Sterility testing is an essential part of in-process and release testing for cellular therapy products. The FDA specifically recommends that sterility testing be performed as outlined in 21 CFR 610.12. Furthermore, because they are concerned that antibiotics may interfere with the accurate assessment of sterility testing, the FDA requires preliminary bacteriostasis and fungistasis testing according to the USP """"""""<71> Sterility Test"""""""" on all samples containing antibiotics, including cells grown in antibiotic-containing media. The methods for sterility testing described in the CFR and USP standards were developed more than 25 years ago, and are labor-intensive and require incubation for 14 days. Since the time these methods were published, more sensitive and rapid methods have been developed for detection of microbial growth in various body fluids. These modern methods include automated systems for the detection of microbial growth in blood and other normally sterile fluids. Despite the fact many laboratories use these automated systems to assess sterility of cellular therapy products, the FDA has not sanctioned this application because there are no published data from any formal comparison of these newer methods with the CFR and USP methods. For these reasons, we have developed a validation protocol comparing the CFR and USP methods with two automated culture methods: the bioMerieux BacT/Alert system and the Becton Dickinson Bactec system. Cell products will be seeded with selected bacteria and fungi and then tested with each method. The test sensitivity and time to detect a positive culture will be assessed. The validation protocol has been submitted to the FDA for their comments and we anticipate this study will be initiated in the Fall, 2002. As supporting data for this FDA submission, we have initiated a preliminary comparison of the CFR and USP methods with the automated bioMerieux BacT/Alert system (currently used in the NIH Microbiology Laboratory). All cellular therapy products (N=187) that have been prepared during August-September were tested in both systems. A total of 7 products had growth detected in one or both systems: 3 products in the BacT/Alert system only, 2 products in the CFR/USP method only, and 2 products in both systems. The clinical significance of these positive cultures will be assessed. This preliminary study will also be used to define (1) the time to detection of a positive culture in the CFR/USP and BacT/Alert methods; and (2) the incidence of false-positive culture signals (e.g., the system appears positive [cloudy; positive signal by the automated system] but no organisms are recovered in the system).
|Khuu, H M; Stock, F; McGann, M et al. (2004) Comparison of automated culture systems with a CFR/USP-compliant method for sterility testing of cell-therapy products. Cytotherapy 6:183-95|