Emerson Exchange 365

Continued Process Verification (CPV) and how it’s changing process validation

Jim Cahill of the Emerson Process Experts Blog recently quizzed me on Continued Process Verification (CPV) and how it’s changing process validation.   Here’s an excerpt from a blog Jim wrote about our discussion. “…this guidance has had a fundamental impact on the pharmaceutical and biotech industries. This is due to the fact that the guideline extends a different dimension into the validation lifecycle than the drug development lifecycle itself… He explained that the FDA started to request more and more statistical analysis for Pharmaceutical Quality/Chemistry Manufacturing and Controls (CMC) submission and science-based proof for proving that the process is in a state of control.”

The traditional three golden batch rule is not exactly true anymore. FDA is not specifying how many more batches will be required because in FDA's current view process validation is not a point activity that will get you FDA approval at the end of stage 2. The three stages defined in the process validation guideline are now instead integral elements of process validation. How much process and product understanding have been developed and what kind of continued process verification program have been incorporated for the commercial operation will have significant impact on FDA's decision. Depending on how much process and product understanding has been demonstrated and how effective the control strategies have been demonstrated to be, the good news is that three golden batches may not necessarily be needed for FDA approval either. FDA explicitly answered No to this question in a question and answer document earlier in 2010. The real challenge the industry is facing may be how to manage the process validation across different platforms, different stages, and different phases as the validation goes from stage1 to stage 3. Level 3 automation and integrated data management through drug development cycle will probably have a big role to play in this direction...

Read the full Emerson Process Experts blog now: full blog here

• lifecycle
• Life Science
• Process validation
• GMP
• Batch Analytics
• PAT
• drugmakers
• FDA
• ICH
• QbD
• PV

zjinemerson

• 20 Feb 2015 7:48 PM
• Cancel

2 Replies

• Victor OBrien

  

  • 23 Feb 2015 6:42 PM

  Has the FDA given any indication that it will go beyond just requesting statistical analysis and science-based proof of a state of control to actually expecting it? It seems more to be that the FDA wants to see a program for process verification at all three stages will increase likelihood of approval, but the FDA hasn't moved toward expecting it as a standard.
  
  Also, you mentioned a lot about stages 2 and 3 and how challenging it is to implement a design space at those stages. What do you think creates those challenges?

  • Cancel
• zjinemerson

  

  • 23 Feb 2015 8:24 PM

  In reply to Victor OBrien:

  I agree that a program for continued process verification will increase the likelihood for FDA approval. CPV is not a reinforcable standard at the moment but most of the pharmaceutical companies are proactively working on such programs expecting FDA to eventually make it mandatory at any time.
  
  Implementing design space is a task at stage 2. The challenge in implementing the design space may include: 1) not all commercially available control platforms have the ability for constraint reinforcement which is one of the main characteristics of a design space. 2) while design space provides quality predictions with certain level of confidence few of the process control platforms have the capability of making control decisions based on these predicted information 3) Complicated dynamics/interactions between the process parameters may exist, which requires advanced control strategy
  
  Process control platform such as DeltaV have upper/lower limit control for set point. DeltaV advance control strategy like model predictive ontrol (MPC) does have the ability for constraint reinforcement on the process parameters. These can all be very useful in implementing design space. However early involvement (such as in Stage 1) of the control expertise is critical in selecting the right control strategy. Decisions such as whether PAT should be used and how the complicated dynamics between the process parameters may be addressed need to be tackled early with access to the right technology.
  
  The challenge at stage 3 is the amount of data and the varity of units being involved. 1) on-line automated monitoring program having access to both real time and historical data 2) powerful algorithum in generating statistics and statistical analysis in real time 3) flexibility in building statistical models for different processes 4) controlled environment for model building and data access 5) consistent plant wide/company wide or even global implementation of CPV program

  • Cancel