RESUMO
Solid pharmaceutical formulations with class II active pharmaceutical ingredients (APIs) face dissolution challenges due to limited solubility, affecting in vivo behavior. Robust computational tools, via data mining, offer valuable insights into product performance, complementing traditional methods and aiding in scale-up decisions. This study utilizes the design of experiments (DoE) to understand fluidized hot-melt granulation manufacturing technology. Exploratory data analysis (MVDA) highlights similarities and differences in tablet manufacturability and dissolution profiles at both the lab and pilot scales. The study sought to gain insights into the application of multivariate data analysis by identifying variations among batches produced at different manufacturing scales for this technology. DoE and MVDA findings show that the granulation temperature, time, and Macrogol type significantly impact product performance. These factors, by influencing particle size distribution, become key predictors of product quality attributes such as resistance to crushing, disintegration time, and early-stage API dissolution in the profile. Software-aided data mining, with its multivariate and versatile nature, complements the empirical approach, which is reliant on trial and error during product scale-up.