Applying the CONSENSUS ‐ LAI virological failure definition in real‐world practice: Evidence from the multicentre LAHIV cohort

Abstract Objectives Definitions of virological failure (VF) for long‐acting cabotegravir plus rilpivirine (CAB+RPV) remain inconsistent across real‐world cohorts. The CONSENSUS‐LAI Delphi process (September 2025) set a shared plasma threshold of ≥200 copies/mL. We evaluated its impact on VF classification and whether its publication influenced prescriber behaviour. Methods Retrospective analysis of virologically suppressed adults initiating CAB+RPV across 11 centres in Tuscany, Italy. Switches attributed to VF by the treating physician were reclassified as meeting CONSENSUS‐LAI plasma criteria or as sub‐threshold. Rates were calculated per 100 person‐years (PY); switches before and after September 2025 were compared with Fisher's exact test. Results Among 297 participants (418 PY), 11 switches were attributed to VF (2.63/100 PY, 95% CI 1.31–4.71). Six (54.5%) met CONSENSUS‐LAI plasma criteria (1.44/100 PY, 95% CI 0.53–3.12), five with a single value >1000 copies/mL, one with confirmed values ≥200 copies/mL; 4 were sub‐threshold and one was cerebrospinal fluid viral escape with undetectable plasma HIV‐1 RNA. Emergent resistance occurred in three cases, all CONSENSUS‐LAI‐defined failures. No resistance was detected in the two sub‐threshold cases tested; all four sub‐threshold VF re‐suppressed, although two of four were switched to a protease inhibitor‐based regimen. Sub‐threshold switches were 1/6 before and 3/5 after CONSENSUS‐LAI publication ( p = 0.24). Conclusions The CONSENSUS‐LAI definition nearly halves the VF rate, aligning it with cohorts using a ≥200 copies/mL threshold. Emergent resistance was confined to CONSENSUS‐LAI‐defined failures, whereas sub‐threshold viraemia, without detected resistance, still drove most post‐consensus switches, highlighting the challenges of translating consensus recommendations into routine practice.