Poor water solubility is a bottle neck in the development of many new drug candidates, and understanding and circumventing this is essential for a more effective drug development. Zafirlukast (ZA) is a leukotriene antagonist marketed for the treatment of asthma (Accolate®). ZA is poorly water soluble, and is formulated in an amorphous form (aZA) to improve its solubility and oral bioavailability. It has been shown that upon dissolution of aZa, the concentration of ZA in solution is supersaturated with respect to its stable crystalline form (ZA monohydrate), and thus, in theory, the bioavailability increases upon amorphization of ZA. The polymers hydroxypropylmethylcellulose (HPMC) and polyvinylpyrrolidone (PVP), often used as stabilizers of the supersaturated state, are in the excipient list of Accolate®. It is not recommended to take Accolate® with food, as this reduces the bioavailability by 40%.
The aim of this study was to investigate the effect of simulated fasted and fed state intestinal media as well as the effect of HPMC and PVP on the supersaturation and precipitation of ZA in vitro.
Supersaturation of aZA was studied in vitro in a small scale setup using the μDiss Profiler™. Several media were used for this study: One medium simulating the fasted state intestinal fluids and three media simulating different fed state intestinal fluids. Solid state changes of the drug were investigated by small angle x-ray scattering. The duration wherein aZA was maintained at a supersaturated state was prolonged in the presence of HPMC and lasted more than 20 h in the presence of PVP in a fasted state intestinal medium. The presence of PVP increased the concentration of drug dissolved in the supersaturated state. The duration of supersaturation was shorter in fed than in a fasted state simulated intestinal media, but the concentration during supersaturation was higher. It was thus not possible to predict any positive or negative food effects from the dissolution/precipitation curves from different media. Lipolysis products in the fed state simulated media seemed to cause both a negative effect on the duration of supersaturation, and an increased drug concentration during supersaturation. In contrast, when testing the effect of a fed state simulated medium compared to the fasted state medium, in the presence of PVP, a clear negative effect was seen on the dissolution/precipitation curved of the fed state medium. The drug concentration during supersaturation was marginally different in the two media, but a precipitation of ZA was seen in the fed state medium, which was not observed in the fasted state medium. Solid state transformation from aZA to ZA monohydrate (mhZA) upon precipitation of the supersaturated solutions was confirmed by small angle x-ray scattering. All of these results can explain the described in vivo behavior of ZA. For ZA simple dissolution experiments in vitro can be used to examine supersaturation, effectiveness of PI and potential food effects on these.