Sampling and diversion strategy for twin-screw granulation lines using batch statistical process monitoring

Area of Science:

• Pharmaceutical Manufacturing
• Chemical Engineering
• Process Analytical Technology

Background:

• Continuous manufacturing is a recognized pharmaceutical technology, but its application is hindered by complex process translation.
• Process stability evaluation and sampling/diversion strategies are critical yet challenging aspects of continuous process design.

Purpose of the Study:

• To develop and validate a novel methodology for evaluating process stability at the product key (PK) level in continuous manufacturing lines.
• To provide a data-driven approach for optimizing sampling, diversion, and continued process verification in pharmaceutical production.

Main Methods:

• A statistical process model was developed and calibrated using data from three manufacturing campaigns.
• The model's efficacy was verified on five independent manufacturing campaigns.
• The methodology focuses on evaluating process stability at the individual product key (PK) level, accounting for both continuous and discrete operations.

Main Results:

• The developed batch statistical process model successfully identified outlying product keys (PKs) within manufacturing campaigns.
• The methodology demonstrated effectiveness in evaluating process stability at the PK level for continuous twin-screw granulation lines.

Conclusions:

• The new methodology offers a robust framework for rationalizing sampling strategies and designing effective diversion strategies in continuous manufacturing.
• This approach provides enhanced perspectives for continued process verification and holds potential for future application in quality decision-making.