506
Superior Virological Suppression with Nevirapine, Zidovudine, and Lamivudine vs Abacavir, Zidovudine, and Lamivudine without Evidence of Clinical Benefit to 48 Weeks: A Randomized Comparison in Patients with Low CD4 Counts in Africa
Sarah Walker*1, C Kityo2, P Kaleebu3, F Ssali2, F Lyagoba3, A Reid4, D Gibb1, C Gilks5, P Mugyenyi2, P Munderi3, and the DART Trial Team
1Med Res Council Clin Trials Unit, London, UK; 2Joint Clin Res Ctr, Kampala, Uganda; 3Med Res Council/Uganda Virus Res Inst, Entebbe; 4Univ of Zimbabwe, Harare; and 5Imperial Coll, London, UK
Background: Abacavir
(ABC) -based 3-nucleoside reverse transcriptase inhibitor (NRTI) regimens have
potential advantages over standard nevirapine (NVP) -based first-line ART in
Africa, as they avoid interactions with tuberculosis (TB) therapy, spare 2
classes for second-line therapy, and have low pill burden. We have previously
shown a trend toward lower toxicity and discontinuation with ABC vs NVP, but
their clinical, immunological, and virological efficacy have not been compared
in Africa.
Methods: A
randomized trial comparing the safety of NVP vs ABC was conducted in 2 Ugandan
centers within the DART trial. We allocated 600 ART-naive adults with CD4
<200 cells/mm3 to combivir plus ABC (n = 300) or NVP (placebo-controlled for the first 24 weeks to the 1° toxicity
endpoint). Plasma HIV-1 RNA was retrospectively assayed at 0, 24, and 48 weeks
using Roche Amplicor. CD4 was measured weekly in real time for 12 weeks, and
clinical events documented and independently reviewed. All analyses are intent
to treat.
Results: Of the
total, 430 patients were women (72%); at baseline, 111 (19%) had WHO stage 4
disease, median age was 36 years (range 18 to 66), CD4, 99 cells/mm3
(1 to 199) and HIV-1 RNA 284,600 copies/mL (mean 5.4 log10, SD 0.7).
In all, 563 (94%) patients completed 48 weeks; 25 (4%) died and 12 (2%) were
lost. In total, 21 (7%) ABC vs 34 (11%) NVP patients were off randomized drug
at 48 weeks/last alive (most had substituted TDF); 62% ABC vs 76% NVP had HIV
RNA <50 copies/mL at week 24, and 62% ABC vs 77% NVP at week 48 (both n = 551, p <0.001); 74% vs 87% were <400 copies/mL at week 48 (p <0.001). Mean CD4 increases from
baseline in ABC vs NVP were +111 vs +134 cells/mm3 at week 24 (p = 0.003) and +147 vs +173 at week 48 (p = 0.006). In contrast, 17 (6%) ABC vs
29 (10%) NVP patients developed new WHO 4 events or died by 48 weeks (HR = 0.57
[95%CI 0.31 to 1.03] p = 0.06,
similar to WHO 3 events p = 0.06).
First events were esophageal Candida
(4A, 3N), extra-pulmonary TB (2A, 5N), cryptococcus (3N), other WHO 4 (4A, 5N),
or death (7A, 11N). Most (13A, 23N) occurred in the first 12 weeks; 9 (3%) ABC
vs 16 (5%) NVP patients died (HR = 0.55 [0.24 to 1.25] p = 0.15), all but 1 (N) in the first 24 weeks and most (7A, 12N)
in the first 12 weeks.
Conclusions: NVP
has superior virological and immunological efficacy compared with ABC over 48
weeks, although viral load suppression and CD4 gains were substantial in both
groups. Further follow-up will important to monitor the observed trend toward
clinical superiority of the ABC arm during this period. The potential
contribution of IRIS to these conflicting results also requires further study.
|