Issue link: https://www.e-digitaleditions.com/i/1248960
14 Pharmaceutical Technology BIOLOGICS AND STERILE DRUG MANUFACTURING 2020 P h a r mTe c h . c o m Manufacturing processes, and weight checks. It is clearly impracti- cal, however laudable the goal. No cleanroom that operates commercially today could satisfy this ex- pectation, especially when line set-up is predomi- nantly a manual process. Restricted access barrier systems (RABS) and isolators would also fail this 'requirement'. There's an inference within such observations that bears further discussion and what might seem a simple, but radical solution. The author developed a media-fill protocol for a lyophilization process some years ago. It was problematic because the production batch record included placement of a thermocouple in a vial on each shelf. This had been the practice in an older facility, but now it had to be performed after automated loading the entire chamber with vials through a 'pizza oven' door. The expected intervention required opening the full lyophilizer door before and after each cycle for placement/re- moval of the thermocouples. The protocol omitted the thermocouples entirely, along with a separate recommendation that thermocouples also be elimi- nated in routine production. After discussion with all involved disciplines, this seemingly heretical ap- proach was adopted. We collectively acknowledged that the presence of thermocouples in the corners of the shelves provided minimal insight into the ly- ophilization process. The cycle was executed by the control system independent of the vial temperatures in any event. Removing the thermocouples elimi- nated manual interventional activities in both the routine process and media fills. Thermocouples in freeze drying chambers have some utility during development and scale-up, but should not be a part of any commercial process. As a rule, the perfect intervention is the one that is eliminated from the process (21). Risk assessment and interventions Over a decade ago, the author co-authored arti- cles on aseptic processing risk assessment (22,23). In that assessment, an arbitrary scoring system was employed for both the type of intervention (inherent or corrective) and proximity (distance from exposed sterile materials). In a more recent publication, these were revised to utilize duration (the time required to execute an intervention) and distance (based on an inverse square model) (24). More than 25 years of client work on interventional practices has revealed some other important les- sons worth sharing (25–27): Just because an intervention is possible does not mean it needs to be included in process simula- tions. Extensive interventions may introduce exces- sive risk to sterile materials such that they should never be performed during operations. When the need for such an invasive activity arises, it is pref- erable to cease further operation, secure all sterile materials, and consider what has been produced as a separate lot. The required activity can then be performed, followed by sanitization/sterilization, and a full re-set of the process/equipment. Further production would be considered a separate lot. Interventions included in the process simulation should be limited to activities that directly affect the sterility of materials. The production operator's activities in the background environment can cre- ate contamination on their gloves and gowns, so these background activities will still be necessary. They are executed following the general principles of aseptic technique, but because they are outside the critical zone, they are not aseptic interventions and do not require inclusion in a process simu- lation. Corrective interventions that disrupt the Contin. on page 43