Purpose: To investigate the causes for drug crystallization in amorphous solid dispersions (ASD). The impact of drug loading, storage temperature, humidity, and exposure to other solvents have been explored for their relative importance as the cause for crystallization in amorphous solid dispersions.
Methods: Indomethacin (IND) and polyvinylpyrrolidone K12 (PVP K12) ASDs were formulated at varying drug loadings (65 – 90% w/w) via cryomilling and melt-quenching. The glass transition temperature (Tg) for each dispersion was measured using differential scanning calorimetry (DSC), and storage temperatures were then assigned (50 – 90°C) to probe the design space both above and below the glass transition. Samples were stored over desiccant (approximately 0% RH) at the temperatures noted previously and monitored for the onset of crystallization using powder X-ray diffractometry (PXRD). Similar stability studies have been performed at varying relative humidities. Additional samples were equilibrated with mixtures of water and organic solvent in a dynamic vapor sorption (DVS) apparatus. The Tg was monitored as a function of solvent identity and content while a method was also developed for quantitation of the two sorbed components.
Results: All dispersions were initially confirmed as amorphous using modulated DSC and initial PXRD scans. Higher drug loading with IND resulted in increasingly lower Tg values and a faster onset of crystallization at all storage temperatures. An exponential dependence of crystallization onset time was noted as a function of drug loading for all temperatures. As the glass transition temperature was approached, an increase in the rate of subsequent crystallization was observed, possibly indicating a switch to diffusionless crystal growth.
Equilibration with various solvents in the DVS unit had a plasticizing effect on the glass transition temperature. A difference in the rate of uptake was also observed between water and organic solvents which may result from differences in affinity and/ or molecular size effects.
Conclusion: Maintaining the stability of the amorphous drug component is critically important for successful pharmaceutical formulations. Increased drug loading and temperature provide a larger thermodynamic driving force for crystallization from which the onset time may be able to be predicted as a function of drug loading. Conversely, small amounts of solvents present have a significant impact on the Tg and crystallization propensity. A better understanding of the causes of destabilization and their relative significance towards causing crystallization will help to make better informed decisions during spray drying or other formulation steps.