Purpose: To identify the mechanism behind the unexpected bio-performance of two amorphous solid dispersions: BMS-A/PVP-VA and BMS-A/HPMC-AS.
Methods: Solubility of crystalline BMS-A in PVP-VA and HPMC-AS was measured by DSC. Drug-polymer interaction parameters were obtained by Flory-Huggins model fitting. Drug dissolution kinetics of spray-dried dispersions were studied under sink and non-sink conditions. BMS-A supersaturation was studied in the presence of pre-dissolved PVP-VA and HPMC-AS. Potency and crystallinity of undissolved solid dispersions were determined by HPLC and DSC. Polymer dissolution kinetics were obtained by mass balance calculation. Bioavailability of solid dispersions was assessed in dogs.
Results: In solid state, both polymers are miscible with BMS-A, while PVP-VA solublizes the drug better. BMS-A dissolves similarly from both solid dispersions in vitro regardless of dissolution method, while the HPMC-AS dispersion performed much better in vivo. At the same concentration, HPMC-AS is more effective in prolonging BMS-A supersaturation; this effect was negated by the slow dissolution rate of HPMC-AS. Further study revealed that fast PVP-VA dissolution resulted in elevated drug loading in undissolved dispersions and facilitated drug recrystallization before complete release. In contrast, the hydrophobicity and slower HPMC-AS dissolution prevented BMS-A recrystallization within the HPMC-AS matrix for >24 h.
Conclusions: The lower bioavailability of PVP-VA dispersion was attributed to BMS-A recrystallization within the undissolved dispersion, due to hydrophilicity and fast PVP-VA dissolution rate. Polymer selection for solid dispersion development has significant impact on in vivo performance besides physical stability.