Interpreting the reasons for the choice and changing of two drug regimens in an observational cohort: comparison of a ritonavir‐boosted protease inhibitor‐based versus a nonnucleoside reverse transcriptase inhibitor‐based first‐line regimen

• Published in: HIV medicine Vol. 15; no. 9; pp. 547 - 556
• Main Authors: Jarrin, I, Hernández‐Novoa, B, Alejos, B, Santos, I, Lopez‐Aldeguer, J, Riera, M, Gutiérrez, F, Rubio, R, Antela, A, Blanco, JR, Moreno, S
• Format: Journal Article
• Language: English
• Published: England 01.10.2014
• Subjects: Adult; Age Factors; Anti-HIV Agents - administration & dosage; antiretroviral therapy; Benzoxazines - administration & dosage; CD4 Lymphocyte Count; cohort study; Coris; Drug Administration Schedule; Drug Therapy, Combination; Female; HIV Infections - drug therapy; HIV Infections - epidemiology; HIV Infections - immunology; HIV Protease Inhibitors - administration & dosage; HIV/AIDS; Human immunodeficiency virus; Humans; Male; persistence; Prospective Studies; response; Ritonavir - administration & dosage; RNA, Viral; Spain - epidemiology; Treatment Outcome; Viral Load; Spain; HIV/AIDS; persistence; cohort study; antiretroviral therapy; response
• ISSN: 1464-2662; 1468-1293; 1468-1293
• DOI: 10.1111/hiv.12144

Objectives We compared reasons for the choice of regimen, time to and reasons for third drug modification, virological response and change in CD4 T‐cell counts in patients started on atazanavir/ritonavir (ATV/r)‐ vs. efavirenz (EFV)‐based first‐line regimens. Methods We included patients from the Cohort of the Spanish HIV Research Network (CoRIS), a multicentre cohort of HIV‐positive treatment‐naïve subjects, in the study. We used logistic regression to assess factors associated with choosing ATV/r vs. EFV, proportional hazards models on the subdistribution hazard to estimate subdistribution hazard ratios (sHRs) for third drug modification, logistic regression to estimate odds ratios (ORs) for virological response and linear regression to assess mean differences in CD4 T‐cell count increase from baseline. Results Of 2167 patients, 10.7% started on ATV/r. ATV/r was more likely than EFV to be prescribed in injecting drug users [adjusted OR 1.85; 95% confidence interval (CI) 1.03–3.33], in 2009–2010 (adjusted OR 1.63; 95% CI 1.08–2.47) and combined with abacavir plus lamivudine (adjusted OR 1.53; 95% CI 0.98–2.43). Multivariate analyses showed no differences, comparing ATV/r vs. EFV, in the risk of third drug modification (sHR 1.04; 95% CI 0.74–1.46) or in virological response (OR 0.81; 95% CI 0.46–1.41); differences in mean CD4 T‐cell count increase from baseline were at the limit of statistical significance (mean difference 29.8 cells/μL; 95% CI −4.1 to 63.6 cells/μL). In patients changing from EFV, 48% of changes were attributable to toxicity/adverse events, 16% to treatment failure/resistance, 3% to simplification, and 8 and 12%, respectively, to patients' and physicians' decisions; these percentages were 24, 6, 12, 14 and 24%, respectively, in those changing from ATV/r. Conclusions ATV/r‐ and EFV‐based regimens meet the requirements of both efficacy and safety for initial combination antiretroviral regimen, which relate to better durability.