Objective: To assess predictive factors of long-term immune restoration in patients who started protease inhibitor (PI)-based HAART and experienced virological rebound after initial complete success.
Method: A retrospective longitudinal analysis of all HIV-infected patients who started their first PI-based HAART and reached viral load below 500 copies/mL was carried out in a large academic center in Italy. Patients were classified either as complete virologic responder (CR) or rebounders (REB) when confirmed plasma viremia was detected thereafter. Immunological outcome was the area under the curve (AUC) of the absolute CD4+ cell count change since the 8th month after treatment initiation (CD4+ T-cell AUC). Association between baseline characteristics, virological outcome, and CD4+ T-cell AUC was assessed by univariate and multivariate analysis.
Results: There were 374 patients who were included in the study. Mean follow-up was 30.2 months. There were 226/374 patients (60.4%) who remained CR while 148/374 (39.6%) presented at least one rebound (REB). Among REB patients, complete viral suppression was regained in 15/42 (35.7%) and 50/106 (47.1%) patients who underwent therapy changes or not, respectively. When multiple linear regression was carried out, previous nucleoside reverse transcriptase inhibitor (NRTI) experience and baseline CD4+ cell count below 350 cells/muL did not impair long-term immune restoration. The occurrence of rebound, its duration (> 18 months), and its magnitude (peak of viral load > 10,000 copies/mL) were independent negative prognostic factors.
Conclusion: The occurrence of viral rebound is independently associated with significantly impaired long-term immunological restoration. The magnitude of viral rebound (< 10,000 copies/mL) and its duration (< 18 months) may be useful to identify those rebounding patients who may still profit from maintaining the current failing therapy if a more aggressive approach may be expected to be deleterious for tolerability reasons or lack of options.