Evaluation of rapid test kits for deoxynivalenol (DON)

Abstract

EU limits for Fusarium mycotoxins in grain will come into force on July 1st 2006. The analytical reference method for deoxynivalenol (DON) is time consuming and expensive.  Therefore, there is a need for rapid DON detection kits that involve a simpler test protocol yet still provide analytical results upon which decisions can be made. Such test kits have the potential for use at intake points within the grain chain to assure safety of supply into the human or animal feed markets.

This project was initiated to evaluate the suitability of commercially available test kits to screen intake samples for DON and to provide reliable quantitative data rapidly.  Specifically, the project set out to provide the cereal processing chain with information on the appropriateness of kits for use in intake situations and, thus, to help cereal processors to meet the requirements of impending legislation.

A set of criteria, against which DON test kits were evaluated, was generated within the project in consultation with the milling industry.  These criteria included test sensitivity and reliability, specificity for DON, cost of analysis (including the cost of ancillary equipment) plus speed and ease-of-use of the assay.

Thus, each kit was assessed in terms of its practical potential for use as a screening tool at specific points in the cereal chain rather than in terms of a complete, statistically robust validation.  Two specific applications within the cereal chain were identified by industry. These were: (i) grain intake where a sample turnaround of < 30 minutes is required and (ii) storage or merchanting facilities where batch analysis prior to delivery is more relevant.

Results from the basic evaluation showed that all test kits were capable of detecting DON in ground wheat samples, i.e. were "fit for purpose," and could be used to screen intake wheat for DON levels.

In order to improve the quality of decision made using any format of rapid test kit, it is recommended that all tests are duplicated.  In addition, for quality control purposes a suitable standard (a pure DON chemical or a wheat based check sample of known DON level close to the agreed threshold) should be tested alongside unknown samples to provide a means of monitoring kit and operator performance.

Mycotoxin distribution is not uniform within a sample and therefore careful sampling is essential for reliable DON results.

The best overall performance across the DON concentration range measured was achieved with the microtiterplate format and fully quantitative test kits. These are higher throughput techniques that take longer to produce results and require greater capital investment in instrumentation to measure colour changes and interpret these in terms of DON concentration.  A threshold at between 800 and 1000 parts per billion (ppb) would help to ensure that samples with unacceptably high DON levels (>1250ppb) are screened out of the cereal chain but would also result in rejection of some samples within the limit.

Within this study, semi-quantitative microtiterplate assays were not favoured as they did not meet the speed or ease-of-use requirements of a grain intake situation nor the sensitivity required for either grain intake or storage/merchant application.

Lateral flow devices proved to be very simple to use and require minimal laboratory equipment or technical experience. In order to introduce objectivity into the assessment the inclusion of a low cost reader is a significant advantage. Such kits are particularly suited to rapid sample turnaround situations (e.g. grain receipt point, where they can be used to screen incoming wheat against an agreed threshold), or as a screening tool in the assessment of the extent of immediate risk posed by DON.

Problems or disputes would subsequently be confirmed using appropriate fully quantitative measurement. Used as a stand-alone test, an adequate margin of safety would be to operate at one level below that closest to the legal limit.  However, such strategy would result in rejection of samples below the limit.  This may be compounded by the tendency of this format of test kit to overestimate DON concentration when only a single standard is used.