Considerations for demonstrating product comparability and process consistency with cell therapy products
View Presentation View Presentation
*Thomas Finn, FDA/CBER 

Keywords: cell therapy, product comparability, process consistency

Cell therapies offer tremendous potential for treating a wide array of serious medical conditions, and may be the best strategy for some very difficult to treat diseases. Some cell therapies are personalized to the degree that product lots are manufactured for each individual patient. But challenges for manufacturing these products are numerous, and for some cell therapies demonstrating consistent product quality can be difficult. Two main areas of concern are 1) that a manufacturing process consistently generates a quality product; and 2) that product quality is being maintained after a manufacturing change. Addressing these concerns is not always easy, and approaches used for traditional drugs or biologics may have to be adjusted to accommodate the more complex nature of cell therapies. Some personalized biologics can vary 100 fold lot-to-lot due to inherent variability of source material from patients. The level of product testing feasible for lot release may be limited for many practical reasons. In some cases limited product quality testing data may need to be supplemented with additional testing in order to establish product comparability or manufacturing consistency. Despite these difficulties, personalized autologous and allogeneic cell therapies also present some unique statistical opportunities. The number of lots generated during the IND phase can be many hundreds or even thousands, and for commercial products thousands per year. One can leverage these large data sets to evaluate manufacturing data in ways that cannot be done for more traditional biologics where small numbers of lots are generated. Ways in which statistics can be very helpful for evaluating cell therapy manufacturing goes beyond questions about product consistency and manufacturing, and can include manufacturing and distribution logistics, data trending to look for patterns among routine manufacturing results, and conducting stability studies. In some cases a statistical analysis can also be part of a post-marketing commitment. In this presentation I will discuss some of the regulatory concerns with traditional approaches to analysis of CMC data for these types of products, and cover some common flaws in study design and analysis.