Objective: To compare the long-term efficacy and tolerability of two efavirenz-containing regimens with those of an indinavir-containing regimen in the initial use of HAART.
Method: HIV-1-infected patients (N = 1266) were randomly assigned to receive one of three regimens: efavirenz, zidovudine plus lamivudine, n = 422; efavirenz plus indinavir, n = 429; or indinavir, zidovudine plus lamivudine, n = 415. Entrance criteria included baseline viral load greater than 10 000 copies/ml HIV-1 RNA, CD4 cell count 50 cells/mul or greater, and no previous use of lamivudine, any non-nucleoside reverse-transcriptase inhibitor or protease inhibitor. The primary endpoint was the proportion of patients (response rate) in each regimen with a viral load under 400 copies/ml at 168 weeks of treatment.
Results: Response rates at 168 weeks were 30% in the indinavir, zidovudine, lamivudine group, 48% in the efavirenz, zidovudine, lamivudine group (P < 0.0001, difference estimate; 97.5% confidence interval (CI) 18.5; 10.9, 26), and 40% in the efavirenz plus indinavir group (P = 0.0018, difference estimate; 97.5% CI 10.2; 2.9, 17.6). Median CD4 cell counts increased above respective baselines by 292 cells/mul (efavirenz, zidovudine, lamivudine and indinavir, zidovudine, lamivudine) and 300 cells/mul (efavirenz plus indinavir). Total discontinuations were 54% (efavirenz, zidovudine, lamivudine), 63% (efavirenz plus indinavir), and 69% (indinavir, zidovudine, lamivudine) of which 13, 12 and 26%, respectively, were caused by adverse events. No new or unexpected increases in the rates or severity of adverse events occurred from long-term treatment with efavirenz-containing regimens.
Conclusion: Long-term HIV therapy with efavirenz-containing regimens, particularly efavirenz, zidovudine, lamivudine, provides significantly greater antiviral activity and tolerability than a regimen of indinavir, zidovudine plus lamivudine.