A randomized trial comparing regimens of four reverse transcriptase inhibitors given together or cyclically in HIV-1 infection--the Quattro Trial. Quattro Steering Committee.

OBJECTIVES: To assess if cyclical therapy with four reverse transcriptase inhibitors is less toxic and as active, based on virological and immunological markers, as the same four drugs given together over a 64-week period. METHODS: An open randomized trial comparing concurrent therapy (T4) with zidovudine, lamivudine, loviride and zalcitabine with the same four drugs given cyclically each for 8 weeks (C4) and with concurrent zidovudine and lamivudine (T2), all given for a total of 64 weeks. The primary endpoint was the change in plasma HIV RNA level from baseline at weeks 32 and 64. Phenotypic and genotypic resistance, CD4+ cell counts, and clinical and laboratory assessments of safety were also compared. Patients were followed for up to a further 32 weeks beyond 64 weeks. Eligible patients had CD4+ cell counts between 50 and 350 x 10(6)/l and no prior antiretroviral therapy. RESULTS: One hundred individuals (34 T4, 34 C4, 32 T2) were recruited between 31 July 1995 and 11 July 1996, of whom 22 had AIDS; the mean (SD) HIV RNA at baseline was 4.9 (0.7) log10 copies/ml and the median (interquartile range) CD4+ cell count was 170 (100-260) x 10(6)/l. A total of 28 T4, 19 C4 and 26 T2 participants were still on the allocated regimen at week 64. A new AIDS event or death was reported in three T4, seven C4 and five T2 participants (P = 0.7). Serious adverse events that were likely to be drug related were observed in three T4, one C4 and four T2 participants. The reduction from baseline in HIV-1 RNA (log10 copies/ml) was greatest in the T4 arm; at 32 weeks the mean reduction (SD) was 1.45 (0.72), 0.42 (0.45) and 1.05 (0.70) in T4, C4 and T2 respectively (global P = 0.0001) and at week 64 1.24 (0.86), 0.73 (0.91) and 0.78 (0.55) respectively (P = 0.02). The pattern of CD4+ change mirrored the changes in HIV RNA. Very few mutations associated with resistance to loviride or zalcitabine were seen. The mutation at codon 215 associated with zidovudine resistance was detected (> 5% of population mutant) in 11 out of 24 T4 participants compared with three out of 21 C4 and 11 out of 20 T2 participants at week 64 (P = 0.02). Further assays of viral resistance including phenotypic assays are ongoing and results will be reported separately.