Validation of sterility testing of cord blood: challenges and results.

ABSTRACT

BACKGROUND: Sterility testing for cord blood (CB) products is mandatory to prevent transplantation-transmitted microbial infections. Here, the automated BacT/ALERT (bioMérieux) culture system was validated to detect microbial contamination in CB units processed at the Canadian National Public Cord Blood Bank. STUDY DESIGN AND METHODS: A three-phase validation was developed. CB units were prepared with pentastarch (Phases 1 and 2) or hetastarch (Phase 3). In Phase 1, CB was spiked with approximately 100 colony-forming units/mL of Pseudomonas aeruginosa, Klebsiella pneumoniae, Staphylococcus aureus, Staphylococcus epidermidis, Bacteroides fragilis, and Candida albicans. Plasma (8 mL) and buffy coat (BC; 0.5 and 8 mL) were inoculated into culture bottles. In Phases 2 and 3, a mix of red blood cells (RBCs) and plasma (4 mL each) was used as the inoculant. In Phase 3, Aspergillus brasiliensis was added as a test organism and microbial concentrations in the by-product RBCs and plasma were determined. The BC fractions were cryopreserved and tested 3 months later. RESULTS: In Phase 1, bacteria failed to grow in CB units containing antibiotics. Thus, antibiotic-free units were used for the other phases. C. albicans was not always captured in plasma, but using a mix of RBCs and plasma, all organisms were detected. The use of pentastarch or hetastarch did not affect microbial recovery. C. albicans and A. brasiliensis were preferentially recovered in RBCs and BC. Cryopreservation did not affect microbial survival during CB processing. CONCLUSIONS: A mix of plasma and RBCs is appropriate for CB sterility testing. Interestingly, fungi preferentially segregate to cellular fractions. The clinical significance of the bactericidal /or bacteriostatic effect of antibiotics in CB merits further investigation.