(441a) Leveraging Process Analytical Technology to Understand Scale-up/Scale-Down Behavior in the Crystallization of a Drug Substance Intermediate

In the synthesis of a pharmaceutical drug substance, scale-up of chemical steps can often result in challenges due to the complexity of the chemistry and the impacts of scaling factors such as mixing, mass transfer, and heat transfer. This presentation will discuss how scale-up challenges were addressed in the penultimate step in the synthesis of an active pharmaceutical ingredient. In this work, Process Analytical Technology (PAT) was utilized to provide a benchmark for this step during pilot-scale processing. We will demonstrate how this approach allows the process development team to troubleshoot an unforeseen challenge and identify development targets for future work.

The penultimate intermediate is isolated from a cyclization reaction via workup of the end of reaction stream and antisolvent addition crystallization. The cis-enantiomer of the penultimate intermediate forms in the cyclization reaction and is thermodynamically soluble in the end of crystallization stream. During pilot-scale demonstration of the cyclization step, poor purge of the cis-enantiomer was observed in the dry cake. In the investigation of this result, scale-down experiments with Focused-Beam Reflectance Measurement (FBRM), Fourier Transform Infrared spectroscopy (FTIR), and in-process microscopy were leveraged to establish fingerprints of the crystallization using process streams generated in the lab and pilot plant. This exercise highlighted differences in crystallization kinetics that were correlated with variable purge of the cis-enantiomer. The data collected also indicated a potential mechanism for entrapment of cis-enantiomer in the isolated cake via solvent occlusion which occurred during the initial portion of antisolvent charge. The process development team leveraged this PAT data to evaluate short-term mitigations that could be implemented at the pilot scale, such as annealing cycles, seed charge, and slower antisolvent charge. At the conclusion of the pilot campaign, scale-up and scale-down data sets were assessed and long-term areas of focus were identified for increasing understanding of the crystallization. Additional improvements to the crystallization involve addressing solvent occlusion through techniques such as cofeed addition and a larger seed charge. In summary, this presentation will explore short-term troubleshooting approaches for process scale-up and long-term process improvements towards a robust end-to-end process.