From the 2005 FDA Science Forum
Testing the Test Kits: An LC/MS/MS method for penicillin in positive control milk
D. N. Heller1 , S. E. Cullison2 , 1CVM Office of Research, Laurel, Maryland, 2University of Montana, Missoula
Every tanker truckload of raw milk shipped in the United States is tested for presence of beta-lactam antibiotics before processing. The FDA Center for Veterinary Medicine, Office of Research, evaluates the rapid test kits which are used in this screening program. The test kits are required to include a negative and positive control sample for quality control purposes. The Pasteurized Milk Ordinance (PMO) of the National Conference on Interstate Milk Shippers (NCIMS) sets strict limits on the concentration of the positive control included with each test kit. The concentration of penicillin G must be 5.0 ng/mL with an allowable error of 10%, i.e., 5 ± 0.5 ng/mL. To date, regulatory agencies have not had a chemical method for verifying the concentration and stability claims for these test kit positive control samples. This method development project aimed to develop and apply such a method based on liquid chromatography-tandem mass spectrometry.
Test kit positive control samples were prepared according to label instructions, by diluting with water or raw control milk, depending on manufacturer. A standard curve was prepared by fortifying raw control milk with penicillin G at levels ranging from 1 - 20 ng/mL. Phenethecillin was used an an internal standard. The samples were extracted twice with acetonitrile. The combined extracts were evaporated and diluted with water. Extracts were further cleaned up on C-18 bonded silica solid phase extraction cartridges. Analysis was carried out on a benchtop triple quadrupole mass spectrometer. On-line liquid chromatography was performed on a short C-18 column, and data were collected using selected reaction monitoring of two product ions from penicillin and one from phenethecillin.
After testing a variety of approaches, the final form of the method was set as described above. Positive control samples from a variety of test kits were analyzed with the provisional method. There was no evidence of any serious problem needing immediate attention; all results fell within the ± 10% tolerance, or nearly so.
This important aspect of the United States' milk safety program was evaluated successfully. The provisional method appears satisfactory for its purpose. Preliminary results with the provisional method do not indicate any major problem with test kit controls. The next step will be to perform a second analyst check and complete the method validation process. At that point the final method will be reapplied to the test kit positive controls, to verify concentration claims and to test the stability of their concentrations over their useful life under recommended storage conditions.
Sarah Cullison served as an intern in the FDA's Windows to Research and Regulatory Science program during 2004. She is presently a graduate student in chemistry at the University of Montana, Missoula.