Individualized Adherence Benchmarks for HIV Pre-Exposure Prophylaxis

Abstract

Tenofovir diphosphate (TFV-DP) concentrations measured with dried blood spots (DBS) can be used to classify adherence to emtricitabine/tenofovir disoproxil fumarate (F/TDF) for HIV pre-exposure prophylaxis (PrEP). A TFV-DP of 700 fmol/punch was previously associated with high PrEP efficacy, and was estimated to represent ≥4 doses/week on average. However, interindividual variability in TFV-DP concentrations may lead to adherence misclassification and decrease the precision of adherence-efficacy relationships. The purpose of this analysis was to evaluate sources of TFV-DP variability to improve the precision of TFV-DP for adherence assessments by incorporating individual characteristics. Data and samples from a 36-week study of TFV-DP in DBS, collected biweekly, among 48 HIV-negative volunteers (25 Females/26 Caucasian/10 African American/14 Hispanic) receiving F/TDF at 33%, 67%, and 100% of daily dosing under directly observed therapy were used for analysis. The simplest pharmacokinetic model to describe TFV-DP accumulation with acceptable performance was a one-compartment constant input model. Covariates, including laboratory values and demographics were ranked in importance of their association with post hoc pharmacokinetic (PK) parameters using random forest analyses. Weight and platelet count were included in the final model and simulations were conducted to generate benchmarks for <2, 2-3, 4-5, and 6-7 doses/week. Based on these simulations, the previously established protective TFV-DP concentration of ≥700 fmol/punch was observed in those taking 2-3 (in individuals ≤110 kg) and ≥4 (in individuals >110 kg) doses/week, amounting to a much lower rate of misspecification (17% vs. 30%) with this individualized model versus previous interpretations. Incorporating body weight and platelet count improved the precision of TFV-DP concentrations for adherence assessments. Previous benchmarks were conservative, indicating that the pharmacological forgiveness of F/TDF may be higher than currently recognized and supports continued investigation of intermittent PrEP dosing regimens. Clinical Trial Registration number, NCT02022657.

Keywords: HIV; adherence; pharmacokinetics; pre-exposure prophylaxis.

Conflict of interest statement

P.L.A. received contract and research funding and donated study medication from Gilead Sciences (paid to his institution). J.J.K. received research funding (paid to her institution) from Gilead Sciences and donated study medication from Gilead Sciences for an NIH-sponsored study. M.E.I., J.R.C.-M., K.M.B., J.Y., L.S., and L.B. have none to declare.

Figures

FIG. 1.

DOT-DBS study dosing schema.

FIG. 2.

Effects of continuous (5th and 95th percentile) and categorical covariate values on constant input clearance.