Data Mining to Determine the Influence of Fluid Properties on the Integrity Test Values

Area of Science:

• Pharmaceutical Manufacturing
• Filtration Technology
• Microbiology

Background:

• European Union Good Manufacturing Practice (GMP) mandates pre-use filter integrity testing (PUPSIT) for sterile products.
• The rationale for PUPSIT is to detect filter damage that may occur post-sterilization or during use, which could otherwise be masked.
• Potential filter plugging by particulates or microorganisms could alter integrity test results, potentially masking defects.

Purpose of the Study:

• To evaluate the impact of bacterial challenge testing (BCT) on filter bubble point (BP) to assess the risk of masking filter defects.
• To determine if bacterial loading during BCT can cause a filter defect to be missed by post-use integrity testing.
• To analyze BP changes in relation to filter specifications and identify fluids posing a masking risk.

Main Methods:

• Collected and compared pre- and post-BCT bubble point (BP) data for 2086 filters across 531 BCTs using diverse pharmaceutical fluids.
• Calculated post-test to pre-test BP ratios for each filter and analyzed their distribution.
• Compared observed BP ratios against a defined risk criterion (ratio of 0.2 μm to 0.45 μm filter BP specifications).

Main Results:

• The average post-test to pre-test BP ratio was 1.00 ± 0.15, indicating no significant average elevation of BP due to bacterial loading.
• Only 1.5% of tested fluids (8 out of 518) exhibited BP ratios meeting the criterion for a potential masking risk.
• The study identified specific fluids and potential risk factors associated with these rare cases of masking risk.

Conclusions:

• Filtration processes that significantly alter BP and pose a risk of masking defects are uncommon.
• Routine bacterial challenge tests can detect potential masking risks.
• The observed BP ratios during BCT provide valuable data for assessing masking potential and informing decisions on PUPSIT implementation.