A study of microcrystalline cellulose variability and how this impacts compaction performance

Abstract

Quality by Design (QbD) is driving the pharmaceutical industry to better understand the impact of raw material variability on the performance and manufacturability of drug products. Identifying excipients critical material attributes (CMAs) is not an easy task as criticality depends on the process and formulation in question. The aim of this thesis was to study the impact of microcrystalline cellulose material variability due to 1) storage relative humidity and 2) inherent variability in the material properties. Initially the impact of varied relative humidity storage conditions on MCC compaction performance was studied (Chapter 3). It was found that storing MCC at high relative humidity (75% RH, 40°C) had a negative effect on the tablet tensile strength of compacts produced from neat MCC by direct compression. In the case of roller compaction, MCC stored at high relative humidity (75% RH, 40°C) prior to incorporation into a dry granulation placebo blend had a negative impact on granule tabletability when tableted at higher compression forces (>180 MPa) on a rotary tablet press. At low and ambient relative humidity the performance of MCC during direct compression and dry granulation was found to be robust. Following this study the focus of this thesis changed to studying the inherent variability in the MCC material. Three different approaches where applied to study MCC variability. These were multi-variate models (PCA and PLS), a compaction model (DPC) and a Raman PCA model. Initially both uni-variate and multi-variate analysis were applied to investigate the inter-lot variability of four Avicel™ MCC lots manufactured from different wood pulp or wood pulp blend starting materials (Chapter 4). PCA analysis identified significant variability across all four samples for the MCC material attributes studied. Differences in crystallinity index, specific surface area and D50 were found to have a significant effect on MCC tabletability over the compression range studied. This work was expanded on in Chapter 5 to include supplier variability. Seven MCC lots manufactured from different suppliers, FMC, JRS and Asahi Kasei were studied using multi-variate analysis. PCA identified that MCC material attribute supplier variability was larger than variability within a single supplier. The Drucker Prager Cap (DPC) compaction model was used to investigate lot to lot variability of the mechanical properties of MCC (Chapter 6). This is the first time this model has been applied to study the lot to lot variability of a material. This work confirmed the dominance of plastic deformation of MCC. In addition the DPC model predicted compact strength results (cohesion) were shown to correlate well with lot performance during tableting on a rotary tablet press (tensile strength). This thesis work also investigated the potential for Raman spectroscopy as an efficient PAT screening technique to identify MCC lots of differing crystallinity index using PCA modelling (Chapter 7). FT Raman spectroscopy PCA modelling was shown to be capable of identifying different MCC lots from each other. It was shown that PC1 of this PCA Raman model separated samples on the basis of lot crystallinity. Possible avenues of research to be considered going forward include further work on validating the Raman crystallinity PCA model. Another area of interest would be to conduct compression studies using MCC lots which have been shown to compress to different tensile strength values with a model API. This approach could assess if the variability in neat MCC during direct compression would have a significant impact on a formulations performance when an API is included in the blend.