(T0930-02-08) API – Polymer Interactions with Dissolution Media: A Key Consideration in Method Development for Amorphous Solid Dispersions

Purpose: Quality control (QC) dissolution methods are designed to capture the critical quality attributes (CQAs) of a dosage form and detect potential changes in product performance. In formulations containing amorphous solid dispersions (ASDs), the physical form of the drug is a typical CQA, as crystallization can significantly affect bioavailability. Consequently, detection and control of crystallinity are often required as part of dissolution method development [1]. Since ASDs introduce additional factors not accounted for by traditional solubility measurements of the crystalline drug, such as media–drug–polymer interactions, the selection of dissolution media becomes a crucial aspect of QC method development. This study investigates how the choice of dissolution media affects the solubility behavior of celecoxib (CLX), a neutral, poorly water-soluble Active Pharmaceutical Ingredient (API), in both crystalline and amorphous solid dispersion (ASD) forms. The goal is to support the development of robust, discriminatory QC dissolution methods that effectively monitor product performance and meet regulatory requirements.

Methods: ASD of CLX and PVP-VA (50:50) was prepared by spray-drying (SD) and confirmed amorphous by X-Ray Powder Diffraction (XRPD) and Differential Scattering Calorimetry (DSC). Solubility of crystalline CLX and ASD was assessed in pH 1.2 [Hydrochloric acid (HCl)], 4,5 [Acetate Buffer (AB)] and 6.8 [Sodium Phosphate Buffer (SPB)], with and without surfactants [Sodium dodecyl Sulphate (SDS), Cetyltrimethylammonium bromide (CTAB), Polysorbate 20 (Tween 20)] at 1x and 5x their Critical Micelle Concentrations (CMC), using the shake flask methodology at 37 ºC. Samples were collected at 1, 4, and 24 hours, centrifuged at 13 400 rpm for 15 minutes and analyzed using High-performance liquid chromatography (HPLC). Physical form during solubility testing was evaluated by XRPD and Raman Spectroscopy. Nano-species were assessed by Dynamic Light Scattering (DLS). QC dissolution testing was conducted to crystalline CLX and ASD in media where ASD solubility exceeded the target dose concentration (444 µg/mL).

Results: CLX formulated as ASD exhibited higher solubility across various pH media when compared with their crystalline counterpart (ASD showed a maximum solubility of 18 µg/mL vs 7 µg/mL for crystalline CLX). Nevertheless, API solubility in these media remained significantly low (Sink Index (SI) < 1), considering both equilibrium and kinetic solubility. As such, the use of surfactants was considered to enhance drug solubility—a common strategy in dissolution method development for poorly water-soluble drugs. In media containing surfactants at 1x and 5x at their respective CMCs, the solubility enhancement followed the order: SLS > CTAB > Tween 20 for crystalline CLX, and SLS ≈ CTAB > Tween 20 for the ASD, independent of buffer pH. In detail, SDS was found to enhance the solubility of both crystalline and amorphous CLX to a similar extent and independent of surfactant concentrations. At 5x CMC, both forms achieved solubility above the target dose concentration. XRPD of the undissolved residue confirmed recrystallization of CLX from the ASD under these conditions, which explains the similar solubility of the two physical form. DLS detected nanoparticles of ~50 nm in the ASD suspended in SDS media, consistent with SLS–PVP-VA nanoaggregates. The strong interaction between PVP-VA and SDS may reduce the polymer’s ability to maintain the drug in a supersaturated state, thereby limiting its solubility-enhancing effect of the polymer in this medium [2]. In contrast, in CTAB-containing media, the ASD showed nearly twice the solubility of the crystalline form. DLS measurements revealed CTAB micelles size increase, suggesting drug incorporation into CTAB micelles. Furthermore, no crystallization of the ASD was detected by XRPD in this medium. Figure 1 presents the dissolution profiles of ASD and crystalline API powders in two different media, tested at target dose concentrations. In SDS-containing media (Figure 1a), both the crystalline and amorphous forms achieved complete drug release. In contrast, this was not observed in CTAB media (Figure 1b), where the solubility limit of the crystalline API restricted its dissolution performance. Figures 1c and 1d display the particle size distribution of dissolved species at the 120 minute time point, measured after dissolution cannula filtration. These results support the findings from the solubility studies, namely the formation of nanoscale species (~50 nm) in SDS media which is consistent with SDS–CLX complexation. Contrasting, in CTAB media only an increase in micelle size was observed. Overall, the behavior of CLX ASDs in different dissolution media highlights the critical role of drug–polymer–surfactant interactions.

Conclusion: The study demonstrated that dissolution media significantly influence the solubility of crystalline and amorphous CLX, with polymer–surfactant interactions playing a key role. While SDS increased solubility, it failed to maintain supersaturation. Tween 20 led to limited improvement and rapid recrystallization. In contrast, CTAB enabled sustained supersaturation without recrystallization, likely due to favorable interactions with PVP-VA. These results highlight the importance of selecting dissolution media that prevent ASD recrystallization and effectively discriminate between solid-state forms.

References: [1] ICH Q6A specifications: test procedures and acceptance criteria for new drug substances and new drug products: chemical substances - Scientific guideline, 2000.
[2] C. Liu et al., “Improving Oral Bioavailability of Sorafenib by Optimizing the ‘Spring’ and ‘Parachute’ Based on Molecular Interaction Mechanisms,” Mol Pharm, vol. 13, no. 2, pp. 599–608, Feb. 2016.

Figure 1 - Dissolution behavior of amorphous solid dispersions under different media 5xCMC SDS + 50 mM Phosphate Buffer (a) and 5x CMC CTAB + 50 mM Phosphate Buffer (b). Particle size distribution of solutions taken from the last timepoint of SDS media (c) and CTAB media (b).