RI-ERP-FINALACTION-Recommendations

OMAMAN-20 C/ In House Validation Report ERP Use Only - March 2015

4

Summary of Results

4.1

Manufacturer’s In-house Study

4.1.1 General remarks The manufacturer’s in-house validation scheme followed the AOAC Appendix N (1), the special guidance document of the AOAC Allergen Community for quantitative gluten methods (4) and long-lasting practical experiences of the manufacturer. One important point in validating a qualitative method is to blind-code the extracted samples before measurement. This was guaranteed by splitting the analysis in two parts. (1) One technician extracted the samples, prepared the surfaces and blind-coded all extracts/surfaces. (2) Another technician performed the dip-stick analysis and documented the read-out of the results in a blinded form prepared by the first technician. Before starting the validation process, both technicians obtained a proper familiarization and training phase. In theory, one would expect that at a certain concentration of gliadin there are positive and negative results for the dip-stick outcome. In practice, this is an exception according to our experience. The normal case is that a clear individual visual cut-off exists for every technician and it is therefore labour-intensive up to impossible to find a gliadin concentration where positive and negative results occur with a comparable frequency (equivalent to POD50%). In consequence, POD response curves give –in some cases- no more information than a simple table with the positive and negative results for each matrix and concentration. In the following chapters, POD curves are presented when worthwhile. The spiking of all spiked samples is performed in the blank matrices prior to extraction. Selectivity Study: Target compounds The WGPAT gliadin material was used as a characterized reference point since its source of origin is described and the contents of gliadin and glutenin were measured independently by chromatography (3). Furthermore, a wheat, rye, and barley flour were analysed. They were supplied and characterized for their variety and their protein contents by Katharina Scherf and Peter Koehler (Deutsche Forschungsanstalt für Lebensmittelchemie, Freising, Germany). For wheat, rye, and barley the varieties were Akteur, Conduct and Marthe, respectively, while the total protein contents were 13.13%, 4.03% and 6.71%, , respectively. Flours were extracted both with ethanol (3.7.2.3) and Cocktail/80% ethanol (3.7.3). WGPAT gliadin was solved both in 60% ethanol and Cocktail/80% ethanol. The resulting solutions were further diluted with both 60% ethanol and Cocktail/80% ethanol to obtain different concentrations. The final dilution step was in sample dilution buffer as described in the test kit insert (3.7.2.5 and 3.7.3 in this document). For each target compound and extracted solution, a probability of detection (POD) curve was determined to estimate the minimal concentration which results in a positive test (five concentrations around the 95% POD; 5 replicates per concentration). At this stage no more replicates were included since only the detectability should be checked. Gliadin concentration for WGPAT gliadin was calculated according to protein content of WGPAT gliadin and dilution. Prolamin concentrations of flour extracts (ethanol and Cocktail extracted) were determined by measurement in the sandwich ELISA RIDASCREEN® Gliadin R7001 (reference method in Codex Alimentarius containing the monoclonal antibody R5 and AOAC OMA first action 2012.01). Analysis was performed as described in chapter 3. Table 1 shows the results of different target compounds extracted with 60% ethanol. 4.1.2

AOAC R search Institute Expert Revi w Panel Use Only

Page 10 of 27

RIDA®QUICK Gliadin Validation report 2015-01-14