A Comparison of Spray drying and Liquid Antisolvent Co-precipitation for the Generation of Two- and Three- component Amorphous Solid Dispersions

Abstract

The thesis focuses on the use of different manufacturing techniques for the generation of amorphous solid dispersions (ASDs) for the enhancement of the solubility of model compounds. ASDs with hydrochlorothiazide (HCTZ) or simvastatin (SIM) were successfully manufactured via spray drying (SD) and liquid antisolvent (LAS) co-precipitation (CP). The LAS CP resulted in the generation of ASDs characterized by component ratios similar to the theoretical values. The novel approach undertaken in this thesis related to the co-processing of materials in the lab-scale chemical reactor with the use of organic solvents only to generate amorphous solid dispersions (ASDs). Moreover, not only binary API-polymer systems were investigated, but also ternary API- polymer-silica formulations, which were generated either from solutions or suspensions in the CP process. The work carried out in this thesis incorporated all aspects of the pharmaceutical drug development process for the CP technique on the laboratory scale. The impact of differences in both API and excipients used were explored, with a diverse range of behaviours displayed by the various co-precipitated API/polymer or API/polymer/silica combinations investigated, which were compared to the equivalent spray-dried formulations. In particular, the effect of CP process parameters on changes to the particle size distribution was investigated. In this thesis, the CP process and feed parameters required for powder generation and the impact of the manufacturing technique on the powder products, as well as the role excipients play on the final product characteristics were studied. All raw materials employed in this study were characterized, followed by the selection of CP parameters. The preformulation studies demonstrated different melting temperatures (Tms), glass transition temperatures (Tgs) and solubilities in organic and aqueous solvents for the two model active pharmaceutical ingredients (APIs). The hydrophilic, polymeric excipients (Kollidon® VA 64 (PVP/VA) and Soluplus®) utilized were characterized by different Tgs (107 °C and 70 °C, respectively) and have been successfully used in the production of ASDs by spray-drying. One of the challenges associated with LAS CP relates to the potential different solubilities of formulation components in the final mixture of solvents, in which the solids should have good solubility and poor solubility (antisolvent), which might result in a disproportionate loss of one of the materials from the co-precipitate product, limiting the robustness of the process. The feasibility of the LAS CP process to be conducted with non-aqueous antisolvents and compounds with a high Tg (HCTZ:117°C) or low Tg (SIM:35°C) was investigated. The successful co-precipitation of ASDs at the screening stage (40 mL) enabled successful co- precipitation at a larger scale (100 mL) with the same parameters for HCTZ-based systems and with further optimized parameters for SIM-based systems. The SEM analysis of particles obtained from precipitated slurries compared to filtered, vacuum-dried powders demonstrated changes in the particle morphologies when additional unit operations were employed. Particle counts and chord length distributions (CLD) were monitored in situ using a Focus Beam Reflectance Measurement (FBRM) probe, and showed differences, depending on the API used, due to different investigated process parameters, such as the type of a stirring device, mixing rate, type of antisolvent, solvent to antisolvent ratio, antisolvent temperature, solvent addition rate, order of solvents addition and other modifications of the process. Binary (API with polymer) HCTZ-based and SIM-based ASDs were manufactured by SD and CP. The influence of the manufacturing process as well as polymer type were studied for their impact on product powder characteristics. Spray-dried and co-precipitated powders showed largely similar Tgs for SIM-based systems, but different Tgs for HCTZ-based materials. Moreover, for both methods, experimentally obtained Tgs were either lower or higher than theoretically estimated Tg values by the Gordon-Taylor equation for SIM and HCTZ-based systems, respectively. However, Tg differences did not have an impact on the physical stability of ASDs kept on storage at 75% RH/25°C for one month, which all remained amorphous, regardless of the manufacturing method used. This work highlights that, despite the suitability of both techniques for producing ASDs and many similarities between the final critical quality attributes (CQAs) of the powders obtained, the SD technique has many advantages over CP. Firstly, it can be viewed as a more environmentally friendly approach due to the lower amounts of solvents which are required to be used per 1 g of powder produced. Secondly, while both SD and CP techniques have the potential to be used in continuous processing, CP requires more post-processing unit operations, such as filtration and drying. However, powders produced by CP have a larger particle size, resulting in a better flowability than powders produced by SD, although tablets prepared with CP-processed powders had a lower mechanical strength. The work of this thesis also highlights the similarity in solubility enhancement of ASDs prepared by CP and SD, relative to their respective crystalline materials. The addition of silica as a third component and a co-precipitation processing aid to ASD formulations addressed the challenges associated with the high stickiness of co-precipitates to the vessel bottom and walls, which was observed on filtration. This behaviour impedes solvent removal because it makes sample handling and spreading on the filter, which is meant to facilitate solvent removal, difficult. By presenting two alternative methods of silica inclusion into the CP process - either (A) suspended in the antisolvent (to which a feed solution of polymer and API is added) or (B) added as a feed suspension (with API and polymer dissolved in the same liquid feed), it was shown that the generation of ternary ASDs with similarly enhanced API solubilities, relative to the crystalline material, can be achieved. These studies also outlined several differences between the CP and SD methods in co-processing all components of the formulation in one step. Similarly to binary API/polymer powders, ternary dried co-precipitated systems were characterized by larger particle size and more irregular particle shapes than ternary spray-dried materials. While only ternary SIM/Aerosil®200 CP systems exhibited poorer flowability than the binary CP samples, ternary SD samples with Syloid®XDP 3050 and Aerosil®200 demonstrated significantly poorer flowability than SD binary materials. It was shown that, in the case of both a high Tg API (HCTZ), which demonstrated (by FT-IR) intermolecular interactions in ternary systems, and a low Tg API (SIM,) which did not demonstrate intermolecular interactions in ternary systems, there were significant differences in the Tgs of ternary systems between SD and CP-processed materials, especially in the case of SIM-based systems, as opposed to binary SIM/polymer powders, dependent on the silica grade used. The work conducted and presented in this thesis highlights an already well-established process in the pharmaceutical industry, namely spray-drying, and a just-emerging technology, namely liquid antisolvent co-precipitation, that can both be applied to ASD production. Both these methods have the potential to be used in a continuous mode. However, the larger particle size favours CP as a candidate for downstream tabletting processes involving direct compression, due to enhanced flowability. While processing APIs with a high or low Tg is feasible by these two techniques, the large solvent waste generated by the CP method should be taken into consideration when selecting a process for large scale manufacture.