prevalent changes on the rate
of progression, respectively. The analysis was controlled by age at SC, acute
infection, lag between last HIV- and first HIV+ test, and time from 1/1/80 (missing
before SC, and time dependent thereafter).
Results: Of 1245 individuals [52.8% injecting drug users, 26.2% homosexual
males, 19.4% heterosexuals contacts and, 1.6% not known], 404 developed AIDS.
Of these persons 482 (38.7%) seroconverted from '80 to '87 and the remaining
763 (61.3%) from '88 to '95]. In the multivariate Cox model the RHs of AIDS there
was not a significant effect of year of SC. On the other hand, from the 2nd half of
96 the RH of AIDS started to decrease from the previous calendar time periods
(RH = 0.66, 95%CI: 0.41-1.05, RH = 0.33; 95%CI: 0.17-0.61 for the 2nd half
of 96 and the 1st half of 97, respectively compared to 94-95). Repeating these
analyses stratifying by exposure category, the prevalent effect in the most recent

calendar periods was larger in the individuals with sexual exposure compared to
those infected through injection drug use.
Conclusion: The analysis did not show significant changes by year of SC on
the HIV-disease progression rate. However, a 'prevalent effect' was found during
calendar years '96-'97. Analyses with longer follow-up are needed to verify if this
risk reduction of progression to AIDS will persist. Further analyses will permit to
verify if such an effect is associated with the use of new combination therapies.