The ongoing challenges of applying QbD to biopharmaceutical products
Once again, QbD was a hot topic of discussion at this year’s BPI conference (Sept 21-24, 2010). Perhaps of greatest significance was the keynote address, “Regulatory Modernization-FDA’s Desired State for Product Quality” by Helen Winkle, Director of FDA’s Office of Pharmaceutical Science. Ms. Winkle discussed ways in which the agency is striving to improve the regulation of product quality and the opportunities and challenges of implementing QbD for biotech products. She presented the results of a McKinsey & Company study which found that there was significant confusion and disagreement within the agency as to exactly how QbD should be applied to biotech products and what type of data would be required to define a biotech design space. Collaboration between the field inspectors and the review and compliance sectors of the agency on the handling of QbD was viewed as a challenge that still needs to be met. We wonder why such collaboration and consensus is viewed as something new; surely these folks must collaborate in the normal course of operations, or did it really require the QbD initiative to bring this about?
The McKinsey study found the biotech industry “stepping up to the plate” in adopting QbD and that QbD is “evolving, gaining momentum, and passion throughout the industry.” Meanwhile, FDA wrestles with its internal contradictory views, leaving companies with little or no guidance on the development of QbD programs and the definition of specific process design spaces beyond what is coming from the analysis of submissions in the QbD Pilot Programs. Unfortunately, Ms. Winkle didn’t give any indication whether FDA is satisfied so far with the way that QbD has been implemented by industry. Nevertheless, the potential benefits of QbD on the business side have been discussed and many companies are adopting its concepts as standard business practice. However, successful implementation of QbD requires cooperation across a multitude of disciplines, including process development, manufacturing, and QC, to ensure its incorporation from the very beginning of product design through the entire life cycle of the product.
The development of a biotech process design space requires the careful analysis of process data obtained during development and routine manufacturing along with data from statistically designed experiments to develop appropriate process models and predict process performance. Based on recent publications and presentations, we believe that the amount of data being submitted in support of QbD may be more than is strictly necessary, as most efforts to date (c.f., D Kenett , Cook G, Welin M , AS Rathore , AS Rathore, et al ) have focused on retrospective design space definition of existing processes based upon several years of process experience, multiple DoE studies and detailed multivariate analyses. While these efforts may be necessary to convince FDA that existing processes are well controlled, this large volume of data may not be available at the time of BLA filing for a new product. A clearer idea of just what data and how much would constitute an acceptable QbD-based submission is definitely needed to help companies achieve the overall goal of “…reliably produc(ing) high-quality drug products without extensive regulatory oversight.” We only hope the agency gets its house in order soon and provides industry with a unified clear path forward.