(M1430-10-58) Rapid Material-Sparing Approach to Address Hairline Fractures in Low Drug Load Tablets via Compaction Simulation

Purpose: While manufacturing tablets for Phase I/II clinical supply, fine cracks were observed across the length of the tablet cores. The formulation had previously been considered low risk because it contained relatively low API loading ( < 20%) and balanced ratio of plastic to brittle filler, and the production of defective tablets jeopardized product integrity and quality. This study presents a systematic approach to troubleshooting and identifying the root cause of hairline fracture defects. Compaction simulation is first used to characterize the compaction properties of neat API and starting formulation, then leveraged to make optimized formulation recommendations for scale-up. This case study highlights the efficacy of using compaction simulation for rapid and material-sparing troubleshooting in tablet formulation, particularly during critical Phase I/II scale-up when significant changes in formulation composition, such as altering drug loading or incorporating additional excipients, are limited.

Methods: Styl’One Evo Compaction Simulator was used to characterize the API and starting formulation blend (Formulation A) by generating slugs across a compression force range of 4-45kN using 11.28mm flat faced tooling and simulating a Korsch XL100 press profile operated at 25rpm. Visual appearance and ejection profile of slugs were studied. Next, compression profiling of Formulation A pre-granulation and final blend was performed using the same parameters as the neat API study. Weight, thickness, hardness, visual appearance, and ejection profiles of the slugs were recorded. The data from the neat API and Formulation A was used to make recommendations for optimizing the formulation and process. Compression profiling was then conducted on the optimized formulation (Formulation B) to determine how the changes affected compaction performance and appearance of the final tablet. A preliminary scale-up batch using Formulation B was manufactured with automated equipment and the same tooling as the initial batch where hairline fractures were observed. The tablets from the scale-up batch were inspected for defects. Finally, a development batch was produced on automated equipment using Formulation B and new tooling that was selected for future manufacturing. Tablets visual appearance and start-up ejection pressure were recorded.

Results: Neat API was found to be sticky, brittle, and abrasive based on the observation of sticking on tableting tooling. It displayed significantly high and irregular ejection pressure of ~24MPa at a compression force of ~20kN. Formulation A pre-granulation and final blends had good tabletability and achieved acceptable tensile strengths of 1.7-2.5MPa at the target compression range of ~15-25kN. However, both pre-granulation and final blends had high ejection pressures – ~15MPa pre-granulation and ~7MPa final blend at ~20kN compression force. Additionally, the pre-granulation blend had a noisy ejection signal, likely due to higher fines content leading to increased contact of API with tooling. As a result, formulation optimization recommendations were made to increase lubricant and plastic filler concentrations. Additionally, a process optimization recommendation was made to increase target solid fraction during roller compaction to minimize fines in the final blend and reduce API contact area with tooling. Formulation B pre-granulation blend and final blend had similar tabletability as the Formulation A blends, achieving acceptable tensile strengths of 1.7-2.5MPa at the target compression range of ~15-25kN. However, Formulation B registered lower ejection pressures and had a less noisy ejection signal. The recommended formulation optimization was successful in lowering ejection forces without negatively impacting overall compression properties of the blend. When Formulation B tablets were produced using automated equipment and identical tooling to the original defective production, no hairline fracture defects were observed. Formulation B was therefore selected for scale-up using automated equipment and new tooling chosen for upcoming campaigns. The tablets produced did not exhibit hairline fracture defects and registered lower ejection pressures than the original defective tablets. The ejection pressure difference was larger than would be expected due to tooling change alone, indicating that the formulation optimization contributed to successful tablet performance.

Conclusion: Neat API compression properties can have a major impact on compression properties of formulated blends even in formulations that don’t generally stand out as high risk. Compaction simulation can be an API sparing tool to quickly understand these properties and provide formulation/process optimization recommendations for full scale manufacturing – the compaction simulation used < 5g API. Additionally, ejection signals may be a key process output in understanding strain during tablet ejection and may be monitored to verify resolution of such strain in optimized formulations. The above study shows that the formulation updates (increased plastic filler and lubricant) and process updates (increased solid fraction of granules) produced acceptable tablets without defects. The optimized formulation performed well regardless of tooling used, providing flexibility for future manufacturing campaigns. The systematic approach applied in this case study to understanding the cause of hairline fracture defects observed during Phase I/II production can be used in other programs to ensure successful root cause analysis and phase appropriate optimization for scale-up while conserving time and material.

Neat API, Formulation A, Formulation B - Ejection Pressure vs Time at Nominal Compression Force (~20 kN)

Formulation A & Formulation B – Tabletability Plot (Compression Force vs Tablet Tensile Strength)

Formulation A & Formulation B – Automated Equipment Tablet Characteristics