(463e) An in silico Tool for Quantitative Kinetic Predictions of API Degradation

Pharmaceuticals are susceptible to oxidation and hydrolysis that often lead to potency loss and the formation of harmful impurities. To ensure drug product efficacy and quality, it is critical to elucidate degradation pathways and products of active pharmaceutical ingredients (APIs). Contemporary approaches to studying API stability rely on forced API degradation experiments. In a typical API degradation study, researchers conjecture API degradation pathways and products at first and then verify the hypotheses experimentally. This trial-and-error approach is inefficient as API degradation experiments are time-consuming and common analytic methods (e.g., HPLC-MS, NMR) consume significant amounts of API, which is expensive to synthesize and purify especially at early stages of drug R&D. Moreover, researchers often have limited intuition of what degradants to expect for a new API, making HPLC-MS analyses challenging.

Alternatively, supplementing the contemporary experimental approach with novel computational chemistry methods has the potential to significantly accelerate API stability studies and reduce R&D costs. In this talk, we present the development of a self-improving software that can readily predict API degradants, automatically build API degradation kinetic models, and reliably identify critical API decomposition pathways. Pfizer and MIT researchers collaboratively investigated the aqueous oxidative degradation chemistry of imipramine. The imipramine degradation model constructed by the MIT team using the newly developed software successfully predicted degradants observed experimentally by the Pfizer team. The model and experimental results also provide novel insights into imipramine degradation pathways and kinetics.

Based on the promising results, we believe the quantitative in silico tools that we are actively developing will greatly assist the pharmaceutical research community in performing drug stability studies. The computer-assisted drug stability analyses will guide future drug degradation experiments, assist in API formulation improvements to resist oxidation, and potentially reduce drug development costs.