Population PK Modeling of Dapivirine Released from Vaginal Rings
Stefan Zeiser (1), Willem Hettema (1), Kees Bol (2), Paul van den Berg (1), Annalene Nel (3), Jeremy Nuttall (3), Edwin Spaans (1)
(1) Kinesis Pharma BV; (2) Kinesis Singapore Pte Ltd; (3) International Partnership for Microbicides, Silver Spring, MD, USA
Objectives: Dapivirine is a non-nucleoside reverse transcriptase inhibitor with potent antiviral activity against HIV-1. International Partnership for Microbicides has developed a vaginal ring containing dapivirine (25 mg) to protect women against HIV infection through sexual intercourse. These rings are placed in the upper third of the vagina where they slowly release the active drug. A population PK model was built to describe the time course of vaginal fluid levels of dapivirine at the introitus and cervix. The goal was to predict dapivirine concentrations after various intervals of ring removal, and to predict vaginal fluid concentrations when the rings are inserted successively every four weeks.
Methods: Non-linear mixed effects PK modeling using NONMEM was performed based on vaginal fluid dapivirine concentrations at the area of the introitus and cervix. The model consists of one compartment with a first order drug elimination process. Release and distribution of dapivirine to the introitus and cervix are described by a first and zero order process. The model characterizes the fraction of dapivirine that is distributed to the introitus and cervix, and the fraction that is released by zero and first order processes.
Results: Besides a small bias at the early release phase and during the washout phase, the model described all data well. Fixed effects parameters could be estimated with high precision (CV<18%). Inter-individual variability (IIV) was determined to be significant on all model parameters except for the parameter which determines the fraction of drug that is distributed by a first order process. Except for the volume of distribution (CV=60.4%), IIV could be estimated with good precision (CV<42%). The goodness of fit plots and visual predictive checks revealed no indications of additional bias or any model misspecification. Simulations showed that, when the ring was removed after a period of four weeks, the vaginal fluid concentrations at both locations dropped to values of about 60, 45, and 20% of the level prior to ring removal at 8, 12, and 24h after removal, respectively. Simulation of successive 4-weekly ring replacement showed virtually no accumulation of dapivirine in vaginal fluid levels.
Conclusions: Despite the high variability in the data, the PK of dapivirine in vaginal fluids could be described adequately by the presented model. The high precision in parameter estimates allows the use of this model as a potential tool for predicting exposure in several ring application scenarios.
