Background: Baseline and follow-up predictors of new AIDS-defining events (ADE) or death among patients who started HAART with CD4+ T-cell counts > or =200 cells/mm3 have rarely been assessed simultaneously.
Methods: A prospective observational cohort study (1996-2002) is reported. HIV-infected patients initiating HAART with a CD4+ T-cell count > or =200 cells/mm3 were studied. Baseline and time-varying factors were tested for the prediction of new ADE/death using Cox regression models.
Results: A total of 896 subjects were studied over a median of 5.1 years. The incidence of a new ADE was 1.6 (95% confidence interval 1.3-2.1) per 100 person-years. Among baseline factors, higher CD4+ T-cell counts before HAART were associated with lower risk of ADE/death, but not after adjustment for time-varying factors. On a multivariable analysis including both baseline and time-varying covariates, longer delay from HIV diagnosis to HAART was an independent predictor of ADE/death (per year, hazard ratio [HR] 1.06; P = 0.025) and was independent of CD4+ T-cell count before treatment. Longer time spent with HIV RNA <400 copies/ml (per month, HR 0.96; P = 0.003) and higher latest CD4+ T-cell count (per log2 cells/mm3, HR 0.65; P < 0.001) were found to be protective.
Conclusions: Patients with higher CD4+ T-cell counts before HAART initiation had a better prognosis. However, except for the delay in starting HAART, viroimmunological evolution outweighed the effect of baseline factors. Moreover, suppressing HIV replication for as long as possible could improve the clinical outcome. Prospective randomized clinical trials to assess the optimal timing of HAART initiation are both feasible and urgently needed.