Realization of an Effective Ongoing/Continued Process Verification System
Backgroud
Tiofarma B.V. is a family-owned pharmaceutical company with more than 30 years’ experience and a diverse product portfolio. Together with their partners, they develop and manufacture medicines for the EU and the US. Recently, Tiofarma approached Progress to help develop and implement an Ongoing Process Verification (OPV) system for their manufacturing facilities. The term OPV is used by the EMA, whereas the US FDA speaks of a Continued Process Verification (CPV) system. As a senior consultant, I was asked to lead the project, bringing both my statistical background and the ability to relate to different disciplines.
CPV guidelines
With the publication of the FDA Process Validation Guideline in 2011 and the revision of Eudralex Annex 15 in 2015, OPV/CPV has been established as the third stage in the validation lifecycle of pharmaceutical products. Since then, there has been a lot of attention, including publications from, for example, ISPE, PDA or BioPhorum on this subject. Meanwhile, such a statistical process control system is increasingly demanded by authorities and industrial partners.
Team effort
Setting up an OPV/CPV system requires a thorough understanding of the statistics behind the process control system. At the same time, a thorough understanding of the product, the manufacturing process and the analytics are required. Performing OPV/CPV must therefore be a team effort and there needs to be mutual understanding within the team. To address the various challenges of setting up a OPV/CPV system, a team was assembled within Tiofarma with the required disciplines, including R&D, QA, QC or Analytical Development, manufacturing experts and a data analyst.
The challenges
Performing OPV/CPV includes data acquisition, data analysis and visualization, reporting and responding to the results. Procedures and tools have to be established for each of these steps.
Data acquisition must be reliable, but preferably with limited effort. Digital data sources are preferred, but availability depends on the ‘digital maturity’ of the company, in ISPE terminology.
The implementation of OPV/CPV for a specific product requires selection of the most informative variables to be trended, which includes Critical Quality Attributes (CQA, QC results of the release testing), but also In-Process Controls (IPC), Critical Material Attributes (CMA) and (Critical) Process Parameters ((C)PP). Monitoring insignificant variables consumes precious time and may distract from the important data. Which variables to select may depend on the data itself, for example by omitting categorical data or variables with excellent process capabilities. But of course it also depends on the potential impact of variables on the final drug product quality (criticality), which requires a thorough understanding of the product and the process. In addition, the type of control chart and statistical tests to be used for each variable must be determined. This may again depend on the data, for example whether the data follow the normal distribution well. Procedures had to be established for rational selection of variables to be trended and statistical analyzes.
Statistical software had to be implemented for data visualization and analysis. Furthermore, procedures had to be put in place for responding to Out Of Trend (OOT) results flagged by the system. Data does not lie, but the observed variation is caused by many different factors related to the product, raw materials, manufacturing process and analytics, and only a team of experts can interpret the observed trends in a useful way.
Results
We improved the availability of digital data sources to which Tiofarma’s data analyst made a major contribution. Statistical tools were set up and related quality documentation provided, including training of key personnel. Together with the highly skilled Tiofarma team, selections of the most informative variables, the type of control charts and the statistical tests were made in a structured, rational and practical way. We also included how the team can effectively assess observed trends in the system.
Conclusion
With the implementation of the OPV/CPV system, based on the experience and shared knowledge of previous similar projects performed by Progress, a powerful tool was delivered. This allows the capable Tiofarma team to not only provide the intended ‘assurance to remain in a state of control’, but also improve insight into the process and provide objectives for further optimization.
