Further Considerations Towards an Effective and Efficient Oncology Drug Discovery DMPK Strategy

Multi-task convolutional neural networks for predicting in vitro clearance endpoints from molecular images

Optimization of compound metabolic stability is a highly topical issue in pharmaceutical research. Accordingly, application of predictive in silico models can potentially reduce the number of design-make-test-analyze iterations and consequently speed up the progression of novel candidate molecules. Herein, we have investigated the question if multiple in vitro clearance endpoints could be accurately predicted from image-based molecular representations. Thus, compound measurements for four commonly investigated clearance endpoints were curated from AstraZeneca internal sources, providing a sound basis for building multi-task convolutional neural network models. Application of several increasingly challenging data splitting strategies confirmed that convolutional neural network models were successful at capturing implicit chemical relationships contained in training and test data, similar to what is commonly observed for structural fingerprints. Furthermore, model benchmarking against state-of-the-art machine learning methods, including deep neural networks and graph convolutional neural networks, trained with structure- and graph-based representations, respectively, revealed on par or increased accuracy of convolutional neural networks with clear benefit of multi-task learning across all clearance endpoints. Our findings indicate that image-based molecular representations can be applied to predict multiple clearance endpoints, suggesting a potential follow-up to investigate model interpretability from molecular images.

Identification and Strategies to Mitigate High Total Clearance of Benzylamine-Substituted Biphenyl Ring Systems

For most oral small-molecule projects within drug discovery, the extent and duration of the effect are influenced by the total clearance of the compound; hence, designing compounds with low clearance remains a key focus to help enable sufficient protein target engagement. Comprehensive understanding and accurate prediction of animal clearance and pharmacokinetics provides confidence that the same can be observed for human. During a MERTK inhibitor lead optimization project, a series containing a biphenyl ring system with benzylamine meta-substitution on one phenyl and nitrogen inclusion as the meta atom on the other ring demonstrated multiple routes of compound elimination in rats. Here, we describe the identification of a structural pharmacophore involving two key interactions observed for both the MERTK program and an additional internal project. Four strategies to mitigate these clearance liabilities were identified and systematically investigated. We provide evidence that disruption of at least one of the interactions led to a significant reduction in CL that was subsequently predicted from rat hepatocytes using in vitro/in vivo extrapolation and the well-stirred scaling method. These tactics will likely be of general utility to the medicinal chemistry and DMPK community during compound optimization when similar issues are encountered for biphenyl benzylamines.

Prediction of In Vivo Pharmacokinetic Parameters and Time-Exposure Curves in Rats Using Machine Learning from the Chemical Structure

Animal pharmacokinetic (PK) data as well as human and animal in vitro systems are utilized in drug discovery to define the rate and route of drug elimination. Accurate prediction and mechanistic understanding of drug clearance and disposition in animals provide a degree of confidence for extrapolation to humans. In addition, prediction of in vivo properties can be used to improve design during drug discovery, help select compounds with better properties, and reduce the number of in vivo experiments. In this study, we generated machine learning models able to predict rat in vivo PK parameters and concentration-time PK profiles based on the molecular chemical structure and either measured or predicted in vitro parameters. The models were trained on internal in vivo rat PK data for over 3000 diverse compounds from multiple projects and therapeutic areas, and the predicted endpoints include clearance and oral bioavailability. We compared the performance of various traditional machine learning algorithms and deep learning approaches, including graph convolutional neural networks. The best models for PK parameters achieved R² = 0.63 [root mean squared error (RMSE) = 0.26] for clearance and R² = 0.55 (RMSE = 0.46) for bioavailability. The models provide a fast and cost-efficient way to guide the design of molecules with optimal PK profiles, to enable the prediction of virtual compounds at the point of design, and to drive prioritization of compounds for in vivo assays.

Machine Learning Models for Human In Vivo Pharmacokinetic Parameters with In-House Validation

Prior to clinical development, a comprehensive pharmacokinetic characterization of a novel drug is required to understand its exposure at the site of action and elimination. Accordingly, in vitro assays and animal pharmacokinetic studies are regularly employed to predict drug exposure in humans, which is often costly and time-consuming. For this reason, the prediction of human pharmacokinetics at the point of design would be of high value for drug discovery. Therefore, we have established a comprehensive data curation protocol that enables machine learning evaluation of 12 human in vivo pharmacokinetic parameters using only chemical structure information and available doses for 1001 unique compounds. These machine learning models were thoroughly investigated and validated using both an independent hold-out test set and AstraZeneca clinical data. In addition, the availability of preclinical predictions for a subset of internal clinical candidates allowed us to compare our in silico approach with state-of-the-art pharmacokinetic predictions. Based on this evaluation, three fit-for-purpose models for AUC PO (R_test² = 0.63; RMSE_test = 0.76), C_max PO (R_test² = 0.68; RMSE_test = 0.62), and Vd_ss IV (R_test² = 0.47; RMSE_test = 0.50) were identified. Based on the findings, our machine learning models have considerable potential for practical applications in drug discovery, such as influencing decision-making in drug discovery projects and progression of drug candidates toward the clinic.

Blood retinal barrier and ocular pharmacokinetics: Considerations for the development of oncology drugs

Tyrosine kinase inhibitors (TKIs) are an example of targeted drug therapy to treat cancer while minimizing damage to healthy tissue. In contrast to traditional oncology drugs, the toxicity profile of targeted therapies is less well understood and can include severe ocular adverse events, which are among the most common toxicity reported by these therapeutics. Inhibition of Mer receptor tyrosine kinase (MERTK) promotes innate tumor immunity by decreasing M2-macrophage polarization and efferocytosis. This mechanism offers the opportunity for targeted immunotherapy to treat cancer; however, the ocular expression of MERTK increases the difficulty for developing a targeted drug due to toxicity concerns. In this article we review the pharmacokinetic (PK) parameters and in vitro absorption, distribution, metabolism, and excretion (ADME) assays available to evaluate ocular disposition and assess the relationship between clinical PK and reported ocular events for TKIs to allow backtranslation to preclinical models. Understanding the ocular disposition in the context of PK and safety remains an evolving area and is likely to be a key aspect of developing safe and efficacious oncology drugs, devoid of ocular toxicity.