Risk-Based Approach to Method Transfer
Jeffrey Staecker, PhD (jstaecker@biohpia.com) BioPhia Consulting, Inc., November 1, 2018
An article summarizing the CMC Forum “Methods on the Move: Addressing Method Transfer Challenges for the Biopharmaceutical Industry” was recently published in BioProcess International and can be found at https://bioprocessintl.com/business/cmc-forums/methods-on-the-move-addressing-method-transfer-challenges-for-the-biopharmaceutical-industry/. The article summarizes various ideas and approaches to test method transfer.
I was fortunate to both co-chair and present at this Forum where I focused on using a “Risk-Based Approach to Method Transfer”. My presentation can be found on the BioPhia website (BioPhia website link inserted here). Some key principles from my presentation include:
Method transfers can introduce changes in bias and/or precision.
The commonly employed risk tool of FMEA (Fault Mode Effect Analysis) can be utilized in evaluating the risk of method transfer.
The risks associated with method transfers are impacted by how “close to the edge” the process as measured by current testing is running. That is, a centered process with results far away from specifications is less at risk than a process being measured as running near a specification extreme.
Many companies use Cpk to monitor processes which can be leveraged to predict probability of errant results from introduction of bias and/or imprecision.
A statistical approach can be devised to ensure that only an acceptable level of bias/imprecision is introduced during method transfer.
After transfer, appropriate SPC (statistical process control) monitoring can ensure that method at receiving site continues to operate within pre-defined limits.
An alternative to method transfer is to use the identical statistical approach for all method transfers regardless of process/method performance. The danger of this approach is to introduce a small amount of bias or imprecisions that passes method transfer criteria but results in inappropriately passing or failing material. I personally experienced this when a method was transferred for a process running at the high end of the specification range. The method transfer introduced a small bias (less than 2%) that resulted in inappropriately failing drug product.