Inter-laboratory comparison study for pyrrolizidine alkaloids in animal feed using spiked and incurred material

Abstract

Pyrrolizidine alkaloids (PAs) are hepatotoxic metabolites produced by plants. PAs in animal feed can cause acute or chronic intoxications in animals and can be transferred to milk. An inter-laboratory comparison study among 12 laboratories, using their own methods of analysis, was conducted for the detection and quantification of PAs in animal feed. The participants were asked to quantify PAs in a blank test sample, a blank test sample to be spiked with a provided spiking mixture of seven PA standards, and a test sample contaminated with common groundsel (Senecio vulgaris). Ten of the participating laboratories used an LC-MS/MS method, one used an LC-ToF-MS method, and one used a GC-MS method. None of the laboratories reported false-negative samples, while two laboratories reported false-positive results in the blank sample. z-scores were calculated for each laboratory for seven PAs in test samples B and C. z-scores varied considerably between laboratories for the concentrations of the free bases and less for the N-oxides, probably due to the lower levels of the free bases as compared with the N-oxides in the contaminated feed. Questionable or unsatisfactory results for the z-scores were obtained for 8% of the cases for the spiked sample and for 12% of the incurred sample. Three laboratories scored consequently positive or negative results. No preferred method for quantification of PAs in feed could be identified within the methods used for this study due to the relatively small number of participants. It was concluded that this inter-laboratory study shows that the methods used for PA detection need further development for accurate estimation of PAs in contaminated feed.