Effectiveness and pitfalls of initial highly active antiretroviral therapy in HIV-infected patients in routine clinical practice

Abstract

Objective: To assess the long-term effectiveness of and factors associated with response to protease inhibitors (PIs) in a cohort of treatment-naive HIV-infected patients. Design and setting: Prospective study in a tertiary care centre. Subjects: A total of 207 treatment-naive patients starting PIs from March 1996 to May 1998. Main outcome measures: Clinical, virological and immunological outcomes, and adherence to therapy after 12 months. Results: Baseline median CD4 cell count and viral load were 160 cells/mm3 and 5 log10 copies/ml, respectively. After 48 weeks, 168 patients (81%) reached plasma HIV-RNA levels below 400 copies/ml, and the mean increase in CD4 cell count was 196 cells/mm3. Clinical events were observed in 29 patients (14%) after a median time of 100 days on therapy, yet mortality was extremely low (0.9%). By multivariate analysis, adherence over 90% [relative risk (RR), 16.66; 95% confidence interval (Cl), 5.26-50; P=0.00001] and AIDS diagnosis at baseline (RR 0.35; 95% Cl, 0.14-0.90; P=0.02) were the strongest predictors for virological suppression. An immunological recovery over 100 cells/mm3 was significantly associated with an initial virological response (RR 2.94; 95% Cl, 1.31-6.66; P=0.009) and adherence over 90% (RR 3.44; 95% Cl, 1.61-7.69; P=0.005). There were high rates of change with the first PI (40%), mostly due to adverse events (51%), but it did not compromise long-term effectiveness. Conclusions: Initial PI treatment in the clinical setting is able to reach equally good outcomes as those found in controlled trials. Changes in therapy due to toxicity do not compromise a successful outcome, which clearly depends on an adequate adherence to therapy.