HIV-1 virologic failure in the RESINA cohort: lessons from two decades of real-world data

To quantify virologic failure (VF), identify predictors, characterize resistance patterns at failure, and evaluate time to resuppression in the RESINA cohort.

ART-naïve adults initiating ART in 2001-2024 were followed. VF was defined as at least one HIV-1 RNA > 200 copies/mL after suppression or ≥ 0.5-log₁₀ rebound. Participants were grouped by treatment era (2001-2007, 2008-2013, ≥ 2014), reflecting availability of drug classes. Genotypes at baseline and VF were interpreted using the HIV-GRADE algorithm. Predictors of VF were assessed with logistic regression; time to resuppression (< 50 copies/mL) after first VF with Cox models and Kaplan-Meier plots.

Among 5136 participants, 139 (2.7%) had VF; rates declined across eras (4.7%, 2.6%, 1.7%). Independent predictors were injection-drug use (OR 1.74), CD4 < 200/µL (OR 2.32), and ART start in 2001-2007 (OR 1.95); MSM acquisition was protective (OR 0.32). At failure, 36 patients showed resistance, often multiclass (61%); INSTI resistance was rare (n = 5). After first VF, 122/139 cases resuppressed (median 147 days). Male sex predicted faster resuppression (HR 1.81); higher failure VL trended to slower resuppression (HR 0.84 per log₁₀). INSTI-based switches consistently achieved resuppression in descriptive analyses and were not associated with multiclass resistance.

VF was uncommon and declined over time, reflecting improved regimen potency and tolerability. Failures were associated with late presentation and IDU, consistent with adherence barriers. Resistance often involved multiple classes, while INSTI resistance remained infrequent. Early, genotype-guided optimization, preferably to INSTI-based therapy, combined with targeted adherence support may improve outcomes.