Markus Krumme, Ph.D.
The case study describes, in brief, the process of an E2E continuous process as an example and how the technical control strategy needs to be reflected in the regulatory control. It introduces a fairly generalizable concept for material traceability, the tracking of the state of control of the individual unit operations and the translation into the state of control of the process chain as a whole. These elements are the more relevant the longer the chain of an E2E process is and as such the choice of an E2E process forces the balancing of quality aspects, traceability and practicability.
The key message is that the underlying process dynamics needs to be characterized in a mighty enough way to cover all unit operations from chemical reactions, over purification, crystallization and the pharmaceutical operations, as the material progresses seamlessly and more importantly without material holding point from liquid based processes into solid processes, which have the fundamental difference of much different axial dispersions. A general concept and options are presented and briefly explained.
The longer the process chain is, the more impactful is the backmixing. The concept of truly continuous processes from start to finish needs to be described as a probability distribution without sharp boundaries which is different from the batch concept. The proposed concept of detect-flag-divert manages the risks associated with the long chain but requires a new consistent traceability concept and ideally is considered during process design already.
An alternative of a hybrid process description is also given, which is a bit more pragmatic and potentially operationally easier, but less consistent and offers less opportunities for advanced controls.
The last element is the discussion of a new concept for a qualification approach that avoids a classical validation effort. This is based on the fact that the batch size is decoupled from equipment constraints and on the concept of continuous monitoring of the health state of the process that allows to dynamically extend the qualified runtime of the process as the demand grows.