(333g) A Hybrid Multivariate Model for Understanding Fluid Bed Wet Granulation Process Parameters Impact on Tablet Physicals

Throughout the drug product development phase, models are developed and utilized to demonstrate process understanding. Systemic models provide the benefit of combining multiple unit operations, which allows the researcher to explore multivariate interactions across the production train. However, many complex models are time consuming to construct and require copious amount of data to generate high prediction accuracy. This work describes the development of a simple and easy to implement multivariate model of a drug product manufacturing process which utilizes fluid bed wet granulation followed by continuous direct compression.

Fluid bed wet granulation is a complex process as it involves several phenomena such as fluidization, granulation, and drying occurring simultaneously. To bypass mechanistic description of complex dynamic relationships between process parameters and responses, a partial-least-squares model was developed from the experimental data that was used to predict granule particle size. Subsequently, the granule particle size was used as an intermediate variable to relate fluid bed wet granulation process parameters to the resulting tablet physicals (e.g., tablet tensile strength). Combined, these relationships generate an integrated multivariate model which uses fluid bed wet granulation process parameters and compression stress as inputs to predict tablet tensile strength.