Sulav Duwal, Max von Kleist
Systems Pharmacology & Disease Control, Dep. of Mathematics and Computer Science, Freie University Berlin, Germany
Objectives: Systems Pharmacology aims to understand clinical mechanisms of action (MOA) of drugs to enable optimal therapy. This requires to understand how in vitro testable insights translate into clinical efficacy. Two modelling approaches are used, each on its own insufficient: In a top-down approach, a minimal PK-PD model is fitted to available clinical data, usually enabling limited mechanistic understanding and scalability. A bottom-up approach builds on mechanistic insights derived from in vitro experiments, but may not be representative for the clinical situation. However, a valid bottom-up approach may allow to explore untested clinical scenarios.
We focused on nucleoside reverse transcriptase inhibitors (NRTIs) used in HIV-1 treatment. NRTIs are administered as prodrugs which, after intracellular phosphorylation, exert their effect by competitively inhibiting reverse transcription of the viral genome. Our objective was to assess the validity of previously developed MOA model [1] for NRTIs using a top-down approach.
Methods: We employ bottom-up and top-down approaches concomitantly and predict the clinical potency (IC50 values for inhibition of target cell infection) of lamivudine (3TC), emtricitabine (FTC) and tenofovir (TDF) [3]. In the top-down approach, we established the link between the plasma prodrug PK and the intracellular NRTI-triphosphates using various parameter estimation techniques and subsequently coupled this composite PK model to viral kinetics [2] to estimate the IC50. In the bottom-up approach inhibition of reverse transcriptase-mediated viral DNA polymerisation by the intracellular NRTI-triphosphates is mechanistically modelled [1]. By using disparate datasets to parameterize the respective approaches, we were able to assess the clinical predictive power of the bottom-up approach.
Results: For all NRTIs, the final model consisted of a 2-compartment model with first order absorption with saturable uptake & anabolism. Estimated IC50s (0.17,1.02 and 0.74μM for TDF, FTC and 3TC) from the top-down approach showed good agreement with MOA derived IC50s (0.1,0.82 and 1.72μM). We noted that the top-down IC50s are highly sensitive to the intracellular PK data.
Conclusions: Validating MOA models by top down approaches is an involved, yet rewarding step in Systems Pharmacology. In our case, we were able to translate in vitro parameters into measure of clinical efficacy allowing us to benchmark treatment protocols prior to human trials.
References:
[1] von Kleist M, Metzner P, Marquet R and Schütte C (2012) PLoS Comp. Biol. 8: e1002359
[2] von Kleist M, Menz S and Huisinga W (2010) PLoS Comput. Biol. 6: e1000720
[3] Duwal S, von Kleist M (2016 ) European Journal of Pharmaceutical Sciences doi:10.1016/j.ejps.2016.01.016
Reference: PAGE 25 () Abstr 5730 [www.page-meeting.org/?abstract=5730]
Poster: Drug/Disease modeling - Infection