High rate of long-term clinical events after antiretroviral therapy resumption in HIV-positive patients exposed to antiretroviral therapy interruption.

OBJECTIVE: We analyzed the incidence rate of long-term events in patients on antiretroviral therapy (ART) previously exposed to therapy interruption. DESIGN: A single-center cohort study involving participants in ART interruptions (ARTI) clinical trials (n = 10) was conducted. METHODS: Non-AIDS events after ART resumption were analyzed. A control group not exposed to ARTI was randomly selected from the same cohort and a propensity score of belonging to ARTI group was estimated based on age, sex, CD4+ nadir value, time from HIV diagnosis to ARTI, time from HIV diagnosis to starting ART and time of suppressed viral load, and used to adjust effect estimates. RESULTS: One hundred and eighty-one patients were included, 136 in ARTI and 45 in the control arm. Median time of known HIV-1 infection was 21 years and median time from ART resumption to first non-AIDS event was 5.2 years. A significantly higher proportion of patients with ARTI had an event as compared with control group [raw percentages: 43% (n = 53) vs. 23% (n = 10), P = 0.015]. These differences were confirmed when only the non-AIDS events occurring after ART resumption were analyzed [adjusted hazard ratio (aHR) = 2.43, 95% confidence interval (CI) 1.15-5.12]. The logistic model adjusted for the propensity score indicated that patients with an ARTI had a four-fold higher risk of having at least one non-AIDS event (P = 0.002). CONCLUSION: We found a higher risk of having at least one non-AIDS event years after ART resumption in HIV-infected patients exposed to ARTI as compared with controls. These data should be taken into consideration for future functional cure clinical trials.