Critical evaluation of LC-MS-based methods for simultaneous determination of deoxynivalenol, ochratoxin A, zearalenone, aflatoxins, fumonisins and T-2/HT-2 toxins in maize

The results of a proficiency test for the LC-MS/(MS) determination of up to 11 mycotoxins (aflatoxins B-1, B-2, G(1) and G(2), fumonisins B-1 and B-2, ochratoxin A, deoxynivalenol, T-2 and HT-2 toxins and zearalenone) in maize were evaluated to identify possible strengths and weaknesses of various methodologies used by the 41 participating laboratories. The majority of laboratories (56%) used mixtures of acetonitrile:water for extraction. Other laboratories used methanol:water mixtures (17%) or performed two consecutive extractions with phosphate buffer solution (PBS) followed by methanol (15%). Few laboratories used mixtures of acetonitrile:water:methanol (7%), water:ethyl acetate (2.5%) or PBS alone (2.5%). The majority of laboratories (58%) used a clean-up step prior to chromatography. The remaining laboratories analysed crude extracts (37%) or used a mixed approach (5%). The amount of sample equivalent injected into LC-MS/(MS) ranged between 0.1-303 mg for purified extracts and 0.08-20 mg for directly analysed crude extracts. External (54%), matrix-matched (22%) or stable isotope-labelled internal standards calibration (24%) were used for toxin quantification. In general, extraction mixtures of water with acetonitrile, methanol or both provided good results for quantitative extraction of mycotoxins from maize. Laboratories using sample extract clean-up reported acceptable results for the majority of mycotoxins. Good results were also obtained by laboratories that analysed crude extracts although a high variability of results was observed for all tested mycotoxins. Matrix-matched calibration or isotope-labelled internal standards efficiently compensated matrix effects whereas external calibration gave reliable results by injecting <= 10 mg of matrix equivalent amounts. Unacceptable high recovery and high variability of fumonisin results were obtained by the majority of laboratories, which could not be explained and thus require further investigation. These findings provide the basis for the optimization and selection of methods to be used in future interlaboratory validation studies to derive their performance characteristics for simultaneous determination of mycotoxins in maize.