Background: The use of HAART combining 2 nucleoside analogues reverse transcriptase inhibitors (NRTIs) plus one protease inhibitor (PI) or 2 NRTIs + 1 non-nucleoside reverse transcriptase inhibitor (NNRTI) has shown comparable efficacy. The study was designed to compare long term (2 years) effectiveness of two antiretroviral (ARV) treatment strategies in patients not previously treated: starting with a nelfinavir based HAART switching to nevirapine in case of failure or side effects or the reverse sequence.
Methods: This multicenter, randomized, open label clinical trial enrolled ARV-naïve HIV patients with CD4 counts below 500 cells/mm3. They were randomly assigned to start ddI + d4T + nelfinavir (switching to ZDV + 3TC + NEV in case of failure or toxicity) (PI-NEV arm) or ddI + d4T + nevirapine, switching to ZDV + 3TC + NFV in case of failure or toxicity (NEV-PI arm). The primary study endpoint was the Kaplan-Meier estimates of the time to failure after switching to second regimen if necessary (considering failure as two consecutive plasma HIV-1 RNA determinations above 200 copies/mL, death, a new category C event or toxicity leading to treatment discontinuation of the second regimen) after a minimum follow-up of two years.
Results: A total of 137 patients were evaluable (67 and 70 in the PI-NEV and NEV-PI arms respectively). Baseline characteristics did not differ among groups. Kaplan-Meier estimates of time to failure did not show differences between the two arms neither in the on-treatment (OT) analysis (log rank test, p = 0.81) nor in the intent-to-treat (ITT) analysis (p = 0.58). At 24 months, the estimated proportion of patients free of failure were 72% and 66% respectively in the PI-NEV and NEV-PI arms OT analysis (p = 0.54) and 73% and 64% in the PI-NEV and NEV-PI arms in the ITT analysis (p = 0.49). The difference in the median in CD4+ lymphocyte count at 24 months was not significantly different in the two groups: 393 and 307 CD4 cells/mm3 in the PI-NEV and NEV-PI arms respectively (p = 0.167). The incidence of adverse events (AEs) in the two arms was very similar: 50 (75%) in the PI-NEV and 54 (70%) in the NEV-PI group, as it was for grade 3-4 AEs leading to drug switching.
Conclusion: At two years both treatments strategies (PI-NEV vs NEV-PI) had a high and comparable efficacy and were generally well tolerated.