Embedding QbD elements into drug development could make approval timelines and inspection success more predictable, say a Genentech team.
Quality-by-Design (QbD) takes effort but, say researchers, a science- and risk-based approach consistent with its philosophy can be adopted without excessively increasing workloads. In fact, companies, regulatory agencies and patients can all benefit from the approach.
“Overall, we believe [our QbD system] will have a favourable impact on drug product quality for patients and associated cost for companies and regulatory agencies”, say a Genentech team writing in the Journal of Pharmaceutical Sciences .
The proposed risk-based biopharm drug development model uses prior knowledge of processes, risk assessments and experimental design. Using prior knowledge unnecessary experiments can be eliminated, keeping extra work to a minimum while creating a stronger development plan.
Following the proposed model will create more standardised regulatory submissions. The logic of these filings will be easier to follow, say the team, and will give more confidence in the process and product control strategy.
“QbD submissions may be longer but stronger submissions, and there may be less need for supplemental submissions during the lifecycle of a product”, say the Genentech team.
Compliance may also benefit from defining acceptable deviations. For example, studying the impact of a temperature change on quality during design space work, defined on page 9 of this document , could minimise the effort required if this event actually happens.
Overall the team sees a number of benefits for those adopting QbD in biologics development. “An enhanced rigorous and standardised process…may result in greater opportunities for flexibility and post-approval process optimisation as well”, says the paper.
Extra costs relate to initial investment in development of small-scale models of processing steps and robustness studies. Investing in robustness studies supports work on design space and there are also benefits to using small-scale models.
“It is worth the investment to design small-scale models or at-scale surrogate models for each unit operation. This is the best way to study the impact of the process on product quality and interactive effects when needed”, says the paper.
Linear scale down of some operations, such as thawing and freezing, is impossible though. Consequently, the team recommends creating models that mimic product stresses in a worst-case scenario. These models test parameters outside the standard manufacturing ranges.