AEGiS-13IAC: Evaluation of the effect of time from seroconversion on AIDS: results from CASCADE (Concerted Action on SeroConversion to AIDS and Death in Europe).

13th International AIDS Conference

Durban, South Africa - July 9-July 14, 2000

Evaluation of the effect of time from seroconversion on AIDS: results from CASCADE (Concerted Action on SeroConversion to AIDS and Death in Europe).

Int Conf AIDS 2000 Jul 9-14; 13:(abstract no. TuOrC367)

Babiker A, Pezzotti P, Walker S, behalf of the CASCADE Collaboration on
A. Babiker, MRC Clinical Trials Unit, 222 Euston Road, London, NW1 2DA, United Kingdom, Tel.: +44 20 7670 4719, Fax: +44 20 7670 4818, E-mail: agb@ctu.mrc.ac.uk

OBJECTIVES: Previous studies have suggested that individuals with similar CD4 count and different time from seroconversion have similar rates of progression to AIDS. We evaluated whether there was a residual effect of time from seroconversion on progression to AIDS from specific CD4 count levels.

METHODS: Pooled data of persons with well documented times of HIV seroconversion from 19 cohorts in 11 countries were analysed. We considered time to the first AIDS-defining event following CD4>500, CD4>350 and CD4>200 allowing for late entry. We explored the effect of time from seroconversion to the CD4 threshold adjusting for age, sex, exposure category, and presentation during acute infection, in Cox models stratified by study. The residual effect of time from seroconversion after adjustment for serial CD4 and calendar year at risk (time-dependent covariates) was also estimated. Analyses were censored at 31/12/1995 to exclude a confounding effect of HAART.

RESULTS: 3825, 3006 and 1804 seroconverters reached CD4 thresholds of 500, 350 and 200 respectively, of whom 1274, 1192 and 985 developed AIDS. As expected there was a strong effect of time from seroconversion to CD4 threshold, with a 1 year increase associated with an increase in risk of progressing to AIDS of 14% (95% CI 10-17%), 12% (9-15%) and 10% (7-13%) respectively (all p>.0001). Adjusting for serial CD4 counts reduced the effect of time from seroconversion: but a 1 year increase was still associated with an increase in risk of 7% (3-10%), 6% (3-9%) and 4% (1-7%) respectively (p = 0.01, 0.01 and 0.08). The effect of time from seroconversion appeared to be non-linear, with no apparent increase in risk after 4 years, and 1 year increase before 4 years associated with an increase in risk of 12%, 11% and 11% respectively.

CONCLUSIONS: We found a significant residual effect of time from seroconversion to CD4 threshold on time to AIDS after adjusting for sex, age, exposure, acute infection and CD4 counts over time.

Keywords: AEGIS, Acquired Immunodeficiency Syndrome, HIV Seropositivity, CD4 Lymphocyte Count, Antiretroviral Therapy, Highly Active, Proportional Hazards Models, Europe, Death, Time
000709
TuOrC367