Measuring quality in real-time improves the overall performance of a manufacturing process. The International Foundation Process Analytical Chemistry (IFPAC) is:
…a world-wide, not-for-profit organization dedicated to the advancement of Process Analytical Technology and Chemistry.
They are holding the IFPAC Annual Meeting later this month in the Washington D.C. area. This meeting provides a forum for discussions in the latest trends and real-life applications in the fields of Quality by Design, Process Analytical Technology, Quality Metrics, Continuous Manufacturing and Process Control applications for the pharmaceutical, biotechnology, generic, chemical, petrochemical, food, and related industries.
Many processes, such as those in clinical manufacturing and facility start-up, do not have enough batch history data for building a consistent Multi-variate Data Analysis (MVDA) model but may have plenty of insights from process development and pilot plant for establishing an ideal batch (golden batch) profile for process monitoring purpose. MVDA on-line is however generally considered only useful for long running commercial processes with many historical batches.
We would like to present a method/an architecture which will allow a process with as few as one batch history data to apply on-line MVDA for process monitoring to bridge this gap in process monitoring. The method involves a CPV server that includes an historian, a simulation server based on random or peudo-random calculation such as Monte Carlo, a quality predictor with interface for customer design space, and a batch analyzer.
This new method/architecture would make on-line MVDA relevant to stage 2 of process validation and allow design space to be input to CPV program in drug development. The CPV server will work with on-line MVDA platform such as DeltaV Batch Analytics (BA) to provide a seamless transition for process monitoring from development to clinical manufacturing to commercial production to achieve the ultimate goal of Continued Process Verification (CPV).
Zuwei explained the role of the BioPhorum Operations Group (BPOG), a cross-industry collaboration with the aim of sharing operational best practices in the areas of drugs substance manufacturing, process development and fill finish. The BPOG has recommended on-line MVDA, such as DeltaV Batch Analytics, as a powerful extension to current CPV programs in pharmaceutical industry.
He shared that many times pharmaceutical and biotech manufacturers face the challenge of building models in Batch Analytics when there is not enough number of historical batches.
By using a new, innovative approach, the MVDA model can now be built on only one historical batch and whatever vision the manufacturer has for the design space. This novel method involves using simulated batches in the model building process. In combination with the dynamic time warp innovation , which allows batches of different time lengths to be aligned and analyzed, on-line multivariate data analysis can be more readily adopted.
Process Validation Lifecycle
Zuwei noted that DeltaV Batch Analytics is now applicable for clinical manufacturing and green field start-up facility as well as commercial productions with more than 25 batches. This means that CPV can now be achieved by working through stages 1, 2 and 3 and does not needed to be treated as a disconnected piece.
Recently, the licensing for DeltaV Batch Analytics was changed to support the deployment of up to five models while using the DeltaV Advanced Batch license. These five models can be deployed on a single unit or may be spread across up to five units.
This licensing change helps to support the trend toward online MVDA as part of a Continued Process Verification program.
If you’re attending the IFPAC Annual Meeting, make sure to catch Zuwei’s presentation. You can also connect and interact with other pharmaceutical, biotech and DeltaV experts in the Life Sciences and DeltaV groups in the Emerson Exchange 365 community.
