Computational Predictability of Microsponge Properties Using Different Multivariate Models.

Area of Science:

• Pharmaceutical Sciences
• Materials Science

Background:

• Metronidazole benzoate-ethyl cellulose microsphonges (MBECM) are developed for drug delivery.
• Understanding formulation and process impacts on MBECM properties is crucial for optimizing drug release.

Purpose of the Study:

• To evaluate principal component regression (PCR) and multiple linear regression (MLR) for predicting metronidazole benzoate-ethyl cellulose microsponge properties.
• To establish a knowledge-rich design space for MBECM development.

Main Methods:

• MBECM were prepared using a quasi-emulsion solvent diffusion method.
• Box-Behnken design was employed for minimum experimentation.
• Data was analyzed using principal component analysis (PCA), PCR, and MLR.

Main Results:

• PCA identified two distinct MBECM formulation groups based on parameters like solvent concentration, polymer concentration, and stirring speed.
• Optimized PCR and MLR models predicted particle size, entrapment efficiency (EE), and actual drug content (ADC).
• MLR models exhibited better predictability compared to PCR for MBECM properties.

Conclusions:

• Multivariate modeling, particularly MLR, effectively predicts MBECM properties.
• Developed MBECM in Carbopol hydrogel showed prolonged drug release and good biocompatibility.
• Minimum experimentation coupled with multivariate analysis provides a robust approach for optimizing microsponge formulations.