PUPSIT (Part 2 of 2)

My opinion on PUPSIT

My first work experience involved addressing deviations related to an aseptic RABS filling line equipped with a fixed, stainless-steel PUPSIT system (non-redundant). PUPSIT was preferred because it did not use redundant filtration and reused the sterilization filters, which meant the filters had to undergo multiple sterilization cycles in an autoclave that could potentially damage them. During those two years, I rarely encountered a failed inline pre-use test. I recall one failed test caused by an operator who had dropped the filter without reporting it, resulting in severe damage to the filter or filter housing. Defects of this kind would never result in the masking of defects and a successful post-use test. On the other hand, I encountered several deviations related to the incorrect opening, closing, or handling of various valves during the pre-use test. This could result in the humidification buffer entering the product’s fluid path, and thus potentially diluting the first volume, if this is not noticed or reported by the operator. Other deviations related to a potential sterility risk caused by failed filter tests following the use of protective vent filters or by the insufficient closure of a valve leading to an open waste container containing wetting fluid in the Grade B area. These deviations clearly illustrate the drawbacks associated with the complexity of PUPSIT systems, and this was not even a redundant system. It should be noted, however, that the valve design was suboptimal due to the relative proximity of the valves—a problem that can be easily remedied by modern automated systems.

In May 2019, I transitioned to QA consulting and worked with various business units of the same aseptic manufacturing client. This client routinely uses redundant single-use filtration systems for nearly all of its isolator filling lines. Redundant filtration itself also reduces the need for PUPSIT, as the risk of a defect in the second filter being masked by particles, or of the filter becoming contaminated during filtration, is significantly reduced by the protective effect of the first filter (provided the first filter is intact). To this day, I have not encountered any filter-related nonconformities resulting from improper handling of the valves during, for example, bioburden sampling. This is likely due to the reduced number of valves and operator actions. However, since this was a single-use system that had been fully assembled by the filter supplier, they encountered a supplier issue due to poor tubing/valve connections that caused leaks. Another challenge they face is the need for an adequate risk assessment for their non-PUPSIT system to comply with the new update to Annex 1, as mentioned above.

Based on my personal experience and data from various studies, I do not fully agree with the update to Annex 1, and I strongly believe that PUPSIT should not be implemented as a standard in aseptic pharmaceutical manufacturing due to its complexity and additional risks. However, I also understand that it is necessary to address certain processes or products with an increased risk of masking defects. That is why it is important to understand the filtration process and the risk of masking defects. My personal advice to companies regarding the Annex 1 update and PUPSIT? Do not rush PUPSIT and its implementation! It is important to take sufficient time to understand the risks involved in the design and use of PUPSIT systems. First, determine the risk that a filter will be damaged and that such damage will subsequently be masked, so that you can identify the steps that can be taken to reduce that risk and weigh the residual risk against the risks associated with PUPSIT.

References:

1. Volume 4 EU Guidelines on Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use. Annex 1: Manufacture of Sterile Medicinal Products, 2008
2. Volume 4 EU Guidelines on Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use. Annex 1: Manufacture of Sterile Medicinal Products, 2022
3. Ferrante, S., et al. “Test Process and Results of Potential Masking of Sterilizing-Grade Filters.” PDA Journal of Pharmaceutical Science and Technology, Accepted Article (Published online May 28, 2020)
4. Ferrante, S., et al. “Test Process and Results of Potential Masking of Sterilizing-Grade Filters.” PDA Journal of Pharmaceutical Science and Technology, Accepted Article (Published online May 28, 2020)
5. Parental Drug Association website, accessed on December 21, 2022: https://www.pda.org/pda-letter-portal/home/full-article/the-use-of-scientific-data-to-assess-and-control-risks-associated-with-sterilizing-filtration