Integration of In Silico Degradation Prediction, Mechanistic Stress Testing, and Accelerated Stability Modeling for...

A stability-assessment framework integrating Zeneth in silico degradation prediction, AIBN-mediated radical autoxidation, and ASAPprime® accelerated kinetic modeling was developed for olaparib drug substance and a prototype immediate-release tablet. Zeneth predicted benzylic oxidation as the dominant degradation pathway (pathway score 912). Under AIBN stress, Impurity 6 formed at 2.15%, with radical-initiator dependence confirmed by an AIBN-free control (< 0.1% total impurity); aqueous hydrogen peroxide (1.0%) provided degradant mass for LC-MS/MS characterization. Impurity 6 was confirmed as the benzylic ketone (C₂₄H₂₁FN₄O₄) against an authenticated reference standard (Schymanski Level 1; mass error + 0.7 ppm); the reprocessed HRMS/MS fragmentation ladder was fully reconciled with this assignment, and Impurity 6 was chromatographically resolved from its isomeric Impurity 4 (relative retention times 1.45 versus 1.02). Isoconversion ASAPprime® studies (drug substance: 50-80 °C/11-75% RH; formulation: 40-70°C/11-75% RH) were validated against 12-month real-time data. The drug substance followed first-order kinetics (Ea = 91.2 ± 6.7 kJ·mol⁻¹) with negligible humidity sensitivity (B = 0.004 ± 0.003; p > 0.05), yielding a model-derived 95% confidence shelf-life estimate of 3.4 years (MAPD = 5.3%; Q² = 0.96). The formulation showed accelerated kinetics (Ea = 85.4 ± 5.9 kJ·mol⁻¹; B = 0.014 ± 0.005; p < 0.05) and a shelf-life estimate of 2.6 years (MAPD = 5.2%; Q² = 0.93). These projections require confirmation by longer-term real-time data. The workflow provides a development stage stability risk assessment tool; Impurity 6 was structurally confirmed at Level 1 against an authenticated reference standard, and the formulation was single batch and unoptimized.