Continuous Manufacturing Abstracts

A Case Study in GMP Continuous Flow Drug Substance Production: Ensuring Quality

Kevin Cole, Ph.D.

Over the past decade, Lilly has invested significantly in developing both drug substance (DS) and drug product (DP) processes using continuous manufacturing (CM). Over this time, we have recognized the ability of CM  to expand the available processing design space, increase process safety, improve product quality, and aid manufacturing flexibility while minding overall product cost.  Our group has demonstrated numerous examples within development and cGMP manufacturing, and this case study will focus on the utilization of CM to produce cGMP active pharmaceutical ingredient (API) within Lilly manufacturing for a mid-stage small molecule clinical asset.

For this example, the registered sequence featured 8 continuous unit operations, which included reactions, an extraction, distillations, and a crystallization/filtration.  Handling of a high containment intermediate wasMany of the CM operations included the use of process analytical technology (PAT). 

Using this process as an example, we would like to expand upon the following questions as part of this case study:

  • How did you define the boundaries of the material that is subjected to a release decision?

Due to the multiple continuous unit operations running simultaneously for this product, surge tanks were often used between operations. These vessels decouple the unit operations and simplify the operational logistics including the start-up and shutdown transitions.  Drawbacks to the use of surge vessels include the increase in material in the process at one time as well as broadening of the residence time distribution (RTD) and expanded deviation boundaries.

  • How do you identify material that has questionable quality (potentially non-conforming material) and how do you deal with it?

During the campaign, a drift from steady state reaction conditions was observed via the on-line HPLC chromatograms, which indicated that the process was drifting away from the desired space.  Since the system RTD had been characterized, we were able to rapidly model of the system dynamics, and informed steps could then be taken to ensure that overall product quality was maintained.

  • How do you assure that materials are collected while the process in a state of control? How do you track the state of the process?

In the current process, a combination of Level 1 and 2 control algorithms were consistently demonstrated.  Level 1 controls were used to maintain tolerances on pump flow rates, temperatures, pressures and level control. Level 2 control, which is a combination of end product testing, raw material attributes and operation within set process parameter ranges was also utilized. For this process, another consideration is that the overall residence times of the continuous sections were typically on the order of hours. These longer residence times serve to buffer the system from the impact of upstream disturbances and immediate automated response through Level 1 control is not required. 

  • How can the operator stay informed, in real time, about the state of the process?

The equipment was connected to the DCS in order to achieve automation control and data recording of process parameters such as pressure, temperature, flow rate, level and valve state. Alarms were triggered if certain parameters strayed from the desired setpoints.  In this case, information derived from the PAT was not integrated into the DCS, which required it to be monitored by operations staff.