Objective: To determine the value of plasma HIV-1 RNA and CD4 cell count as predictors of the clinical benefit of antiretroviral treatment.
Design and setting: The CAESAR (Canada, Australia, Europe, South Africa) trial randomized 1840 patients [inclusion CD4 cell count, 25-250 x 10(6)/l] to add either placebo, lamivudine (3TC) or 3TC plus loviride in a double-blinded fashion to baseline treatments (zidovudine, zidovudine-didanosine or zidovudine-zalcitabine) for 1 year.
Patients: This analysis included 487 patients with data on CD4 cell count and HIV-1 RNA after 12-20 weeks of treatment and subsequent follow-up for clinical progression.
Main outcome measures: The correlation between 12-20-week change in CD4 cell count, HIV-1 RNA and progression to AIDS or death in the placebo group was used to predict the clinical benefit of the 3TC-containing arms of the trial, given their effects on CD4 cell count and HIV-1 RNA.
Results: After 12-20 weeks of treatment, HIV-1 RNA fell by 0.37 log10 copies/ml in the 3TC arms versus a rise of 0.05 log10 copies/ml in the placebo arm. The 12-20-week CD4 cell count rose by 35 x 10(6)/l in the 3TC arm versus a fall of 8 x 10(6)/l in the placebo arm. After 12-20 weeks of treatment, a reduction in HIV-1 RNA of 1 log10 at 12-20 weeks predicted a 49% reduction in progression [hazard ratio (HR), 0.51; 95% confidence interval (CI), 0.30-0.87] and a rise in CD4 cell count of 50 x 10(6)/l predicted a 51% reduction in progression (HR, 0.49; 95% CI, 0.33-0.73). Using the model from the placebo arm, the rises in CD4 cell count and reductions in HIV-1 RNA during 3TC treatment predicted a 59% reduction in progression to AIDS or death. The observed clinical benefit was a 57% reduction in progression for the 3TC arms versus placebo (HR, 0.43; 95% CI, 0.26-0.71).
Conclusions: Rises in CD4 cell count and reductions in HIV-1 RNA were reliable in predicting the clinical benefit of 3TC in the CAESAR trial.