Abstract
Tenofovir diphosphate (TFVdp; an active metabolite of oral HIV pre-exposure prophylaxis (PrEP)) is measured in dried blood spots (DBS) to estimate adherence. However, TFVdp's long half-life in whole blood may lead to misclassification following a recent change in adherence. PrEP's other metabolite, emtricitabine triphosphate (FTCtp), has a shorter half-life in whole blood but adherence thresholds are undefined. We characterized DBS TFVdp and FTCtp concentrations across many dosing scenarios. Population pharmacokinetic models were fit to TFVdp and FTCtp DBS concentrations from a directly observed therapy study (NCT03218592). Concentrations were simulated for 90 days of daily dosing followed by 90 days of 1 to 7 doses/week and for event-driven PrEP (edPrEP) scenarios. Thresholds of 1,000 and 200 fmol/punch, for TFVdp and FTCtp, respectively, were reflective of taking 4 doses/week (a minimum target for effective PrEP in men). TFVdp was < 1,000 fmol/punch for 17 days after initiating daily PrEP and > 1,000 fmol/punch for 62 days after decreasing to 3 doses/week. Respectively, FTCtp was < 200 fmol/punch for 4 days and > 200 fmol/punch for 6 days. Accuracy of edPrEP adherence classification depended on duration between last sex act and DBS sampling for both measures with misclassification ranging from 9-100%. These data demonstrate adherence misclassification by DBS TFVdp for 2 months following a decline in adherence, elucidating the need for FTCtp to estimate recent adherence. We provide proof of principle that individualized interpretation is needed to support edPrEP adherence monitoring. Our collective approach facilitates clinicians' ability to interpret DBS results and administer patient-centric interventions.