Quantitative analysis of the effects of multi-component formulation parameters on granule and tablet properties via a combined population balance and statistical predictive model

In pharmaceutical drug product manufacturing, the use of multi-component mixtures in the formulation is a common practice. However, the introduction of each material alters the powder blend properties and affects the granule growth mechanisms in a wet granulation process. In this study, we quantified the effects of varying formulation parameters on granule growth rate, expressed by the liquid amount required for granulation to reach completion, and on granule and tablet quality attributes. Process models, including population balance and statistical models, were developed to represent the granulation, milling, and tableting process, and predict relevant granule and tablet properties. Overall, prediction of granule properties had an accuracy of above 80% and the prediction of different tablet properties ranged from an R2 of 0.68 to 0.9. These models are computationally efficient and can be simulated serially to predict the performance of different multi-component formulations by using the pre-blend properties as model inputs.