Development of a robust optimization strategy for primary drying in lyophilization, considering intra- and...

Area of Science:

• Pharmaceutical Engineering
• Chemical Engineering
• Process Systems Engineering

Background:

• Conventional freeze-drying models lack robustness due to unaddressed process variability (vial-to-vial, batch-to-batch).
• Existing models limit Quality by Design (QbD) applications in freeze-drying process development.
• Need for advanced modeling to capture inherent process variations for reliable scale-up and manufacturing.

Purpose of the Study:

• To develop a multi-batch stochastic-spatial simulation framework (MBS3F) for mechanistic modeling of primary drying.
• To explicitly incorporate spatial and stochastic variability into freeze-drying process modeling.
• To enable robust process design and identify optimal operating conditions aligned with QbD principles.

Main Methods:

• Developed a multi-batch stochastic-spatial simulation framework (MBS3F).
• Linked vial heat transfer coefficient (Kv) to shelf position to model intra-batch and batch-to-batch variability.
• Implemented a three-stage optimization strategy: deterministic design space definition, probabilistic robustness evaluation, and statistical process control (ICH Q8-Q10 aligned).

Main Results:

• Model validated against experimental data (5% w/v NaCl placebo) with R²=0.97 for product temperature and 4.4% relative error for primary drying time.
• Successfully accounted for spatial heterogeneity and process variability in primary drying.
• Demonstrated the framework's ability to identify operating conditions ensuring product quality and equipment performance.

Conclusions:

• The MBS3F provides a realistic and reliable approach for defining the freeze-drying design space.
• Integrating stochastic variability with spatially resolved mechanistic modeling enhances process understanding and control.
• The framework supports robust freeze-drying process development and optimization under QbD.