Background We aimed to mimic the ACTG 5257 trial, comparing raltegravir (RAL), ritonavir-boosted atazavavir (ATV/r) and ritonavir-boosted darunavir (DRV/r) in the observational setting. Methods All the ICONA patients starting a first cART with 2NRTI + ATV/r, DRV/r or RAL were included. Primary end-point was treatment failure, i.e. virological failure (confirmed HIV-RNA > 200copies/mL > 6 months therapy) or discontinuation for any reason of the third drug. Secondary end-points: virological failure50 (50 copies/mL threshold), and discontinuation of the third drug due to intolerance/toxicity. Cox regression analyses were run to compare the risk of outcomes between the three regimens. Results 2249 patients were included, 985 (44%) initiated ATV/r, 1023 (45%) DRV/r and 241 (11%) RAL; median follow-up of 3.6 years (IQR: 2.3-5.2). After controlling for baseline confounding factors, patients given ATV/r showed a 26% higher risk of treatment failure (TF) vs. DRV/r (AHR 1.26, 95%CI 1.11-1.43); patients on RAL had a lower risk of TF vs. ATV/r (AHR 0.81, 95%CI 0.66-0.99). The probability of virological failure50 was significantly lower for people initiating RAL vs. DRV/r (AHR 0.46, 95%CI 0.24-0.87) or ATV/r (AHR 0.52, 95%CI 0.27-0.99). In addition, RAL was associated to a lower risk of discontinuation for toxicity vs. both DRV/r (AHR: 0.37, 95%CI: 0.19-0.72) and ATV/r (AHR: 0.18, 95%CI: 0.09-0.34). ATV/r was associated with a higher risk of discontinuing due to toxicity (AHR 2.09, 95%CI 1.63-2.67) vs. DRV/r. Conclusions In our observational study, we confirmed higher risk of treatment failure and lower tolerability of ATV/r-based regimens as compared to those including DRV/r or RAL.
Keywords: Cohort study; antiretroviral regimens; boosted-atazanavir; boosted-darunavir; raltegravir; therapy discontinuation.