HIV drug resistance, early treatment outcomes and impact of guidelines compliance after protease inhibitor-based second-line failure in a dedicated resistance clinic in western Kenya: a retrospective cohort study.

INTRODUCTION

Data on drug resistance, viral outcomes and guidelines compliance following protease inhibitor (PI)-based second-line failure in low- and middle-income countries are limited, particularly in the era of dolutegravir-containing antiretroviral therapy (ART).

METHODS

We conducted a retrospective cohort study of people living with HIV (PLWH) ≥3 years old with second-line viral failure (VF, ≥1000 copies/ml) at the Academic Model Providing Access to Healthcare from 2011 to 2021. We assessed resistance prevalence and patterns at second-line VF, stratified by PI (atazanavir/ritonavir or lopinavir/ritonavir), and examined correlations of resistance and treatment strategies with VF at 6-18 months post-genotype. Analyses employed inverse probability weighting, adjusting for calendar year, age, gender, ART duration, PI at genotyping and class-specific resistance, and considered guidelines-supported versus unsupported strategies.

RESULTS

Of 187 participants (median age 41 years, 54% female, 41% on atazanavir/ritonavir, 59% on lopinavir/ritonavir-based ART), 91% had any resistance (NRTI 79%, NNRTI 80%, major PI 37%, dual-class 36%, triple-class 37%). Predicted resistance to third-line options was 67% for etravirine or rilpivirine and 10% for darunavir/ritonavir. Despite higher resistance detected on atazanavir/ritonavir versus lopinavir/ritonavir, predicted darunavir/ritonavir resistance was similar. At median 9 months post-genotype, 95% of 173 participants with available data were on a guidelines-supported regimen (55% second-line; 45% third-line, 86% dolutegravir-based), of whom 28% had post-genotype VF. Of the 5% not on guidelines-supported regimens, 71% had post-genotype VF. Adjusted odds of VF were higher for guidelines-unsupported versus supported regimens (OR = 4.52; 95% CI 1.02-26.24), and odds of VF were 97% lower for those on third-line versus second-line (OR = 0.07; 95% CI 0.02-0.20).

CONCLUSIONS

We found high levels of drug resistance and early VF following PI-based second-line failure in Kenya. Treatment guidelines compliance and switches to third-line, even within guidelines recommendations, improved early viral outcomes. Findings highlight the vulnerability of PLWH with advanced ART experience and resistance profiles, and the importance of following guidelines and improving access to third-line and drug resistance testing, particularly in the new ART era.