Background:  Achieving and maintaining viral suppression is one of the primary goals of treatment in HIV patients. Yet, few studies have explored the predictive value of longitudinal adherence data on viral rebound.

Methods:  Individuals in the Swiss HIV Cohort Study (SHCS) on ART, with RNA <50 copies/mL over the previous 3 months and who completed ≥1 adherence questionnaire prior to June 1, 2006 were included. Information on factors, such as adherence and ART regimen, were collected at each semi-annual follow-up visit. Non-adherence was defined as missing ≥1 dose of ART or taking a drug holiday (missing 2 consecutive doses) in the previous 28 days. Viral rebound was defined as two consecutive RNA >50 copies/mL. Cox regression models with time-independent and dependent covariates were used to evaluate time to viral rebound.  

Results:  We analyzed data from 2846 individuals with 12,807 visits (median 5 visits per individual) over a median follow-up time of 2.5 years. Non-adherence was reported on 25.2% of visits. The rate of viral rebound was 20.4%; 23.2% in those who missed ≥1 dose compared with 13.9% with no missed doses; 46.1% in those who reported taking a drug holiday compared with 14.9% in those without. In unadjusted analysis, non-adherence was associated with an increased risk of viral rebound (hazard ratio (HR) 1.60, 95% confidence interval (CI) 1.31 to 1.96). In an adjusted model, non-adherence remained highly significant (HR 1.48, 95%CI 1.19 to 1.83). Variables significantly associated with an increased risk of viral rebound were smoking (HR 1.37, 95%CI 1.12 to 1.69), taking drugs other than ART (HR 1.30, 95%CI 1.01 to 1.68), >5 previous ART regimens (HR 1.34, 95%CI 1.05 to 1.71), dosing frequency (3 times daily compared to twice daily HR 1.89, 95%CI 1.11 to 3.24), and ART regimen (boosted protease inhibitor (PI), HR 1.41, 95%CI 1.08 to 1.84; non-boosted PI, HR 1.46, 95%CI 1.07 to 2.00; compared with non-nucleoside reverse transcriptase inhibitors [NNRTI]). Covariates associated with a decreased risk of viral rebound were time virally suppressed at baseline (HR 0.84, 95%CI 0.79 to 0.90) and time on ART (HR 0.77, 95%CI 0.71 to 0.82).

Conclusions:  A simple self-report adherence questionnaire predicts viral rebound. Non-adherence is associated with a 48% increase in the incidence of viral rebound. After adjusting for non-adherence, several condition- and treatment-related factors—such as dosing frequency, regimen type, and number of previous ART regimens—were significantly associated with an increased risk of viral rebound.