(39b) A Perspective On the Use of Mechanistic Models Alongside Statistical Design of Experiments to Define a Design Space in Drug Substance

In Chemical Process Development for small molecule drug substance, the design of an optimal synthesis and manufacturing process must meet key elements of safety, product quality, efficiency, robustness, economics, intellectual property constraints and greenness/sustainability. Defining a design space, in which to operate the process parameters across the sequence of unit operations to meet the design goals, encompasses the work of many chemists and chemical engineers through a range of activities (lab experimentation, data analysis, modeling, and plant scale-up). The resulting data package and design space form the basis for the eventual Chemistry, Manufacturing and Controls section of regulatory filings for New Drug Applications.

The Quality by Design approach emphasized by the FDA in the last decade and the compression of timelines to develop a drug candidate provides a strong need for efficient process development and rigorous scientifically-based approaches to develop this design space. Elements of this approach encompass high-throughput experimentation, modeling (both mechanistic and empirical), data visualization, statistical design of experiments, and appropriate scale-up verification. This talk will provide a perspective on the approach at Bristol-Myers Squibb incorporating these elements.