First-line HIV treatment outcomes following the introduction of integrase inhibitors in UK guidelines

Objective To investigate the characteristics and outcomes of people who initiated different antiretroviral therapy (ART) regimens during the era of integrase strand transfer inhibitors (INSTIs). Design UK-based observational cohort study. Methods UK Collaborative HIV Cohort study participants were included if they had started ART between 1 January 2012 and 30 June 2017. Virological failure was defined as the first of two consecutive plasma HIV RNA more than 50 copies/ml, at least 6 months after starting ART. Follow-up was censored at ART discontinuation, class switch or death. The risk of virological failure among those on INSTI, protease inhibitor or nonnucleoside reverse transcriptase inhibitor (NNRTI) regimens was compared using Kaplan-Meier and Cox regression methods. Results Of 12 585 participants, 45.6% started a NNRTI, 29.0% a protease inhibitor and 25.4% an INSTI regimen. Over a median follow-up of 20.3 months (interquartile range 7.9-38.9), 7.5% of participants experienced virological failure. Compared with those starting an NNRTI regimen, people receiving INSTIs or protease inhibitors were more likely to experience virological failure: INSTI group adjusted hazard ratio 1.52, 95% confidence interval 1.19-1.95, P = 0.0009; protease inhibitor group adjusted hazard ratio 2.70, 95% confidence interval 2.27-3.21, P less than 0.0001, likelihood ratio test P less than 0.0001. Conclusion First-line INSTI regimens were associated with a lower risk of virological failure than protease inhibitor regimens but both groups were more likely to experience virological failure than those initiating treatment with a NNRTI. There is likely to be residual channelling bias resulting from selected use of INSTIs and protease inhibitors in specific clinical contexts, including in those with a perceived risk of poor adherence.