Leveraging AQbD Principles for Development of Challenging Drug Substance Stability-Indicating Methods

Abstract

Drug substance stability-indicating methods (SIMs) must undergo a litany of optimization and robustness stresses to ensure that the methods can be successfully validated, transferred, and routinely executed within a variety of different laboratories. These studies typically expose limitations with the methods that require flexibility and strategic development to address. Incorporation of an analytical quality by design (AQbD) strategy leverages risk assessments and a thorough evaluation of critical method parameters to achieve the desired goala method that delivers robust, high-quality data. Herein, we describe the development, challenges, and solutions implemented for a pair of pharmaceutical drug substance liquid chromatography SIMs. In this case study, the analytical target profile was defined, followed by an iterative approach to method development, which consisted of automated stationary phase and mobile phase screenings, followed by software-based optimization and robustness evaluations. Aspects of the collective tracked components (e.g., low UV-response, poor solute retention, and amine-rich compounds) presented several challenges for this case study, but were addressed through nontraditional method conditions, such as divided primary and supplemental methods and carefully controlled concentration of a chaotropic mobile phase modifier. Due to timeline constraints, impurities designated as critical were refined in parallel with analytical method development. New information from a forced degradation study (i.e., coelution of an oxidative degradant with the active pharmaceutical ingredient) triggered a method redesign, which was rapidly achieved by leveraging the knowledge gained during earlier method development. The optimized methods were then tested empirically through validation-level chromatographic robustness experiments and critical method parameters were identified. Altogether, this workflow enabled thorough characterization of many possible conditions so robust, validatable, and easily transferrable methods could be generated to support long-term stability studies and commercial manufacturing.