Industry Group Exploring Number of Batches for Early Stage Validation

Deciding how many early stage validation batches to run is still a challenge for pharmaceutical manufacturers in implementing FDA’s 2011 process validation guidance.

Joanne Barrick, an advisor of global validation at Eli Lilly, said at an Oct. 8 process validation conference that “selecting the relevant Stage 1 batch to include may be one of the remaining challenges of implementing the lifecycle approach in process validation. Is it a statistical question? Maybe or maybe not.”

Barrick also said that “I am currently unaware of any published document that addresses this question.”

She announced at the International Society for Pharmaceutical Engineering (ISPE)/Product Quality Research Institute conference in Silver Spring, Md., that ISPE plans to release a discussion paper to clear up the confusion.

Barrick is a member of the ISPE team that is developing these papers.

She polled the participants attending the meeting and found that 45% of the workshop participants said they are still defaulting to three batches without scientific or statistical justification, 22% use statistics, 12% use historical process performance, 12% use knowledge and data from similar processes and 10% use risk assessment tools.

Barrick said that many of the details of how many batches to run were not specified in FDA’s process validation guidance and were left up to industry to figure out.

The general framework laid out in FDA’s 2011 process validation guidance is for manufacturers to adhere to the principles endorsed in the International Conference on Harmonization Q8, Q9 and Q10 trio of quality guidelines and take a science- and risk-based approach to designing process validation programs.

The guidance proposed a three-stage lifecycle approach to process validation. Stage 1 is for process design, Stage 2 is for process qualification and Stage 3 is for continued process verification (Also see "FDA Guidance Exempts Legacy Products from Early Stage Process Validation; Endorses Use of PAT" - Pink Sheet, 1 Feb, 2011.).

ISPE issued two discussion papers in 2012 to help manufacturers estimate the number of batches to run during Stage 2 and Stage 3 as well as how to establish robust continuous verification programs for drugs once they are released for commercial manufacture (Also see "Pharmaceutical lndustry Proposes Ideas for Number of Batches, Monitoring Production" - Pink Sheet, 28 Feb, 2013.).

Barrick said the upcoming discussion paper will propose ideas for answering the question, “how do I decide whether the process is capable and robust and whether I have enough confidence to begin releasing batches to Stage 2.”

Barrick said that the upcoming discussion paper will specifically consider the following questions in guiding manufacturers’ decision on how many Stage 1 batches to run:

• Was the Stage 1 batch produced in a commercial facility and on the same equipment as the commercial process?
• Is the process scale dependent? Barrick said that “if this is not known the answer should probably be yes, you should assume that it is indeed scale dependent.”
• If the process is scale dependent do models characterize the impact of scale? Have these models been verified with full scale data? Barrick said that “maybe you are filtering on one column and you have five identical columns you will be using for the commercial process.”
• Is the batch size the same as for routine production?
• Is the process control strategy similar to the commercial control strategy?
• Was the batch used for clinical trials?
• What is the source of the active pharmaceutical ingredient? Was the API made from a validated process?
• Is the age of the batch important?

Barrick said that the upcoming discussion paper may use a red, green and yellow scoring system to assess risk.

Barrick said that the number of batches in Stage 1 may be lower than the number of batches in Stage 2.