Troubling statement by the Prolamin Working Group on the determination of gluten in fermented foods containing partially hydrolyzed glutenTricia Thompson
Gluten Free Watchdog is very troubled by a statement recently released by the Prolamin Working Group “on the determination of gluten in fermented foods containing partially hydrolyzed gluten.” While we have a great deal of respect for each of the authors, the insinuation that the FDA did not lead with science in determining that “the R5 Competitive ELISA method is not suitable for the detection and quantification of gluten in any fermented or hydrolyzed food” comes across as exceedingly dismissive of the work done by FDA in writing this rule. The PWG should be very familiar with the work of the late Eric Garber, former research scientist at FDA. His scientific thinking is all over the FDA rule. Great scientific minds can disagree and whether the competitive ELISA is fit for purpose is obviously open to debate. BUT completely discounting the FDA’s well thought out scientific approach to this ELISA comes across as a knee jerk reaction that aims to defend industry, and in particular the beer industry.
The PWG statement can be read in full at: https://www.frontiersin.org/articles/10.3389/fnut.2020.626712/full.
The FDA’s scientific explanation of their view of the competitive ELISA is available at https://www.federalregister.gov/documents/2020/08/13/2020-17088/food-labeling-gluten-free-labeling-of-fermented-or-hydrolyzed-foods (scroll down to section 2. INNOVATION IN DEVELOPING METHODS FOR FERMENTED, HYDROLYZED, OR DISTILLED FOODS).
Portion of the “Statement of the Prolamin Working Group on the determination of gluten in fermented foods containing partially hydrolyzed gluten“:
“On August 12, 2020, the U.S. Food and Drug Administration (FDA) has finalized a rule related to gluten-free labelling for foods containing fermented, hydrolyzed ingredients. The FDA believes that there is no scientifically valid analytical method effective for determining gluten in fermented or hydrolyzed foods. In the absence of an analytical method, the FDA has decided to evaluate gluten-free claims on these foods based only on evidence that the food or ingredient used is gluten-free before fermentation or hydrolysis. For example, barley-based beers from which gluten is removed during brewing using special filtration, adsorption and/or enzymatic treatment are therefore excluded from bearing a gluten-free label.
The Prolamin Working Group (PWG) acknowledges that the FDA rule is a regulatory act and might have to take into consideration several aspects other than scientific evidence, including risk assessment. Nevertheless, the PWG thinks that science has to be the most important driver for regulatory acts in risk management.”
Portion of FDA’s Federal Register Notice “Food Labeling; Gluten-Free Labeling of Fermented or Hydrolyzed Foods“:
2. INNOVATION IN DEVELOPING METHODS FOR FERMENTED, HYDROLYZED, OR DISTILLED FOODS
(Comment 4) A few comments stated that a valid method exists to quantify gluten in a product that has been fermented or hydrolyzed, like beer, and pointed to the R5 Competitive ELISA test with inactivated protease enzyme.
(Response 4) When compliance with § 101.91(b) is based on an analysis of the food, FDA will use a scientifically valid method that can reliably detect the presence of 20 ppm gluten in a variety of food matrices, including both raw, cooked, or baked products (§ 101.191(c)). As stated in the 2011 notice and the 2013 gluten-free food labeling final rule, a scientifically valid method for purposes of substantiating a “gluten-free” claim for food matrices where formally validated methods (e.g., that underwent a multi-laboratory performance evaluation) do not exist is one that is accurate, precise, and specific for its intended purpose and where the results of the method evaluation are published in the peer-reviewed scientific literature. In other words, a scientifically valid test is one that consistently and reliably does what it is intended to do (76 FR 46671 at 46673; 78 FR 47154 at 47165). The R5 Competitive ELISA test has potential as a quantitative method, and we acknowledge that, under the appropriate test conditions, the R5 Competitive ELISA can generate reproducible results. The commercial R5 Competitive ELISA marketed for the detection of hydrolyzed (or fermented) gluten has, by design, an advantage over sandwich ELISA-based methods by not requiring the presence of two antigenic epitopes (antibody binding sites) to detect the presence of gluten peptides. Further, because the immunopathogenesis associated with celiac disease only requires a single immunopathogenic element, the R5 Competitive ELISA is theoretically more appropriate as an assay.
However, as currently designed, the R5 Competitive ELISA method is not suitable for the detection and quantification of gluten in any fermented or hydrolyzed food (e.g., beer, yogurt). The lack of appropriate reference standards for the detection and quantification of gluten subjected to fermentation or proteolysis (hydrolysis) makes the results generated by the R5 Competitive ELISA difficult, if not impossible, to interpret. As currently supplied, the calibration standard in the R5 Competitive ELISA is allowed to proceed for a specified amount of time at a specific temperature. If the hydrolytic conditions (time, temperature, or composition under which the hydrolysis is occurring) associated with the production of the sample being analyzed were different from those used to make the calibration standards, the peptide profile is likely to be different, and the assay is unlikely to generate accurate results. The Association of Official Analytic Chemists Official Methods of Analysis (AOAC OMA) First Action award to the R5 Competitive ELISA stated that the hydrolyzed gluten being used as a calibration standard may not be suitable, and users should establish their own standards before relying on the calibration standard (Ref. 3). Specifically, minor fluctuations in temperature and time, as well as the specifics of the proteolysis, could result in a different range of peptides, making the calibration standards not suitable.
Further, it is not known how to interpret the immunopathogenicity based on the amount and profile of gluten peptides detected. The threshold of 20 ppm gluten was based on studies examining the immunopathogenicity of intact gluten. Whether the biological activity on a per mg basis is the same for gluten peptides, as was measured with intact proteins, is unknown; the answer may depend on the peptide profile.
Thus, we have concerns regarding the use of the R5 Competitive ELISA in the detection of gluten in fermented or hydrolyzed foods or ingredients because of the challenge in demonstrating that it is suitable for the intended purpose of interpreting the immunopathogenicity based on the amount and profile of gluten peptides detected and whether the method performs reliably (i.e., is a scientifically valid method). While the method may perform reproducibly as indicated by the American Association of Cereal Chemist International (AACCI) validation (Ref. 4), it does not mean that the method is suitable for the intended purpose of detecting and quantifying, with sufficient accuracy, the gluten protein content in fermented and hydrolyzed foods, or assessing the immunopahogenicity or equivalent amount of intact gluten proteins.
Finally, the procedure of adding a controlled amount of an artificially prepared hydrolysate to food as required by the testing protocol (a process called “spiking”) may give an inaccurate reading because it does not reflect the assay’s ability to detect gluten that has been added to the food before processing and hydrolyzed during production. For this reason, it is Start Printed Page 49247important that, whenever possible, methods be validated using gluten that is added to the food before processing. The inability to detect any gluten using the R5 Competitive ELISA (below the limit of detection) is not an indication of complete elimination or even a reduction of gluten. Another complexity is that not all the immunopathogenic sequences of gluten have been identified. Further, the R5 antibody does not recognize all immunopathogenic sequences (e.g., glutenin-derived) and, therefore, gluten could be present in a form that is not detectable (Ref. 5).