Risk-Based and Data-Driven Strategies for Lifecycle Management of Comparability Protocols

Presented by José Menezes

Professor José C. Menezes is CEO at 4Tune Engineering Ltd., Portugal.

Comparability of biotechnology products that are subject to changes in their manufacturing process is a key activity and a quality commitment that companies “pledge” to regulatory authorities in their product filings. Such assessments are required routinely over a product’s lifecycle to ensure that its quality consistency translates into a high degree of confidence that intended manufacturing changes have no risk to patients and that safety and efficacy of those biologic drugs are not affected.

Process changes are implemented over a product’s lifecycle to improve process economics and product quality as well as to introduce new technologies and adjust production capacity to demand. Since ICH Q5E was introduced more than a decade ago, comparability protocols (CP) involve comparing outcomes of different analytical techniques one by one on preand postchange product samples (e.g., structural and physicochemical characterization, biological activity, and product purity). Information extracted from each technique is limited to (for example) a few peaks or the mean of key indicators computed over a number of batches before and after a change is introduced. Because not all information provided by each method is used, and because comparison is made one technique at a time, an overall quality comparison is not achieved. The possibility of missing quality profile mismatches or of overstating the degree of comparability is real. The current practice of basing the use of analytical techniques on prior knowledge of what regions of their analytical domains should be taken into account (and which should not) does not make use of the new and more powerful ways of extracting and combining unused information made available by the different methods used.

This presentation focused on a different approach to comparability analysis, which my company has used successfully in its projects. This type of analysis has regulatory support in a recent FDA draft guidance (published in April 2016) that revisits and updates the concepts in ICH Q5E by establishing a framework to promote continuous improvement in product quality over lifecycle. There is an expectation that other recent FDA documents (2014–2016) could propose more robust control strategies for commercial products by encouraging the use of quality risk management (QRM) as a new good manufacturing practice (GMP) requirement over a product’s lifecycle and as a way of building quality by design (QbD) elements into legacy filings. This new approach uses multivariate analysis to extract relevant information from whole analytic domains and then combines information from several analytical methods together — in principle, all those that were filed as to characterization methods for a product.

Our approach not only can assess quality consistency holistically (by using the entire set of bioanalytical methods in the filing and to their fullest potential), but it can also establish the link from quality consistency to clinical consistency (safety and efficacy are guaranteed by design). The residual uncertainty in comparability is estimated in the quality domain and can be extended to the pharmacological domain if data are available to establish that correlation. We have a proof-of-concept successfully completed in this area. In that particular project, we have used our proprietary platform iSEE™. This platform is a relational system for data, information, and knowledge aggregation that can not only aggregate data from different analytical techniques for different batches (before and after the change), but also stores all chemical and manufacturing controls (CMC), processing conditions for each unity operation (endto- end), and clinical performance data to enable the type of analysis described. The formal risk assessments needed for each characterization technique and for the overall comparability protocol are performed within iRISK™, a state-of-theart risk management platform proprietary of our 4Tune Engineering Ltd.

In conclusion, we assist clients with comparability protocols over their entire product lifecycle using multiple analytics and proprietary platforms that support science- and risk-based comparability assessments. Our approach ensures improved late-stage development of biologics and a close control and significant mitigation of risks during commercial manufacturing.