To evaluate the frequency of discontinuation of the first highly active antiretroviral regimen (HAART) and the factors predictive of discontinuing for toxicity and failure in a population-based cohort of HIV-positive individuals in Italy, naïve from antiretrovirals at enrolment. The study population consisted of individuals who initiated HAART and had at least one follow-up visit. The primary end-points were discontinuation of any component of HAART for drug toxicity and discontinuation for failure. Survival analyses were performed to identify predictive factors for reaching the two end-points. Eight hundred and sixty-two individuals initiated HAART; in 727 of them (84.3%) this consisted of two nucleoside reverse transcriptase inhibitors (NRTI) and one protease inhibitor (PI). Over a median follow-up of 45 weeks, 312 patients (36.2%) discontinued therapy: 182 (21.1%) discontinued due to toxicity, 44 (5.1%) due to failure. The probability of discontinuing HAART at 1 year was 25.5% [95% confidence interval (CI), 21.9-28.9] due to toxicity and 7.6% (95% CI, 4.9-1 0.3) due to failure. Independent factors associated with discontinuation for toxicity were: gender [relative hazard (RH) = 0.51; 95% CI, 0.32-0.80 for men versus women], type of treatment (indinavir-containing regimens, RH = 1.94; 95% CI, 1.10-3.41 and ritonavir-containing regimens, RH = 3.83; 95% CI, 2.09-7.03 versus hard-gell saquinavir) and time spent on treatment (RH = 0.89; 95% CI, 0.80-0.98 for each additional month). Discontinuation due to failure was independently associated with the most recent HIV-RNA (RH = 3.20; 95% CI, 1.74-5.88 for log10 copies/ml higher), and with type of treatment (indinavir-containing regimens, RH = 0.21; 95% CI, 0.06-0.78 and ritonavir-containing regimens, RH = 0.23; 95% CI, 0.04-1.26 versus hard-gell saquinavir). If the current HAART regimen caused no toxicity, less than 10% of naïve patients discontinue their first HAART regimen because of failure after 1 year from starting therapy.
