Track: Manufacturing and Analytical Characterization - Chemical - Process Design and Controls - QbD and Assessment of Process Parameters
Category: Late Breaking Poster Abstract
A Quality by Design Approach to Optimize and Accelerate Formulation and Process Development Leading toward Registration Batches Manufacturing
Purpose: To improve an existing formulation composition and manufacturing process that was designed for early phase development. Process optimization activities focused on improving the flow characteristics of the granules, minimizing or eliminating segregation and increasing the manufacturing yield with the aim of achieving a robust process prior to the product registration campaign. Methods: Pre-blend (16.17% w/w API load) was prepared in Bohle Bin Blender (300 L at 70% fill volume with 480 revolutions) and further roller compacted using Gerteis Minipactor. Final blend (300 L Bohle bin blender at 55% fill volume with 60 revolutions) was encapsulated at two strengths using Bosch 705 encapsulator. Both pre-blending and final blending processes were evaluated for number of revolutions vs. %RSD to establish the optimal bending time. Critical material attributes, formulation variables and Critical Process Parameters (CPPs) of the roller compaction process were identified. Roller compaction process parameters such as roll force, roll speed, roll gap, and granulator speed on the Critical Quality Attributes (CQAs) of the drug product were studied in sub-batches (Batch size:1 kg). Design of experiment (DoE, JMP software) with 4 factors and 2 levels was utilized to evaluate the parameters of significance in the manufacturing process. Effect of roll force (5-9 kN), roll speed (2-6 rpm), roll gap (1-3 mm), and granulator speed (45-85 rpm) on properties such as powder flow, solid fraction of compacted granules (analyzed using AccuPyc® and GeoPyc®), content uniformity and dissolution was evaluated as response variables to establish design space. Effect of granulator screens (0.8 mm and 1.25 mm vs target 1.0 mm) was studied in separate experiments. Based on these responses, risk levels were identified, and acceptable processing ranges were established. Point optimization analysis was performed based on the input of ideal target response variables to identify and establish target process parameters. Data processing to interpret experimental results was carried out on Minitab® Version 19. Impact of Filler 1 to Filler 2 ratio (1:3 to 3:1), disintegrant levels (2 to 8%) and lubricant levels (0.2 -0.8%) were also studied for their impact on drug product CQAs. Clinical batches manufactured (batches at 85 kg scale using optimized process parameters) were placed on stability to generate data to support registration filing. Results: Solid fractions of compacted ribbons studied at various roller compaction conditions were within the range of 0.63-0.78 with optimum solid fraction values of 0.70 (Figure 1). Similarly, optimum ranges for roll speed, roll gap, roll force and granulator speed were established with predicted desirability of more than 90% (target parameters selected: roll force of 7 kN, roll speed of 4 rpm, roll gap of 2.5 mm, and granulator speed of 65 rpm). The order of impact of process parameters in influencing granule properties was roll force > roll gap > granulator speed > roll speed. Uniformity sample analysis demonstrated that risk to uniformity reduced by adjusting the number of blender revolutions and maintaining the ideal blender fill volume throughout the scale up. The process was scaled up successfully from 11 kg to 85 kg by maintaining the number of revolutions constant from 40L bin to 300L bin with RSD values less than 2. Blends consisting of varying amounts of filler, disintegration and lubricant levels processed at optimized conditions resulted in acceptable dissolution profiles confirming robustness of the formulation developed. Selection of appropriate capsule size and dosing disc allowed manufacturing of a range of doses for clinical supply to support ongoing clinical trials and to finalize suitable doses for registration batches. Parallel formulation and manufacturing process parameters optimization allowed the program to accelerate towards registration batch manufacturing phase once the final capsules strengths are confirmed. Conclusion: The adoption of parallel formulation development and QbD approach towards process optimization resulted in the identification of CPPs. Risks were successfully reduced to acceptable levels within the studied design space, and a robust manufacturing process producing a desired quality product for future registration and commercial manufacturing was defined. The work performed established flexible unit doses to meet the clinical needs and resulted in significant reduction in both drug substance consumption and the overall product development time.