There is no disputing the fact that the Generally Recognized As Safe (GRAS) process is under attack: first, by the General Accountability Office report, then the PEW report and more recently, by the National Resources Defense Council (NRDC). In common with all of these reports is an obvious lack of in-depth knowledge and only a superficial understanding of the rationale behind the GRAS process. In the latter two instances, regarding PEW and the NRDC (which incidentally, include the same people) the commentaries have been subject to innuendo, hyperbole, conflation and unsupported allegations of corruption among GRAS experts. However, despite all the sensationalism and attempts to shake consumer faith in the GRAS process, absolutely no evidence is offered of the public having been harmed as the result of a GRAS determination.
However misguided these attacks may be we should look on this occasion as an opportunity to discuss possible improvements to a nearly 60 year-old process. But before we discuss potential improvements to the GRAS process, we should examine some solutions proposed by Mr. Neltner [of NRDC] and his colleagues.
First, their demand that all GRAS determinations should undergo mandatory Notification. My response is no, not all GRAS determinations should be notified and one the reasons for not so doing, is the ability of companies to keep R&D and marketing information proprietary. The safety of the ingredient will have met the “threshold of common knowledge” through publication of the safety data and, it is the responsibility of the user to limit his use of the ingredient to its acceptable daily intake (ADI), just as he would have to do with an approved food additive – not doing so would be a violation of the law. Further, why should a company invest all the time and money in research and development when anyone could then piggy-back on a GRAS Notification? At the very least, this discourages innovation. It does not make sense for extensive public disclosure of proprietary information and marketing plans, simply on the basis of these unsubstantiated threats. Realistically, if all new drug candidates were required post comprehensive information on specifications, manufacturing, uses and safety, would we have many new drugs today?
Second, the suggestion that the GRAS process should be discarded in favor of an expedited food additive petition (FAP) such as a seen with the Food Contact Substance (FCS) Notification program for packaging ingredients, where the use and the identity of the ingredient are held by FDA as confidential information. Although the promise of confidentiality is seductive, this is a straw man argument and is unworkable on several levels. First, the FCS system was designed for only small amounts of ingredients (in the small ppm range and lower), which makes a safety decision relatively easy to achieve, in fact a large percentage of packaging ingredients finding their way into foods is likely below the Threshold of Regulation. Second, like mandatory minimum sentencing laws, expert judgment (and often, common sense) is removed from the process because of the mandatory testing thresholds imposed; whether the dogmatic criteria of the Redbook or FCS are used, most candidate food ingredients are used at much higher levels and would be required to adhere to extensive testing requirements, including lifetime animal studies, thereby eliminating a speedy route to market, both of which would also be a financial barrier to small companies and an impediment to innovation.
Third, GRAS should be eliminated and everything should go through the standard food additive petition (FAP) process. This is unworkable on the basis of at least two points. First, as above, the FAP process (and FDA), unfailingly default to the worst case scenario and out of an abundance of caution by a very conservative agency (having nothing to lose from asking for more data and everything to gain by postponing a decision), the approval process can be prolonged interminably. Secondly, from a “boots on the ground” viewpoint, FDA has approved approximately 15 new petitions (not just amendments) in the last 20 years; this is a rate of less than one FAP per year. On the other hand, the GRAS notification program has administratively processed approximately 487 Notifications since 1998 to July of 2014, a rate 31 Notifications per year. If FDA was required to review these Notifications as petitions, it would require over 300 years to dispose of these Notifications as FAPs. This does not even include the 30 – 100 flavor ingredients GRASed by FEMA each year. Clearly, the greatest beneficiary of the GRAS process is FDA itself!
The GRAS process is too important to be distracted by these simplistic solutions, all of which have the potential to eventually derail a historically successful process. We need to stop and think about a difficult problem that deserves more of our attention; however any solution must involve an honest effort by both FDA and industry.
As regards the role of FDA, the agency needs to step up to the plate, work with industry and stop playing the role of a bystander.
First, it has been over 25 years since the last food ingredient survey, wherein ingredient manufacturers voluntarily reported manufacturing data. These surveys had been funded by FDA through the National Academy of Science for several years, beginning in the early 1970’s. The survey was dropped for cost considerations, but FDA could always look toward a more cost-conscious vendor or industry could even volunteer compilation of the data. There are several models that could be applied to contain costs and ensure security of the data submitted. FDA has long argued that actual use data would be more accurate, but the prospect of manufacturers sharing their formulas with anyone (including FDA) is simply not realistic. Historically, the ingredient survey data was regarded as fairly accurate with a high reporting rate by companies and was always regarded as a good source of exposure data.
Second, FDA needs to exercise some enforcement and call out manufacturers who make claims of ingredients that are GRAS, but simply not notified. Just a few letters each year to such manufacturers requiring documentation of a legitimate GRAS determination would be stimulus enough to improve the overall quality of GRAS determinations, if poor quality GRAS determinations are as rampant as implied.
Third, a GRAS Notification should be treated as a rebuttable presumption of the safety-in-use of the ingredient. FDA should stop playing the role of final arbiter of GRAS determinations. Not all of the food ingredient safety expertise in the US is confined to FDA – Congress recognized as much when the concept of GRAS was elucidated in the Act. Further, if FDA requires 3-5 years (by its own admission), to determine if a food additive is safe how can it determine whether an ingredient does not meet the criteria for GRAS in only six months? Also, by FDA’s own admission, the threshold for an acceptable GRAS has been a moving target. Further, the agency sometimes confuses GRAS Notification with GRAS affirmation – there is a stark difference between the two. The folly of a deep dive for data at FDA and a dissection of each Notification is subject to all the difficulties described in the Congressional Shays Committee report of the early 1990’s on the failure of the GRAS affirmation process. A Notification should require no more than identification and description of the ingredient, the aggregate level of consumption, proof of publication of safety data and evidence of expertise of the Experts declaring GRAS status. A Notification is a Notification, not a request for GRAS affirmation.
Lastly, if FDA has doubts about the safety-in-use of an ingredient following review of the Notification or safety data, it should be allowed to request more information, but with the understanding that the data would be placed in a Food Master File that would not be accessible outside the FDA. If, after this second review of data, there is a dispute about the safety-in-use of the ingredient, FDA needs to take definitive action as described in §409 of the Act.
Industry also needs to step up to the plate as well if it expects the GRAS process to continue and not become the subject of politically motivated witch hunt by Congress.
First, industry should exhibit more concern about vicarious liability. Production of a GRAS dossier should be the product of competent food ingredient safety experts (best exemplified by peer-reviewed publications in the field) and certainly not by persons without a history of demonstrable expertise in food ingredient safety. All too often there has been more concern by some in industry about the cost of safety testing or the cost of a comprehensive GRAS dossier to determine the choice of a path forward. Costs and ROI should always be a consideration, but low bidder should not be the deciding factor. The Parke Doctrine elucidates the potential vulnerabilities of decision makers and FSMA rules now in place could result in civil action as well as criminal prosecution, for an unwise choice. The Parke Doctrine also makes it especially clear that neither a consultant nor an expert panel constitute a firewall for poor choices.
Second, industry should exhibit more concern about its greatest vulnerability, the appearance of a conflict of interest. Third party compilation of data is superior to in-house judgment about what should be included in a dossier and thus tamps down the potential for accusations of cherry-picking information. Also, whether the dossier is compiled by in-house staff or a consultant, the expert panel should be independent of either and expect a fee regardless of the decision rendered.
Third, just as EFSA has instituted a periodic review of food ingredients, industry needs to consider a similar step of vetting ingredients according to a rational system of prioritization. Each company should prioritize a review schedule for the ingredients it uses, especially considering the amount used by the company and that used by others based the FDA re-initiated food ingredient survey data (see above). Also, the bulk of ingredients in §184 of the CFR were GRAS affirmed on the basis of information available in the late 1970’s – a distant age in terms of scientific progress and achievement and, newly developed testing methods and more recent concerns regarding potential safety issues. For those ingredients with extensive use by industry, independent reviews by companies or even a pooling of resources by consortia could undertake reviews to ensure continued safety-in-use of ingredients. We cannot rationally expect, nor should want FDA to perform this review in camera or without the input of industry expertise.
Let me sum up by quoting from Lyndon Johnson, “any jackass can kick down a barn, but it takes a carpenter to build one.” I believe the barn-kicking naysayers will stir up enough doubt and fear among consumers that they will stimulate an increase in their membership rolls or whatever their eventual goals may be – and although their Parthian shots, snarky comments and McCarthyisms are frustrating to most of us, they are sure to gain the attention of a naïve public with a short attention span. And, for the life of me, I cannot figure out why venerated groups keep sharing their megaphone with those that have such limited knowledge of such a complex subject and nothing constructive to offer.
So then, it is up to us, the carpenters, FDA and industry, to shore up and improve the barn, which we can do by working cooperatively, with each entity sharing in the costs and resources to ensure the continued safety-in-use of ingredients and the preservation of a truly unique example of self-policing, the process we know as GRAS.