Pharmacokinetic-pharmacodynamic modelling of MK-2048 in ex vivo cervical tissue
Katherine Kay (1), Sarah Cook (1, 2), Charlene Dezzutti (3, 4), Peter Anderson (5), Lane Bushman (5), Cory Shetler (4), Dana Tirabassi (4), Kenneth Marks (4), Lisa Rohan (3,4), Robert Bies (1)
(1) State University of New York at Buffalo; (2) Current affiliation: Enhanced Pharmacodynamics, LLC; (3) University of Pittsburgh; (4) Magee-Womens Research Institute; (5) University of Colorado
Objectives:
An estimated 1.8 million individuals worldwide acquired new HIV infections in 2016; the equivalent of around 5,000 new infections per day. The development of safe and efficacious HIV prevention strategies is therefore a global health priority. The Film Antiretroviral Microbicide Evaluation (FAME) program aims to incorporate MK-2048 – a potent HIV integrase inhibitor developed by Merck & Co., Inc., Kenilworth, NJ, USA - into a vaginal film formulation capable of providing one week of coitally independent protection from HIV infection from unprotected vaginal intercourse. This work focuses on characterizing the extended pharmacokinetic (PK) and pharmacodynamic (PD) properties of MK-2048 in HIV-1 infected human cervical tissue explants.
Methods:
The extended PK data explant study included three cervical tissue samples (subjects) treated with 100 µM of MK-2048; two tissues per explant per time point. The samples were washed and frozen at 0, 0.5, 1, 6, 24, 48, 72, 96 & 120 hours. Homogenate tissue concentrations were fitted to a PK model using non-linear mixed-effects modelling implemented in NONMEM V7.3.0 [1]; initial amounts of MK-2048 were set to concentrations observed in 0-hour samples. The PK-PD explant study included five tissues (subjects) with 2-explants per tissue, per treatment group. Explants were infected overnight with HIV-1 in the presence or absence of MK-2048, then washed and serially sampled for HIV-1 p24 antigen at 1, 4, 7-8, 11, 14-15, 17-18 & 21-22 days. For each tissue and treatment group, two additional explants were used for MK-2048 PK measurements on day 1. The viral dynamics of the control group were fitted to a PD model using non-linear mixed-effects modelling implemented in NONMEM V7.3.0; tissue PK parameters were fixed to those fitted in the extended PK model. Different PD growth models were tested including linear, exponential, simeoni (linear followed by exponential growth and transit compartments for damage phase) [2], signal distribution (with a differing number of transit compartments) [3] and the Perelson model of HIV dynamics [4].
Results:
The PK of MK-2048 was well characterized by a 2-compartment model with linear elimination. The coefficient of variation (CV) for the residual variability was 19% within replicates and 15% CV was found for other sources. The between-subject variability on drug elimination rate was 1008% CV with a 67% relative standard error. Viral growth for the control group (i.e. samples treated with HIV only) was best characterized using the Perelson model including uninfected target cells, infected cells and free virus.
Conclusions:
The extended PK model predicted MK-2048 drug concentrations in cervical tissue and the Perelson model best described viral dynamics in the control group of the human explant tissue. The final extended PK and control PD models informed the PK-PD modelling process of this new compound and the HIV-1 viral dynamics in human explant studies. The results of this work can provide suggestions on the concentration of MK-2048 required to prevent viral replication in humans. They can be incorporated in to physiologically-based pharmacokinetic (PBPK) models of vaginally administered drugs to predict patient concentration time profiles in vivo and used to develop PBPK-PD models able to predict whether MK-2048 is likely to prevent new, sexually-transmitted, HIV-1 infections.
References:
[1] Beal SL, Sheiner LB, Boeckmann AJ & Bauer RJ (Eds.) NONMEM Users Guides. 1989-2011. Icon Development Solutions, Ellicott City, Maryland, USA
[2] Simeoni, M., et al. (2004). "Predictive pharmacokinetic-pharmacodynamic modeling of tumor growth kinetics in xenograft models after administration of anticancer agents." Cancer Res 64: 1094-1101.
[3] Yang, J., et al. (2010). "Comparison of two pharmacodynamic transduction models for the analysis of tumor therapeutic responses in model systems." AAPS J 12(1): 1-10.
[4] Perelson, A. S. and R. M. Ribeiro (2008). "Estimating drug efficacy and viral dynamic parameters: HIV and HCV." Stat Med 27(23): 4647-4657.