When the time comes around to plan for process validation, inevitably the question always comes up “how many validation lots are we required to do?”. There is good news and bad news. The good news is that the FDA does not have a requirement for a certain number of PPQ lots. The bad news is that the FDA expects manufacturers to justify the number of lots by providing a clear rationale based on product knowledge and process understanding. Usually, the assumption is that the magic number is three lots. Unfortunately, manufacture of three consecutive lots may not be sufficient to demonstrate process reproducibility.
I recently came across a discussion paper by the International Society for Pharmaceutical Engineering (ISPE) on this topic titled “Stage 2 Process Validation: Determining and Justifying the Number of Process Performance Qualification Lots". Although the authors clearly state that the approach presented in the paper does not reflect a consensus position of the industry, it was what they thought would be best practice, and I think this approach is a great one.
The basic premise is that the number of lots required is driven by a risk assessment. This risk is assessed based on the level of product knowledge and understanding of the manufacturing process, as well as the control strategy of the process. The authors link the residual risk identified from this assessment to the number of validation lots required. A process with higher risk may require more than 3 lots in order to provide enough assurance that the variability between lots is adequately controlled before commercial distribution.
The authors also suggest that the risk assessment plan they present be performed periodically during development. The earlier this is started in the development cycle, the more likely the development teams can design the product to minimize or eliminate the risk or devise robust control strategies. Ideally, there would be no high residual risk items identified when proceeding to process validation. This is key to developing a validation strategy, minimizing the number of validation lots required and providing visibility to allow adequate planning should large numbers of lots be required.
The paper provides a detailed workflow for determining the risk and number of validation lots required. The following steps to determining risk were presented.
- Assess product and process knowledge and understanding (how severe is the risk and what is the probability that it will occur)
- Assess control strategy (can controls be identified to mitigate each risk)
- Determine residual risk after controls are applied (how much risk is left after controls are applied)
Three approaches to determining the number of lots were presented based on the residual risk (the second and third are statistically based).
- Use of rationale and experience
- Target process confidence and target process capability (CpK)
- Expected coverage
For each approach, tables with examples that connect residual risk, recommended number of lots and rationale or acceptance criteria (for statistically based approaches) are provided. These can be used as templates and adjusted to fit your needs. As you might expect, the higher the risk the more lots are required to validate that the process is robust. Conversely, if there is a lot of process history and the risk is minimal, fewer batches may be justified.
I highly recommend reviewing this paper if you are interested in a systematic approach to developing your PPQ plan and want some guidance in determining the number of lots to make during process validation.