Evaluation of rapid test kits for quantification of deoxynivalenol in naturally contaminated oats and wheat

The aim of this study was to evaluate the performance of rapid test kits for analysis of deoxynivalenol (DON), a mycotoxin commonly found in small-grain cereals. Concentrations of DON and its acetylated derivative 3-acetyldeoxynivalenol (3-ADON) were analysed in naturally contaminated wheat (n=56) and oats (n=81) using two enzyme-linked immunosorbent assays, i.e. Ridascreen (R) DON ('Standard ELISA') and Ridascreen (R) FAST DON ('Fast ELISA'), and two lateral flow devices, i.e. Rida (R) QUICK DON ('Quick LFD') and Rosa (R) DON (Rosa LFD'). Analysis by the LC-MS/MS reference method showed a median DON + 3-ADON concentration of 1,168 mu g/kg (90th percentile 3,264 mu g/kg) in wheat and 2,250 mu g/kg (90th percentile of 9,008 mu g/kg) in oats. Rosa LFD offered the best reliability, achieving detection that was stable across toxin levels in both matrices. In addition, Rosa LED had the highest accuracy (i.e. proportion of correct results) at a DON level of 1,250 mu g/kg in wheat (0.83) and 8,000 mu g/kg in oats (0.96). Standard ELISA achieved the highest accuracy at a DON level of 1,750 mu g/kg in oats (0.85). Detection by Quick LFD in wheat and Fast ELISA in both wheat and oats differed significantly among toxin levels (P <= 0.05), with recoveries outside European Union requirements (70-120%) in some cases. The correlation with results from LC-MS/MS analysis in oats was best for Rosa LFD (R-2=0.79). In wheat, however the correlation was best for Fast ELISA (R-2=0.81). Rosa LFD was regarded as well-suited for on-site detection. Standard ELISA allowed simultaneous testing of several samples and achieved detection that was stable across toxin levels in both matrices, making it a useful alternative in laboratories. To our knowledge, this is the first report describing the performance of rapid test kits for detection of DON in naturally contaminated oats.