Category: Formulation and Quality
Purpose: Amorphous solid dispersions (ASDs), in which an amorphous drug is formulated with an amorphous polymer, generate a transient supersaturated drug solution upon dissolution, where the concentrations achieved are significantly higher than the equilibrium solubility. Non-sink conditions are essential for assessing the in vitro dissolution performance of a supersaturating system, as this allows the achievable supersaturation to be characterized, as well the kinetics of nucleation and growth. If crystals are present within an ASD, these crystals directly result in lost solubility advantage, and may seed crystal growth thereby reducing the amount of dissolved drug available for absorption. Herein, the impact of the polymer on the stabilization of the attained supersaturation, as well as evaluation of its inhibitory effect on crystal seed growth was investigated for amorphous solid dispersion formulations with and without residual crystals.
Methods: Indomethacin and PVPVA are used as the model drug and polymer. Hot melt extrusion (HME) was used to produce ASDs containing 0-25% residual crystals, by processing at temperature and residence times conditions insufficient to produce fully amorphous systems. The impact of crystal seeds on non-sink dissolution performance was assessed through two study designs. First, the impact of bulk crystal seeds on maintenance of supersaturation and the effect of polymeric additives on crystal growth inhibition was explored. Second, the non-sink dissolution performance of indomethacin-PVPVA HME ASDs containing 0-25% residual crystallinity was investigated. Scanning electron microscopy (SEM) images demonstrating the microstructure of partially dissolved and grown crystals seeds were also collected.
Results: The ability of the polymer to maintain supersaturation in the presence of crystal seeds is shown in Figure 1. In the absence of polymer at high supersaturation, indomethacin (50 µg/mL) nucleates and desupersaturates rapidly. In the presence of bulk crystal seeds, the rate of desupersaturation is proportional to amount of seeds added. When PVPVA is added to the solution (50 µg/mL), supersaturation is maintained for 12 hours. Slow (or no) desupersaturation results in the presence of crystal seeds. Figure 2 shows the non-sink dissolution profile of 1:1 indomethacin:PVPVA ASDs containing various levels of residual crystals (0-25%). The loss in solubility advantage roughly corresponds to the amount of residual crystalline content (2A), and minimal seed growth is observed (slight desupersaturation is noted). Several batches, containing < 2% crystals, have similar dissolution profiles (2B), and show no evidence of seeded crystal growth. Similar results are found for physical mixtures of ASD and bulk crystal seeds (2C). Essentially, all available amorphous drug dissolves, while the crystal seed growth is poisoned by the presence of the polymer.
Differences in the microstructure of dissolving crystals and crystals seeds grown from supersaturated solutions are seen in the absence and presence of polymer (Figure 3). Notably, the surface is much rougher when dissolving in the absence of polymer. In the presence of polymer, the surfaces are smoother, indicating that polymer is interacting with the crystal surfaces. Under the same conditions at high supersaturation (an additional 50 µg/mL of indomethacin), indomethacin crystals grow with smooth surfaces in the absence of polymer and with a high degree of irregularity in the presence of polymer. Similar patterns are observed for crystals residual from an HME ASD. After dissolution of the amorphous material, a small residual crystal remains with rounded features, but a highly non-uniform structure. Such a small, defective, and irregular crystal would be expected following the high temperature and shear conditions of an HME process. Residual crystals from an HME ASD, after dissolution of the amorphous material but at higher supersaturation conditions (an additional 50 µg/mL of indomethacin), show needle-like growth protruding from all crystal surfaces, suggesting underlying disorder in the processed crystal.
Conclusion: By studying the dissolution performance of ASDs with residual crystals under non-sink conditions, the loss of solubility advantage can be observed. The impact of the polymer to maintain supersaturation and poison crystal growth was also illustrated. The defective nature and higher surface energy of residual crystals in the ASD can be compared to bulk crystals through imaging techniques.
Lynne Taylor– Professor, Department of Industrial and Physical Pharmacy,College of Pharmacy, Purdue University, West Lafayette, Indiana