Factors Influencing the Dispersibility of Glycopyrronium Bromide and Indacaterol Maleate – Combined In Vitro and In Silico Study

Abstract

The development of dry powder inhalers (DPIs) for pulmonary drug delivery is complex, requiring optimization of variable factors to ensure effective lung deposition. This study investigates the factors influencing the dispersibility of glycopyrronium bromide (GLP) and indacaterol maleate (IND) in adhesive mixtures using both in vitro and in silico approaches. The formulation was designed to match the reference listed drug (RLD), using lactose and magnesium stearate as excipients. Key variables examined included mixing energy, carrier particle size distribution (PSD), and active pharmaceutical ingredient (API) particle size characteristics across multiple suppliers.

A Next Generation Impactor (NGI) was employed to assess the aerodynamic particle size distribution (APSD) of 67 formulations. The collected impactor data were analyzed using machine learning (ML) models, leveraging the h2o AutoML framework. Stacked ensemble models demonstrated high predictive accuracy (R²: 0.940 for GLP, 0.969 for IND), identifying key formulation parameters affecting dispersibility. SHAP analysis revealed that GLP dispersibility was influenced primarily by GLP PSD (d90, d50, SPAN), lactose d10, and mixing energy, while IND was more dependent on lactose PSD and its own particle size.

The findings confirm that both APIs interact with each other within the formulation, significantly impacting their reciprocal deposition profiles. These insights highlight the challenge of developing bioequivalent DPI formulations and emphasize the importance of PSD control, mixing energy optimization, and advanced ML modeling in predicting therapeutic equivalence. The study provides a predictive framework to support the development of generic inhalation products, improving regulatory approval pathways and ensuring effective pulmonary drug delivery.

Graphical Abstract