MULTIVARIATE DATA ANALYSIS FOR TABLETING PERFORMANCE IMPROVEMENT AT INDUSTRIAL SCALE. A CASE STUDY FOCUSED ON UNDERSTANDING AND AVOIDING THE OCCURRENCE OF CAPPING AND LAMINATION

Abstract

The objective of this work was to investigate the granule properties and processing conditions leading to capping and lamination phenomena occurred for a high drug load, wet granulated, film coated product. Assessment of historical industrial data was combined with parameters resulting from compaction simulator testing of representative samples. A systematic approach, based on batch statistical modelling, was applied to identify patterns between batches linked to changes in input variables. Batch evolution modelling revealed the typical process fingerprint and expected variability and confirmed that defects were not linked to process evolution excursions. Class-based modelling and ANOVA allowed the identification of statistically significant differences between input variables (active ingredient particle size and granulation water amount) and adjustments to be done to improve the process. The model built with the parameters resulting from the compaction simulator confirmed the inter-batch differences and suggested the mechanism for capping/lamination-higher elastic recovery rate. The strain rate sensitivity (SRS) values showed a product sensitivity to certain processing conditions, including dies wear. SRS and bulk density showed a negative correlation and could be used to predict the appropriate tableting conditions to increase tablet hardness.