Appropriate Limits for Microbial Contamination after PBSC Processing and the New European Pharmacopoeia Method for Microbiological Control of Cellular Products.

Aim: Patients undergoing haemopoietic stem cell transplantation are at risk of developing life-threatening infections. Peripheral Blood Stem Cell (PBSC) products represent a potential infection source and previously reported contamination rates of PBSC range from 0 – 4.5%. Our objective was to retrospectively investigate PBSC processing at the Centre for Blood Cell Therapies (CBCT) to determine its effectiveness in preventing contamination and to determine if the contamination rate limit is appropriate, as well as determining if the new European Pharmacopoeia (EP) method for microbiological control of cellular products (2.6.27) will have any bearing on these contamination rates.

Method: CBCT employs multiple systems to prevent PBSC contamination including functional closed processing steps and appropriate aseptic process validation. A 4 sample (post collection, sterile additive, pre freezing, administration line) microbial testing system with inoculation of PF BacTAlert bottles (BioMerieux) allows identification of contamination sources. Any products with measurable contamination with initial (post apheresis) negative results was deemed to have processing-associated contamination. Our unit, by agreement with our manufacturing regulator (Therapeutic Goods Administration) has imposed a process-associated contamination rate limit of 1% of all procedures. A retrospective analysis from 2001 to 2005 collated all positive microbial events and classified contamination as process-associated or derived from other sources. The EP method was used to validate PF and pairs of adult BacTAlert FA (aerobic) and FN (anaerobic), and BA(aerobic) and BN(anaerobic) bottles with a final inoculum of 1% of the product volume containing 10 CFU and 100 CFU of E. coli, P. aeruginosa, S. aureus, C. albicans, C. pseudodiphtheriticum, B. fragilis, A. fumigatus and C. sporogenes. The final dilution was in representative patient cryopreserved PBSC consented for donation for this purpose to demonstrate the suitability of the method for detecting microbial contamination in PBSC in the presence of cryoprotectant. Bottles were incubated for at least 7 days for automated detection, all detected organisms were typed, and dilute inoculums were plated on TSA plates to confirm actual CFU inoculum.

Results:1770 PBSC collections were processed over the 4-year period (May 2001–June 2005). Overall, 20/1770 (1.7%) PBSC products were contaminated, with 4/1770 (0.34%) deemed as process associated. The remainder resulted from other sources, including endogenous (11/1770, 0.94%), detected at infusion (1/1770, 0.09%) and false positives (4/1770, 0.34%). The EP method was positive for all inoculums into PF, FA and BA bottles save for B. fragilis and C. sporogenes, whilst FN was negative for all inoculums save for E. coli, S. aureus, B. fragilis and C. sporogenes and BN was similarly negative to FN for all inoculums with the addition of P. aeruginosa.

Conclusion: We have demonstrated that whilst there is a process-associated contamination rate of 0.34%, this is within our regulator agreed limit of 1%. We have also demonstrated that a routine BacTAlert (FA and FN) detection method with 1% of PBSC will readily detect 10 CFU with no discernable inhibitory effects according to a pharmacopoeia method.