 |
Discordance between virological/immunological and clinical outcomes at 48 Weeks, in a randomised comparison of ZDV/3TC/NVP and ZDV/3TC/ABC in 600 patients with low CD4 counts in Africa
Presented by Paula Munderi, Uganda.
Munderi P.1, Walker S.2, Kityo C.3, Kaleebu P.1, Ssali F.3, Lyagoba F.1, Reid A.4, Gibb D.2, Gilks C.5, Mugyenyi P.3, on behalf of the DART Trial
1MRC/UVRI Uganda Research Unit on AIDS, Entebbe, Uganda, 2MRC Clinical Trials Unit, London, United Kingdom, 3Joint Clinical Research Centre, Kampala, Uganda, 4University of Zimbabwe, Harare, Zimbabwe, 5World Health Organisation, Geneva, Switzerland
Objectives: To investigate potential reasons for discordant virological/immunological and clinical outcomes to 48 weeks, in a randomised trial of zidovudine/lamivudine/nevirapine(NVP) versus zidovudine/lamivudine/abacavir (ABC).
Methods: 600 ARV-naive adults with CD4<200 cells/mm3 in 2 Ugandan centres were randomised to Zidovudine/Lamivudine plus ABC or NVP (placebo-controlled to the primary toxicity endpoint at 24 weeks). Plasma HIV-1 RNA was retrospectively assayed using Roche Amplicor, CD4 measured in real-time, and major clinical events independently reviewed. All analyses are ITT
Results: Median baseline CD4 was 99 cells/mm3 and RNA 284,600 copies/ml (balanced across groups). 563 (94%) patients completed 48 weeks: 25 (4%) died and 12 (2%) were lost. At 48 weeks virological suppression was superior with NVP (77% vs 62% ABC <50 copies/ml, p<0.001) as was mean CD4 increase (+173 versus +147 cells/mm3; p=0.006). However, there was a clear trend towards higher rates of death (HR=1.82,p=0.15), new WHO stage 4 event or death (HR=1.76,p=0.06) and new WHO stage 3/4 event or death (HR=1.49,p=0.05), in the NVP arm. There were fewer ART substitutions with ABC, and few substitutions before clinical events in either group, suggesting that moving to more virologically/immunologically potent regimens was not driving the clinical results. Although most clinical events occurred before 12 weeks, there was no evidence of heterogeneity in treatment effect before and after 12 weeks, for death (HR=0.57,0.48 respectively, hetp=0.86) or other outcomes; nor of heterogeneity by baseline CD4, suggesting that IRIS was unlikely to be a major factor. Adjusting for most recent CD4 and HIV-RNA did not explain the clinical difference between randomised groups.
Conclusions: Differences between outcomes may be a chance finding: if real, they suggest a disconnect between early virological/immunological and clinical outcomes, which may influence the way surrogate markers are interpreted. Given these early differences, attenuation over time may still not result in clinical benefit with NVP in the medium term.
Back to the session -
Back to the Programme-at-a-Glance
|
|