Predicting In Vivo Efficacy from In Vitro Data: Quantitative Systems Pharmacology Modeling for an Epigenetic Modifier Drug in Cancer

Abstract

Reliably predicting in vivo efficacy from in vitro data would facilitate drug development by reducing animal usage and guiding drug dosing in human clinical trials. However, such prediction remains challenging. Here, we built a quantitative pharmacokinetic/pharmacodynamic (PK/PD) mathematical model capable of predicting in vivo efficacy in animal xenograft models of tumor growth while trained almost exclusively on in vitro cell culture data sets. We studied a chemical inhibitor of LSD1 (ORY‐1001), a lysine‐specific histone demethylase enzyme with epigenetic function, and drug‐induced regulation of target engagement, biomarker levels, and tumor cell growth across multiple doses administered in a pulsed and continuous fashion. A PK model of unbound plasma drug concentration was linked to the in vitro PD model, which enabled the prediction of in vivo tumor growth dynamics across a range of drug doses and regimens. Remarkably, only a change in a single parameter—the one controlling intrinsic cell/tumor growth in the absence of drug—was needed to scale the PD model from the in vitro to in vivo setting. These findings create a framework for using in vitro data to predict in vivo drug efficacy with clear benefits to reducing animal usage while enabling the collection of dense time course and dose response data in a highly controlled in vitro environment.

Growth‐rate model predicts in vivo tumor response from in vitro data

“…Other published approaches for predicting in vivo drug efficacy can be classified in two groups: complex models of drug‐tumor interactions that rely on target engagement and molecular understanding of cellular growth or machine learning methods that require extensive training data. Overall, these approaches yielded predictions with correlation values up to 0.5, 2 , 7 , 12 , 20 , 21 , 24 which are lower than our method. In addition, our approach strikes a pragmatic balance in terms of data requirements: our growth‐rate model relies on end point in vitro phenotypic data instead of direct measure of drug‐target interaction or time‐course experiments, 12 and does not require extensive training data to be as predictive as other approaches.…”

“…Overall, these approaches yielded predictions with correlation values up to 0.5, 2 , 7 , 12 , 20 , 21 , 24 which are lower than our method. In addition, our approach strikes a pragmatic balance in terms of data requirements: our growth‐rate model relies on end point in vitro phenotypic data instead of direct measure of drug‐target interaction or time‐course experiments, 12 and does not require extensive training data to be as predictive as other approaches. 24 Therefore, our approach is more widely applicable and can potentially be extended to drug combinations 14 , 18 , 47 , 48 or new drugs with limited knowledge of their cell‐intrinsic effects.…”

“…Methods to improve on the prediction of in vivo and clinical responses based on in vitro data can be broadly categorized into mechanistic or statistical models. First, drug effects are modeled at the molecular level, accounting for known underlying biochemical processes in vivo 2 , 11 , 12 , 13 , 14 , 15 , 16 , 17 as reviewed in Danhof et al 2 Other mathematical models have focused on population dynamics to predict how scheduling of dosing regimens and combinations of therapies can improve efficacy and address the emergence of resistance. 18 , 19 Second, statistical and machine learning methods aim to leverage large datasets collected on various cancer cell lines, animal studies, and clinical data.…”

Natural Killer Cell-Derived Extracellular Vesicles as a Promising Immunotherapeutic Strategy for Cancer: A Systematic Review

“…Thus, the use of mere in vitro data to show the anti-cancer potential of NK-EVs is totally insufficient, as the model does not replicate the physiological complexities that make up a living organism [ 62 ]. For more effective translation, a highly defined and complex model of pharmacodynamics (PD) and pharmacokinetics (PK) is needed, such as the one constructed by Bouhaddou et al (2020) [ 63 ]. In term of cancer, the main debacle that has stumped the progression of many medical innovations is the existence of drug resistance, which is very difficult to replicate in vitro [ 64 ].…”

Recent applications of quantitative systems pharmacology and machine learning models across diseases

“…The model was able to predict in vivo drug efficacy extrapolated exclusively from in vitro data. Such a mechanistic approach could reduce the use of animal models, the cost and time in drug development [ 45 ]. In another study that addresses drug therapies, Stroh et al model the activatable antibody, Probody therapeutic (Pb-Tx), designed to keep the antigen-binding site of engineered antibody masks until local proteolytic activation in disease tissue.…”