1). Similar results were obtained excluding the 15 women with previous antiretroviral exposure to prevent mother-to-child transmission. Six HIV-related severe pulmonary or central nervous system events (four in A and two in N), reported as WHO stage 4 events but judged not to meet diagnostic criteria for pneumocystis or toxoplasmosis on blinded review by the ERC, were not included as WHO 4 endpoints because they did not meet the protocol definitions [one patient (in N) subsequently died, and two (one in A and one in N) had other WHO 4 events included in WHO 4/death outcomes]. The trend towards clinical superiority with abacavir remained after including these six severe brain/lung events (Fig. 1). There
was no evidence that the trend towards clinical superiority with abacavir was limited to subgroups defined by centre, year of ART initiation, randomized monitoring strategy or Luminespib nmr pre-ART age, CD4 cell count, HIV-1 RNA, weight or WHO stage (considering the effect size in each subgroup as well as statistical significance). In particular, there was no evidence of heterogeneity in the relative difference between abacavir and nevirapine in those with pre-ART CD4 counts of 0–49, 50–100 and 100–199 cells/μL for death
(HR 0.82, 0.25 and 0.75, respectively; heterogeneity P=0.47), new or recurrent WHO 4 events or death (HR 0.64, 0.30 and 0.99, respectively; heterogeneity P=0.36), new or recurrent WHO 3 or 4 events or death (HR 0.62, 0.78 and 0.69, respectively; heterogeneity P=0.90) check details or other outcomes. Most deaths and disease progression events occurred early after ART initiation (Fig. 2). All but one death (in N) occurred in the first 24 weeks, with most (seven of nine in A and 12 of 16 in N) occurring in the first 12 weeks, and most new or recurrent WHO 4 events and deaths (15 of 20 in A and 25 of 32 in N) also occurred in the first
12 weeks. Despite much smaller overall event rates after 12 weeks, there was no evidence of heterogeneity in the relative difference between abacavir and nevirapine before and after 12 weeks for death (HR 0.58 and 0.48, respectively; heterogeneity P=0.86) or new or recurrent WHO 4 Lonafarnib mouse event or death (HR 0.58 and 0.67, respectively; heterogeneity P=0.84) (similar results were obtained splitting at 4, 8 or 24 weeks). The only outcome where estimates suggested that the relative difference between abacavir and nevirapine might possibly be attenuating or reversing was new or recurrent WHO 3 or 4 events or death (HR 0.56 for 0–12 weeks, HR 0.68 for 12–24 weeks, and HR 1.41 for 24–48 weeks) but, with the small number of events, the statistical evidence for this was weak (heterogeneity P=0.22). In contrast to clinical response, immunological response was superior with nevirapine compared with abacavir, with mean CD4 cell count increases of 173 vs. 147 cells/μL at 48 weeks (P=0.006) (Fig. 3 and Table 2).