Between January 2004 and October 2004, 600 individuals
were randomized: 300 to the active nevirapine group (N) and 300 to the active abacavir group (A). selleck Baseline characteristics were broadly similar (Table 1). A total of 563 participants (94%) completed 48 weeks (286 in A and 277 in N); 25 (4%) died (nine in A and16 in N) and 12 (2%) were lost to follow-up (five in A and seven in N). The randomized drug had been substituted/stopped in 21 participants (7%) receiving abacavir vs. 34 participants (11%) receiving nevirapine by 48 weeks/last follow-up (exact P=0.09). The majority had substituted abacavir/nevirapine with tenofovir DF for adverse events (five in A and 12 in N; mostly suspected hypersensitivity while on the blinded drug), or to start anti-tuberculosis treatment
as per protocol (five in A and 17 in N), or for personal reasons (one in A). The remainder had stopped ART for adverse events (two in A and one in N) or personal reasons Selleck LBH589 (one in A and three in N), or changed to the opposite drug for pregnancy (one in A) or adverse events (two in A) or in error when unblinded at 24 weeks (four in A and one in N). Fifty-one participants (8%) had substituted stavudine for zidovudine, mostly for anaemia/neutropenia (25 in A and 26 in N). In the abacavir group, 94.8% of person-time spent under follow-up to 48 weeks was spent on abacavir+lamivudine+zidovudine/stavudine compared with 91.1% on nevirapine+lamivudine+zidovudine/stavudine in the nevirapine
group. Adherence by 4-weekly self-reported questionnaire was similar in the abacavir and nevirapine groups, with means of 3.7%vs. 2.6%, respectively, reporting missing pills in the last 4 days (P=0.32), and 14.5%vs. 13.4%, respectively, in the last 28 days (P=0.70). To 48 weeks, there was a consistent trend towards clinical superiority of abacavir over nevirapine in terms of HIV-related events (Fig. 1). Nine participants in the abacavir group vs. 16 in the nevirapine group had died (HR 0.55; 95% CI 0.24–1.25; P=0.15) and 20 vs. SPTBN5 32, respectively, had experienced a new or recurrent WHO stage 4 event or died (HR 0.60; 95% CI 0.34–1.05; P=0.07). The first new or recurrent WHO stage 4 events were oesophageal candidiasis (four in A and six in N), extrapulmonary tuberculosis (two in A and five in N), cryptococcus (two in A and four in N), Pneumocystis carinii pneumonia (two in A and one in N), herpes simplex (two in A and one in N), toxoplasmosis (one in A and one in N), Kaposi sarcoma (two in N), HIV wasting (one in N), and cryptosporidia (one in N); and 18 participants (seven in A and 11 in N) died without a new or recurrent WHO 4 event being identified after ART initiation. Forty-eight participants in the abacavir group vs. 68 in the nevirapine group experienced a new or recurrent WHO stage 3 or 4 event or died (HR=0.67; 95% CI 0.46–0.96; P=0.03).