Predicting the Water Sorption in ASDs

Counteracting the Loss of Release for Indomethacin-Copovidone ASDs

This work revisits the changing release behavior of indomethacin(IND)-copovidone amorphous solid dispersions (ASDs) when increasing their drug load (DL). While showing congruent release behavior at DL 0.1, ASDs with DLs of 0.3 and higher show incongruent release finally resulting in a complete loss of release. To study and explain this phenomenon, we modeled the release kinetics of these ASDs and looked into their phase behavior both experimentally and theoretically. We applied a diffusion model to accurately describe experimental release profiles for congruent release, incongruent release as well as for loss of release. Predicted concentration profiles for IND, copovidone, and water within the ASD revealed the formation of an ASD layer that almost exclusively contains amorphous IND. Our phase-diagram predictions and experimental data explain this phenomenon by water-induced phase separation in those parts of the ASD which did absorb water from the dissolution medium. Whereas the evolving copovidone-rich phase dissolved, the IND-rich phase remained undissolved and formed a super-hydrophobic cover of the remaining inner core of the ASD, thus finally completely preventing its dissolution. Higher DLs promote phase separation. This leads to the counterintuitive effect that the higher the DL, the lower the absolute amount of IND released. While the ASD containing 6 mg IND (DL 0.1) released 6 mg IND, the one containing 42 mg IND (DL 0.7) released only 1 mg IND. The theoretical approach applied in this work is for the first time able to quantitatively predict that reducing DL or tablet size could be used to overcome this problem.

Simultaneous Water Sorption and Crystallization in ASDs 2: Modeling Long-Term Stabilities

The physical stability of amorphous solid dispersions (ASDs) is a major topic in the formulation research of oral dosage forms. To minimize the effort of investigating the long-term stability using cost- and time-consuming experiments, we developed a thermodynamic and kinetic modeling framework to predict and understand the crystallization kinetics of ASDs during long-term storage below the glass transition. Since crystallization of the active phrarmaceutical ingredients (APIs) in ASDs largely depends on the amount of water absorbed by the ASDs, water-sorption kinetics and API-crystallization kinetics were considered simultaneously. The developed modeling approach allows prediction of the time evolution of viscosity, supersaturation, and crystallinity as a function of drug load, relative humidity, and temperature. It was applied and evaluated against two-year-lasting crystallization experiments of ASDs containing nifedipine and copovidone or HPMCAS measured in part I of this work. We could show that the proposed modeling approach is able to describe the interplay between water sorption and API crystallization and to predict long-term stabilities of ASDs just based on short-term measurements. Most importantly, it enables explaining and understanding the reasons for different and sometimes even unexpected crystallization behaviors of ASDs.

Kinetics of Water-Induced Amorphous Phase Separation in Amorphous Solid Dispersions via Raman Mapping

The poor bioavailability of an active pharmaceutical ingredient (API) can be enhanced by dissolving it in a polymeric matrix. This formulation strategy is commonly known as amorphous solid dispersion (ASD). API crystallization and/or amorphous phase separation can be detrimental to the bioavailability. Our previous work (Pharmaceutics 2022, 14(9), 1904) provided analysis of the thermodynamics underpinning the collapse of ritonavir (RIT) release from RIT/poly(vinylpyrrolidone-co-vinyl acetate) (PVPVA) ASDs due to water-induced amorphous phase separation. This work aimed for the first time to quantify the kinetics of water-induced amorphous phase separation in ASDs and the compositions of the two evolving amorphous phases. Investigations were performed via confocal Raman spectroscopy, and spectra were evaluated using so-called Indirect Hard Modeling. The kinetics of amorphous phase separation were quantified for 20 wt% and 25 wt% drug load (DL) RIT/PVPVA ASDs at 25 °C and 94% relative humidity (RH). The in situ measured compositions of the evolving phases showed excellent agreement with the ternary phase diagram of the RIT/PVPVA/water system predicted by PC-SAFT in our previous study (Pharmaceutics 2022, 14(9), 1904).

Water Sorption in Rubbery and Glassy Polymers, Nifedipine, and Their ASDs

Polymers like poly(vinylpyrrolidone-co-vinyl acetate) (PVPVA) or hydroxypropyl methylcellulose acetate succinate (HPMCAS) are commonly used as a matrix for amorphous solid dispersions (ASDs) to enhance the bioavailability of the active pharmaceutical ingredients (APIs). The stability of ASDs is strongly influenced by the water sorption in the ASD from the surrounding air. In this work, the water sorption in the neat polymers PVPVA and HPMCAS, in the neat API nifedipine (NIF), and in their ASDs of different drug loads was measured above and below the glass-transition temperature. The equilibrium water sorption was predicted using the Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT) combined with the Non-Equilibrium Thermodynamics of Glassy Polymers (NET-GP).The water-sorption kinetics were modeled using the Maxwell-Stefan approach whereas the thermodynamic driving force was calculated using PC-SAFT and NET-GP. The water diffusion coefficients in the polymers, NIF, or ASDs were determined using the Free-Volume Theory. Using the water-sorption kinetics of the pure polymers and of NIF, the water-sorption kinetics of the ASDs were successfully predicted, thus providing the water diffusion coefficients in the ASD as a function of relative humidity and of the water concentration in polymers or ASDs.

Anomalous Water-Sorption Kinetics in ASDs

Anomalous water-sorption kinetics in amorphous solid dispersions (ASDs) are caused by the slow swelling of the polymer. In this work, we used a diffusion–relaxation model with the Williams–Landel–Ferry (WLF) equation and the Arrhenius equation to predict the anomalous water-sorption kinetics in ASDs of poly(vinyl-pyrrolidone)-co-vinyl-acetate (PVPVA) and indomethacin (IND) at 25 °C. These predictions were based on the viscosities of pure PVPVA and pure IND, as well as on the water-sorption kinetics in pure PVPVA. The diffusion–relaxation model was able to predict the different types of anomalous behavior leading to a qualitative and quantitative agreement with the experimental data. Predictions and experiments indicated more pronounced anomalous two-stage water-sorption behavior in the ASDs than in pure PVPVA. This was caused by a higher viscosity of glassy ASD–water mixtures compared to glassy PVPVA–water mixtures at the same distance from their glass transition temperature. These results suggest that this ASD swells more slowly than the polymer it is composed of. The modeling approach applied in this work can be used in the future for predicting diffusion-controlled release behavior or swelling-controlled release behavior of ASDs.