Non-Destructive Pipeline for Analysis of Crystalline Solid Dispersions

Abstract

The continued development of advanced processing technologies for pharmaceutical drug products via crystalline solid dispersion (CrySoD) requires non-invasive analytical methods capable of characterizing resultant products to understand the impact of processing choices on critical quality attributes (CQAs, e.g., crystal size, morphology, and drug content). In this study, we demonstrate such a non-destructive and robust pipeline, using X-ray micro-computed tomography (X-µCT) and a Python-based image processing strategy for modafinil CrySoDs within hydroxypropyl methylcellulose (HPMC) capsules produced via a solution-based additive manufacturing process. X-µCT scans were first processed using radial filtration and smoothing splines to isolate the capsule object, and then the active pharmaceutical ingredient (modafinil) was isolated from the scans using greyscale thresholding based on X-µCT intensities. The CQAs were then evaluated using a combination of heterogeneity index calculations and image segmentation approaches. Watershed segmentation was most successful in crystal size analysis, as it can segment and separately characterize fused crystallites, which are well-dispersed in the capsules. Overall, this work establishes a generalizable, non-destructive workflow that might be used for the structural characterization of CrySoDs and to provide a quantitative framework that links critical process parameters (CPPs) to final products CQAs.

Keywords: CrySoD, Polymorphism, Crystallization, capsule, API, pharmaceutical, modafinil