Exploring the influence of hydrogen bond donor groups on the microstructure and intermolecular interactions of amorphous solid dispersions containing diflunisal structural analogues.

Drug-polymer intermolecular interactions, and H-bonds specifically, play an important role in the stabilization process of a compound in an amorphous solid dispersion (ASD). However, it is still difficult to predict whether or not interactions will form and what the strength of those interactions would be, based on the structure of drug and polymer. Therefore, in this study, structural analogues of diflunisal (DIF) were synthesized and incorporated in ASDs with poly(vinylpyrrolidone-co-vinyl acetate) (PVPVA) as a stabilizing polymer. The respective DIF derivatives contained different types and numbers of H-bond donor groups, which allowed to assess the influence of these structural differences on the phase behavior and the actual interactions formed in the ASDs. The highest possible drug loading of these derivatives in PVPVA were evaluated through film casting. Subsequently, a lower drug loading of each compound was spray dried. These spray dried ASDs were subjected to an in-depth solid-state nuclear magnetic resonance (ssNMR) study, including 1D spectroscopy and relaxometry, as well as 2D dipolar HETCOR experiments. The drug loading study revealed the highest possible loading of 50 wt% for the native DIF in PVPVA. The methoxy DIF derivative reached the second highest drug loading of 35 wt%, while methylation of the carboxyl group of DIF led to a sharp decrease in the maximum loading, to around 10 wt% only. Unexpectedly, the maximum loading increased again when both the COOH and OH groups of diflunisal were methylated in the dimethyl DIF derivative, to around 30 wt%. The ssNMR study on the spray dried ASD samples confirmed intermolecular H-bonding with PVPVA for native DIF and methoxy DIF. Studies of the proton relaxation decay times and 2D 1H-13C dipolar HETCOR experiments indicated that the ASDs with native DIF and methoxy DIF were homogenously mixed, while the ASDs containing DIF methyl ester and dimethyl DIF were phase separated at the nm level. It was established that, for these systems, the availability of the carboxyl group was imperative in the formation of intermolecular H-bonds with PVPVA and in the generation of homogenously mixed ASDs.

Investigating the Potential of Ethyl Cellulose and a Porosity-Increasing Agent as a Carrier System for the Formulation of Amorphous Solid Dispersions.

In the present work, an insoluble polymer, i.e., ethyl cellulose (EC), was combined with the water-soluble polyvinylpyrrolidone (PVP) as a carrier system for the formulation of amorphous solid dispersions. The rationale was that by conjoining these two different types of carriers a more gradual drug release could be created with less risk for precipitation. Our initial hypothesis was that upon contact with the dissolution medium, PVP would be released, creating a porous EC matrix through which the model drug indomethacin could diffuse. On the basis of observations of EC as a coating material, the effect of the molecular weight of PVP, and the ratio of EC/PVP on the miscibility of the polymer blend, the solid state of the solid dispersion and the drug release from these solid dispersions were investigated. X-ray powder diffraction, modulated differential scanning calorimetry, and solid-state nuclear magnetic resonance were used to unravel the miscibility and solid-state properties of these blends and solid dispersions. Solid-state nuclear magnetic resonance appeared to be a crucial technique for this aspect as modulated differential scanning calorimetry was not sufficient to grasp the complex phase behavior of these systems. Both EC/PVP K12 and EC/PVP K25 blends were miscible over the entire composition range, and addition of indomethacin did not alter this. Concerning the drug release, it was initially thought that more PVP would lead to faster drug release with a higher probability that all of the drug molecules would be able to diffuse out of the EC network as more pores would be created. However, this view on the release mechanism appeared to be too simplistic as an optimum was observed for both blends. On the basis of this work, it could be concluded that drug release from this complex ternary system was affected not only by the ratio of EC/PVP and the molecular weight of PVP but also by interactions between the three components, the wettability of the formulations, and the viscosity layer that was created around the particles.