Un-Refined Quality Monitoring

This week, I will address the second of nine common CGMP Quality System shortcomings; Un-Refined Quality Monitoring.

When limited development information is available, Critical Quality Attribute (CQA) and specifications may have been established too loose or too tight. If set too tight, the CQA and/or specifications may be regularly exceeded, thus causing extensive investigations and/or product to be scraped, reworked, and/or released under questionable justifications. If the CQAs and specifications are set to loose, poor product quality may go undetected. Setting precise CQAs, specifications, and processing parameters can be difficult to achieve using limited small-scale and bench-scale data. During a typical technology transfer, a set of CQA parameters and specifications are inherited from the development group’s smaller-scale experiences. Often, these CQA parameters and specifications have not been exactly determined and tend to be a best guestimation. Critical processing parameters are established with the goal of meeting these CQAs and specifications. Phase III clinical lots are made using these processing parameters and specifications. If these clinical trials are successful, then the CQAs, processing parameters, and specifications become the gold standard for determining product quality. I have often heard people simply justify process deviations by stating specifications were met. This may be acceptable justification or it may not, depending upon the how the specifications were determined in the first place.

During commercial scale-up, process validation is typically completed in full-scale equipment using these processing parameters and specifications. Often, the limits of the processing ranges are not tested during process validation and sometimes not even tested with bench-scale studies.

CQA and specification ranges must be established to allow for inevitable processing and testing variability, but obtaining product quality and/or stability data at the range’s extremes is generally not obtained by the time of commercial manufacturing. Due to this limited processing/product knowledge at the start of commercial production and the learning curve associated with new processes, procedures, and assays, manufacturing management often experience difficulties with the product and/or process during the first year of production. As manufacturing works through these difficulties, significant process and product characterization knowledge is obtained. After several months of commercial manufacturing, process/testing variability typically settles into a more stable pattern.

To overcome this shortcoming, the Quality Unit should closely monitor all critical processing parameters using Statistical Process Control (SPC) to enhance their process characterization knowledge. To facilitate this endeavor, product reviews should be completed quarterly instead of annual for new production lines. Ideally, better product and process characterization should be completed prior to initiating commercial manufacturing with the goal of obtaining each critical processing parameter’s edge of failure.