Issue link: https://www.e-digitaleditions.com/i/1520706
20 BioPharm International ® Partnerships for Outsourcing eBook May 2024 www.biopharminternational.com Quality agreement from the start. While this may add initial time and cost, over the long run this will pay out with less process issues, including errors or oversights. Defining a relationship A CDMO should ask a lot of questions and ta ke a consultative approach to building the quality agree- ment. Take a close look at their platform and inter- nal quality standards and how this aligns with the needs of a product. Set the expectations out clearly, including defining the roles and responsibilities for the CDMO and customer. The agreement will clar- if y when and how to incorporate customer specific requirements and identif y any areas for risk to ad- dress from the start. Ensure t hat during t his stage, and t hroughout the relationship, there is always open and honest dialogue about ever y detail. For instance, what is considered a repetitive dev iation and when does a customer feel the need to sign off on it. The cus- tomer may define the second occurrence of an event as repetitive, while the CDMO might define it as the third occurrence. T he CDMO shou ld ca ref u l ly re v iew c u s tomer requests and audit obser vations, accept, and im- plement those that improve the process, but a lso be willing to explain when customer obser vations cannot or will not be implemented. The culture of the CDMO is important. Consider what type of company culture and service approach is comfortable. Is the agile approach of a small- to medium-size CDMO better, or maybe working with a larger multinational CDMO feels more secure? Before st a r t i ng t he request for proposa l (R F P) process, have requirements laid out and source up to 10 CDMOs based on how the offer fits the needs. Consider whether it will just be batch manufactur- ing or are process development, analytical develop- ment, process optimization, and robustness studies needed when drafting the RFP. Once it has been narrowed down to two top con- tenders, always do a pre-contract audit. This is the best way to start an open relationship. For the first audit, look at the maturity of the quality systems. Whether the CDMO is working with comprehensive electronic systems or using paper systems is less im- portant than if their approach meets the needs of the process. Assess if the PQS provides the appropriate level of oversight, and whether the metrics demon- strate control. Ask about who would do the actual work on the product and ensure the qualit y team have a good understanding of risk management. A smaller, more agile CDMO has the advantages of keeping the project moving quickly and taking a proactive approach to solving problems. A larger, formal CDMO will have all the qualit y controls in place to ensure compliance but may slow the project down and make resolving problems more complex. There will always be some element of compromise when creating these relationships. Balancing assays When moving into process development, it is impera- tive to clearly define what impurities and degradation products need to be tested for in both release and sta- bility formats. In the early stages there will be a rolling shelf life based on the stability data, so this must be ac- curate and available quickly, having to perform multiple out-of-specification/out-of-trend (OOS/OOT) investiga- tions to justify the data may delay clinical programs. Trending variable data and extrapolating to the poten- tial OOS time point is much harder with a variable assay. Consider whether there is a small-scale model that can generate enough material for analytical develop- ment instead of waiting for full-scale batches. Always be working on a long-term approach to assay development regardless of phase. Traditionally, there has been the excuse of 'phase-ap- propriate' assays because they have the allure of reduc- ing costs and timelines at the early stages. If not man- aged properly, this is something that gets paid for later many times over when small details create big problems that can impact on commercialization timelines. It is not worth risking generating clinical data with a sub-opti- mal release assay that will then be locked into use for later stages. This can end up creating a great amount of work to demonstrate a new or improved assay is compa- rable and is also expensive and delays the process. When taking an assay developed by a previous CDMO or even the partner company, a full evaluation should be done to identify any deficiencies in the assay and properly optimize the assay early on. Consider what the When moving into process development, it is imperative to clearly define what impurities and degradation products need to be tested for in both release and stability formats.