In 2011, the FDA issued Process Validation: General Principles and Practices, which proposed a lifecycle model for process validation that emphasized the use of a science and risk-based approach for validation.[1] Although this model was originally proposed for process validation, regulatory agencies have confirmed it is equally applicable to cleaning processes and cleaning validation. This lifecycle model when adapted for cleaning validation breaks it down into three stages: Stage 1 - cycle development, Stage 2 - cleaning validation, and Stage 3 – continued cleaning process verification. The purpose of stage 3 - continued cleaning process verification, or periodic monitoring as it will be referred to throughout this memo, is to provide the continued assurance that the cleaning process remains in a state of control through the collection and statistical evaluation of cleaning process performance data.
Historically, it was considered acceptable to monitor the validated state of cleaning through periodic testing of the cleaning process. This was done by performing a periodic revalidation of the cleaning process where one or more validation runs were repeated to ensure the process was in a state of control. Unfortunately, this linear approach to cleaning validation maintenance did not provide a robust picture of the cleaning processes’ state of control, as it solely relied on the success or failure of the periodic confirmatory validation runs. Ultimately, this approach was unable to identify drift from validated parameters which often resulted in programs that fell out of compliance over time. The concept of periodic revalidation has since been removed from regulatory guidances based on the rationale that ongoing activities implemented as part of periodic monitoring are more relevant for determining the state of control for a cleaning process.
Periodic monitoring addresses many of the shortcomings of periodic revalidation, as it provides greater visibility of the cleaning processes’ performance over time, and a logical path for the early detection, investigation, and remediation of process drift, as well as the identification of potential process improvements. So, let’s discuss what a robust periodic monitoring program looks like post-cleaning validation, and why regulatory agencies prefer the lifecycle approach. At its core, this is a program that establishes a system to detect unplanned departures from the validated cleaning process through the continuous collection and analysis of cleaning process performance data that is indicative of the processes’ state of control.
To adequately monitor the controlled state during stage 3, it’s necessary to monitor the analytical data, process data and any deviations stemming from the cleaning process. Analytical data should be collected and analyzed from visual inspection, swab, and rinse samples. To do this, a risk-based sampling plan should be established to collect this cleaning data with consideration for equipment’s design, accessibility, material of construction, and sampling method(s) to be used. Sampling locations chosen should be based on risk with respect to difficulty of cleaning and likelihood of contamination or carryover into the next product. [5] This plan may imitate the one executed in the cleaning validation protocol, or it may be reduced if there is data-supported justification and statistical demonstration of consistently passing results. Samples should be analyzed for the detection of relevant risks such as process residues, cleaning agent and microbiological concerns. If analytical testing of swab and/or rinse samples is already performed after every product changeover cleaning, the product changeover cleaning data may be used in support of monitoring. It is important to note that if a facility manufactures products that have potential to develop nitrosamines or other genotoxic impurities of concern on multi-product equipment, the risk to other products manufactured on the same equipment line must be monitored at product changeover, and the data should be evaluated during periodic monitoring. Additionally, data for the validated critical process parameters that are measured and maintained during execution of cleaning should be gathered and reviewed as they demonstrate performance of the cleaning process for each cleaning run.[7] For automated cleaning processes, validated critical process parameter trend data should be reviewed in tandem with any instances of process alarms, which provide a strong indication of the cleaning cycle’s and the automation’s performance. Lastly, gather and review any cleaning failures, deviations related to cleaning, or change controls that have occurred since the last routine monitoring or the cleaning validation. [7] This is the minimum recommended data that should be evaluated during periodic monitoring as it is a holistic representation of the cleaning processes’ performance over time.
The cleaning process performance data, previously mentioned, should be trended, and compared against the previous monitoring or cleaning validation results, as applicable, for any indication of process drift. Trending of these data sets allows for the identification and investigation of cleaning cycle issues before they result in a failure requiring revalidation. [7] For example, in Figure 1, Trended TOC Rinse Monitoring Results shows the trending data of TOC rinse samples collected to monitor a cleaning process. As you can see over time there has been a gradual increase in residue levels observed in the TOC measurements, this is indicative of process drift and should be investigated to determine the source and required corrective actions before a critical cleaning failure occurs.
Figure 1: Trended TOC Rinse Monitoring Results
The periodic monitoring plan should be executed routinely following completion of cleaning validation. For both automated and manual cleaning processes, a risk-based assessment must be used to justify the frequency that monitoring will be performed for both product contacting and non-product contacting equipment. For automated processes, the frequency of monitoring may be reduced as process knowledge and cycle performance data is gathered and trended, provided there is data-supported justification that the process is statistically unlikely to drift out of the controlled state. However, for manual cleaning processes, monitoring must continue to be executed at least annually even where there is statistical data demonstrating that the process is operating in a controlled state due to the inherent variability of manual cleaning.
Periodic monitoring is essential for maintaining the validated state of cleaning processes, ensuring regulatory compliance, and continuously improving cleaning processes to ensure product quality and patient safety.
References –
1. FDA. (2011). Process Validation: General Principles and Practices.
2. ICH. (2016) Q7. Good Manufacturing Practices for Active Pharmaceutical Ingredients.
3. PIC/S. (2007). PI 006-3. Recommendations on Validation Master Plan Installation and Operational Qualification Non-Sterile Process Validation Cleaning Validation.
4. Health Canada. (2021) GUI 0028_EN: Cleaning Validation Guide.
5. ASTM. (2018). E3106. Standard Guide for Science-Based and Risk-Based Cleaning Process Development and Validation.
6. ISPE. (2020). Guide: Cleaning Validation Lifecycle – Applications, Methods, and Controls.
7. PDA. (2012). Technical Report No. 29 (Revised 2012). Points to Consider for Cleaning Validation.
Contributors: Joanna Joseph, Jenna Carlson, and Alec Fufidio
Comments