FDA’s Hostile Take-Over of GRAS, Part II: FDA’s Assault on Legitimacy of GRAS Expert Panels

March 6, 2018 - 13 minutes read

FDA's Assault on Legitimacy of GRAS Expert Panel

The week of November 13, 2017, is notable because FDA released two documents relating to guidance for conclusions of GRAS1 status and draft guidance for conduct of GRAS Expert Panels.2 Both of these documents revise and expand on “Substances Generally Recognized as Safe: Final Rule” (81 FR 54869, August 17, 2016). In December, our newsletter discussed the guideline for Regulatory Framework,3 this month we will discuss draft guidance for the conduct of GRAS Expert Panels (GEP), focusing on the avoidance of the subjective notion of a Conflict of Interest (COI) or even more subjective, the appearance of a COI.

The notion of a COI was raised prominently in the General Accountability Office (GAO) report on GRAS, released March 5, 2010, and titled FDA Should Strengthen Its Oversight of Food Ingredients Determined to Be Generally Recognized as Safe (GRAS).4 FDA has used the GAO report to consolidate FDA’s take-over of the GRAS process — enhancing the probability that FDA will become the final arbiter of what is GRAS. This draft guidance lays out the means by which nearly any GRAS could be declared invalid; not just those submitted for notification, but all GRAS determinations, past and future as the result of a provision in the Food Safety Modernization Act (FSMA).5

Probably the most emphasized issue in the GAO and PEW reports6 is the issue of bias in terms of COI; that is, that the temptation for an GEP to wave through a substance in which a Panelist might have a personal, career or financial interest. The solution proposed by FDA is simply for the Panelist to declare his COI and be banned from the GEP. However, if the conflicted Panelist’s expertise on the particular subject exceeds the probability of his potential to mislead the GEP (after all, he has already “confessed” his lack of total allegiance to scientific fact), he can be granted a waiver and allowed to provide information, but have no vote in the final outcome. Easy enough? No. In our experience at Burdock Group, the most difficult task is to obtain GEP participants that not only fit the description of an expert as described in the Food Drug and Cosmetic Act §201(s); that is, experts qualified by scientific training and experience to evaluate its [i.e., the ingredient’s] safety but, importantly, can make a decision on the data and information available, avoiding the temptation to always want to “see more data.” In our estimation, the GEP candidates offered an invitation by Burdock Group are honorable men and women, who if conflicted, will decline the invitation to become a member of the GEP.

Beyond selecting honorable people, what qualifies an Expert Panelist candidate? In their publication of 2004, Burdock and Carabin7 wrote:

The statute does not … define the experts other than those “qualified by scientific training and experience”, however, the regulations (21 CFR § 170.3(i)), in the definition of safety, refers to competent scientists. A latter part of this section, 21 CFR § 170.3(i)(3), makes reference to … experts qualified by scientific training and experience to evaluate the safety of food and food ingredients … , but only in the context of determining appropriate safety factors. Therefore, it would seem that although experts must be qualified or competent, expertise in the safety of food is only required in the application of safety factors, not in an overall judgment of safety. Therefore, a toxicologist not familiar with food ingredient safety, could still render an opinion on general safety based on his or her interpretation of the available toxicity data. Likewise, a veterinarian or pathologist could interpret animal data, a physician might be called upon to contribute an opinion on the effect a substance may have on human health and a chemist might opine on the possibility of contaminants formed or excluded during the synthesis or isolation of a product (Hallagan and Hall, 19958; Woods and Doull, 19919). Above all, the expert should be a disinterested party. A company employee may be an expert, but may also have some “emotional equity” or business/career interests that might be perceived as a potential conflict and would therefore not constitute a totally disinterested party appropriate for decision-making.

However, reading the draft guideline and FDA’s comments on advisory panels, gives rise to the thought that GRAS food ingredient reviews of the future may require competence in areas beyond a simple safety assessment, such as special diviners of safety in very specialized areas and possibly require consumer representatives to provide moral and ethical anchorage to scientific aims. Such new ambitions, as hinted at in the draft guidance, could well produce a decision matrix that looks something like the figure below.

GRAS Decision Matrix

In this matrix, the X-axis represents the increasing degree of relative scientific expertise: from relatively little or no scientific expertise, to scientific expertise specific to the substance at hand (and represented by a scientist from the company that has developed the substance or a scientist that is not necessarily an employee of the developing company, but certainly someone who has published research in this area). On the Y-axis, the degree of bias is expressed: from relatively little conscious bias, to bias in the sense that a conflict of interest would be easily discernable, such as someone in the employ of the developing company. Translational scientific expertise are those best described as having an understanding of scientific issues, but lack knowledge in safety or the specific issue at hand.

The first group, the “average” citizen, could have a range of biases from normal human bias to those with nearly grudge-like mistrust of industry, but all having little scientific expertise. If required to add a consumer representative to an expert panel, it would likely not be too difficult to find participants as self-identified “consumer representatives” tend to self-select to serve on FDA advisory committees. Also ineligible are professionals, with or without a conflict or appearance of conflict but unqualified, such as a psychiatrist or engineer assessing the safety of a food ingredient. Further, neither of these groups would fit the statutory requirement of “experts qualified by scientific training and experience”.

Progressing on the axis of scientific expertise from actual (e.g., a toxicologist with expertise in food ingredients) to someone with specific scientific expertise (e.g., one who has published in the area under question). At the lower right, a qualified professional with specific scientific expertise but absolutely no conflict, could not exist — as according to the dogma of COI, no one is like Caesar’s wife and above all suspicion.

Going to the top right of the chart, there are scientists with specific expertise, but theoretically so convinced of the concept’s infallibility or as a stakeholder in the developing company, they could not be depended upon to provide candid information.

Who is left? Two groups: the first is the professional with the appearance (or potential) of conflict. This scientist represents the sole resource of expertise available to the GEP, but because of the appearance (or potential) conflict, a waiver for participation would be required. However, would this person actually want to participate, with the potential of a stigmatizing waiver hanging over his or her head as one whose fealty to the truth has been questioned? The other group — government officials, not necessarily FDA alone but likely including officials at the National Institutes of Health — probably unbiased and knowledgeable scientists of the same group that can confer Significant Scientific Agreement (SSA) for health claims.10

What is the outlook then, for a traditional conclusion of GRAS status — in a word — bleak. If all GRAS conclusions can be picked apart on the basis of subjective notions of unqualified participants on the basis of inadequate expertise or others with specialized knowledge and therefore, the presumed bias that disqualifies them, while remaining without COI, then who remains to satisfy FDA determination of having met both the threshold of knowledge and honesty?

In the words of Thomas P. Stossel:11
The case underlying the COI movement is a mixture of moralistic bullying, opinion unsupported by empiric evidence, speculation, simplistic and distorted interpretations of complicated and nuanced information, superficially and incompletely framed anecdotes, inappropriately extrapolated or irrelevant psychological research results, and emotionally laden human interest stories.


  1. Regulatory Framework for Substances Intended for Use in Human Food or Animal Food on the Basis of the Generally Recognized as Safe (GRAS) Provision of the Federal Food, Drug, and Cosmetic Act: Guidance for Industry
  2. Best Practices for Convening a GRAS Panel: Guidance for Industry.  Site last visited Feb. 19, 2018.
  3. Is This FDA’s Hostile Take-Over of GRAS or Is It Simply Establishing Some Ground Rules? Burdock Group Advisor, December 2017
  4. GAO=General Accountability Office report.
  5. Current Good Manufacturing Practice and Hazard Analysis and Risk-Based Preventive Controls For Human Food. Final Rule. Federal Register 80 FR 55907: “Any raw material or other ingredient that a facility adds to a food product must be lawful. This rule does not address the criteria for determining whether a particular raw material or other ingredient is lawful under the applicable statutory provisions (e.g., under section 409 of the FD&C Act regarding food additives).”
  6. The PEW Charitable Trusts report. Site last visited March 5, 2018.
  7. Burdock, GA and Carabin, IG (2004). Generally recognized as safe (GRAS): history and description. Toxicology Letters 150:3-18.
  8. Hallagan, FB and Hall, RL (1995). FEMA GRAS-a GRAS assessment program for flavor ingredients. Regulatory Toxicology and Pharmacology 21:422-430.
  9. Woods, LA and Doull, J (1991). GRAS evaluation of flavoring substances by the expert panel of FEMA. Regulatory Toxicology and Pharmacology 14:48-58.
  10. A frequent interpretation of the law permitting Health Claims suggests that Congress meant those scientists who may provide SSA could be modeled on the concept of GRAS Expert Panel; however, it was the FDA’s interpretation of the law that required SSA conferees must be from a government agency.
  11. Author of Pharmaphobia: How the conflict of interest myth undermines American medical innovation. Rowman and Littlefield, Lanham, MD 20706.
  12. Rothman, KJ (1993). Conflict of Interest: the new McCarthyism in Science. Journal of the American Medical Association 269:2782-2784.
Tags: , , , ,