Yanes Daniel, Shinebaum Rachael, Papakostas Georgios, Reynolds Gavin K, Swainson Sadie M E
Oral Product Development, Pharmaceutical Technology & Development, Operations, AstraZeneca, Macclesfield SK10 2NA, United Kingdom.
School of Pharmacy, University of Nottingham, University Park Campus, Nottingham NG7 2RD, United Kingdom.
Int J Pharm X. 2025 May 22;9:100339. doi: 10.1016/j.ijpx.2025.100339. eCollection 2025 Jun.
Maintaining flowability of pharmaceutical blends is critical for operational efficiency in state-of-the-art continuous direct compression (CDC) manufacturing, with poor flow potentially resulting in API loss, increased experimental work and increased time to market. Consequently, flowability is a crucial consideration in the design of formulations and must be considered throughout the development process when changes are introduced. Traditionally, understanding flow properties has required testing large amounts of material, particularly when evaluating formulation options. This has led to research into developing predictive flow models to reduce experimental burden. Current models with good predictive capacity, such as using granular bond number, require non-routine measurements such as mechanical surface energy. Three mixture designs, each using three pharmaceutical materials, were developed to investigate flow properties and allow the evaluation of a number of mixing models for predicting flowability with minimal experimental input requirements. The resultant models ranged in complexity from simple first order mixture models to more complex third order models with binary and ternary interaction parameters. An analysis of the experimental cost versus prediction accuracy suggested that while the more complex models delivered the most accurate predictions, a first order mass weighted model using inverse FFC was capable of providing good predictions in return for a more manageable experimental burden, with an R value of 0.68, root mean square error of 2.88 and a mean absolute percentage error of 0.21. This model has the potential to provide valuable insights during early formulation design and development where material is scarce and good flowability is crucial.
在先进的连续直接压片（CDC）生产中，保持药物混合物的流动性对于操作效率至关重要，流动性差可能导致活性成分损失、实验工作量增加以及上市时间延长。因此，流动性是制剂设计中的关键考虑因素，在引入变更的整个开发过程中都必须加以考虑。传统上，了解流动特性需要测试大量材料，尤其是在评估制剂选项时。这促使人们开展研究以开发预测流动模型，以减轻实验负担。目前具有良好预测能力的模型，例如使用颗粒键数的模型，需要进行非常规测量，如机械表面能测量。开发了三种混合物设计，每种设计使用三种药物材料，以研究流动特性，并允许在最少实验输入要求的情况下评估多种预测流动性的混合模型。所得模型的复杂程度各异，从简单的一阶混合模型到具有二元和三元相互作用参数的更复杂的三阶模型。对实验成本与预测准确性的分析表明，虽然更复杂的模型能提供最准确的预测，但使用反FFC的一阶质量加权模型能够以更易于管理的实验负担提供良好的预测，其R值为0.68，均方根误差为2.88，平均绝对百分比误差为0.21。该模型在早期制剂设计和开发阶段（此时材料稀缺且良好的流动性至关重要）有可能提供有价值的见解。
