*This article will be published in the Summer issue of IngredientsOnline.com
Food ingredient regulation in the United States (U.S.) is unique among other regulatory domains in the U.S. and worldwide by incorporating two equally legitimate and parallel processes of compliance to the law – one that is government managed (a Food Additive Petition or FAP) and another that is a self-approval system (Generally Recognized As Safe or GRAS). Further, once a GRAS determination has been made by qualified experts, it can be submitted to FDA for its review in the form of a GRAS notification, but there is no legal or regulatory obligation to do so and many are not submitted to FDA. This means of compliance with federal law offers several advantages and the latter of which, GRAS, can offer a lower cost of regulatory compliance and a faster route to market, but should not be mistaken as a method for circumventing the burden of proof of safety-in-use or as a method for keeping proprietary information secret. Please note also, that regardless of the process of federal compliance (FAP or GRAS), the approval is very specific – the approval is for a particular use of the ingredient (i.e., at a specific concentration for each specific purpose and food category, and including its specifications and method of manufacture), not the ingredient per se (i.e., a generalized approval of the ingredient for any use in any food or amount, produced by any method).
Similarities between a food additive petition and GRAS
FAP or GRAS require much of the same type of basic information and, in general, includes the following: specific chemical identity (including, but not limited to scientific and common names, Chemical Abstracts Service Number, empirical and structural formula) or if derived from a natural substance, the genus and species; specifications and stability information; food categories in which the ingredient would be used (for example, baked goods, hard candy, snack foods – for a full list, refer to 21 CFR 170.3(n)) and the amount used in each category (from this information the exposure information may be calculated); proposed technical effect (for example, will it be used as a buffer, preservative, texturizer – for a full list, refer to 21 CFR 170.3(o)); manufacturing or processing procedures and; safety data, from which a (safe) Acceptable Daily Intake (ADI) is derived.
Three differences between a food additive petition and GRAS
The three primary differences between the two processes (i.e., FAP or GRAS), are (1) the required amount of safety data to be collected for review, (2) the requirement for publication of safety data and (3) the total time required for review and compliance and thus, time to market. Again, both processes will require a certain amount of core information including, but not limited to in vitro genotoxicity and cytoxicity data and in vivo genotoxicity data and; other in vivo toxicity information, which will include performance of a 28-day or 90-day safety study in rats (see table below). All studies should comply with guidelines from the Organization for Economic Co-operation and Development (OECD). As often as not, the aforementioned studies are often sufficient for a review of a GRAS determination because the amount and type of data required for a determination of safety is at the discretion of the independent experts who must shoulder the responsibility for their decision. For a food additive petition, the engine driving in-life study requirements is more informatics based and either using computational toxicology software or the “Redbook” available at the FDA website, the structure of the ingredient is used to indicate the potential hazards based on the chemical structure(s) shared with other substances with known toxicologic outcomes. This hazard information is combined with an overlay of potential exposure (as determined by the amount added to each of the food categories) to determine the type and extent of testing necessary. Additional consideration (and subsequently, safety factors) might be given to potentially large exposure to vulnerable groups, such as sweeteners to which children might be disproportionately exposed.
Thus, computational toxicology (or the Redbook), exposure/consumption calculations and the possibility of vulnerable groups invariably increases the safety data requirements.
FDA has often claimed that the only difference between a GRAS and a FAP is the obligation to publish the safety data supporting the GRAS, although the new guidelines for GRAS (Federal Register 81:54960; August 17, 2016), hint that publication of the entire GRAS dossier may soon be required (including manufacturing information and consumption data). Our experience at Burdock Group has been that even some proprietary manufacturing data must be included in a notification to FDA. While FDA indicates that a FAP is not accompanied with an obligation to publish, little is left confidential as everything in a FAP is either included in the “final rule” published in the Federal Register or is available through a Freedom of Information Act request.
Time to market- GRAS without FDA Notification:
43 weeks to perform core studies (if required) + 17 weeks to complete the GRAS dossier1,2
Time to market from the decision to initiate a process for compliance through the GRAS process or a FAP is also vastly different. That is, the core information (see above), even including a 90-day rodent study and genotoxicity testing (conducted simultaneously), could be completed within a year, during which the GRAS dossier can be compiled and an acceptance for publication of the safety data by a journal editor within two months of receipt of the final reports of the studies. Once this has been completed, the Expert Panel can move ahead on its deliberations for safety-in-use and work is usually complete in two weeks. At this point, the Expert Panel has determined the substance to be GRAS, the substance is compliant with federal law and can be marketed even though the substance has not undergone the FDA notification procedure.
Time to market- GRAS with FDA notification:
79 weeks to perform core studies (if required) + 37 weeks to complete the GRAS, notify the GRAS dossier, and receive a no objection letter from FDA3,4
Additional time is needed if FDA notification is desired. The FDA requires that the published study results be available to the journal readership which is often an additional 2-3 months after the editor’s acceptance. Additionally, FDA supports the view that a publication should be available to scientists for comment for another 3-6 months before submission to FDA as a notification. Notifications require another six months for review by FDA, although there may be some requests by FDA for clarification or additional information, but if a response to FDA questions require more than two weeks, FDA often will suggest the notification be withdrawn and resubmitted when the information becomes available.
Cost and time to market comparison for GRAS vs. FAP
Time to market- Food additive petition:
250-324 weeks5
Time to market for a FAP is considerably lengthened as a result of the time required for study performance and review by the agency. As noted earlier, the type of testing required is less based on expertise but based on something FDA regards as more quantifiable and reproducible, and because a FAP is such an iterative process, FDA likely will have suggested a testing regimen for the petitioner. Therefore, based on a substance of a theoretical intermediate toxicological potential and a cumulative human daily exposure of 25 – 500 parts per billion (ppb) in the diet, the substance will have a Concern Level II, which mandates the following schedule of testing (see table). If the amount consumed would be greater than 500 ppb daily, additional data would be required including chronic toxicity and carcinogenicity studies, considerably delaying the potential approval. Publication of studies is not necessary as the final study reports from the testing laboratory are submitted with the FAP.
When to do a FAP versus a GRAS?
Obviously, for anyone concerned with timing and cost, a FAP would be a poor choice; however, sometimes there is no choice. Historically, FAPs have been the best pathway for controversial substance such as one with particularly active chemical groups (such as those that might combine with cellular DNA) or whose chemically-related kinsmen are known to produce specific hazards such as reproductive toxicity or cancer. Good candidates for a FAP would also include those substances that are known not to be held in particularly high environmental regard such as those containing chlorine atoms. Traditionally, those substances, such as artificial sweeteners, that would experience high consumption by children, a particularly vulnerable group, would be best served with a FAP. Also, for those substances listed in 21 CFR 182.1(b)(2) (a list of otherwise GRAS substances), additional uses of any of these substances requires a FAP.
The main problem cited by most is that a FAP takes too long. GRAS is obviously a more expedient route to approval, but a GRAS must be done with care toward scientific documentation and credibility of the GRAS author. Also, although there is no obligation to report a GRAS to FDA, but under the Food Safety Modernization Act (FSMA), the user has the obligation to demonstrate the safety of the ingredient and a hitherto non-notified GRAS may be demanded by FDA. If the GRAS does not adequately demonstrate the safety-in-use of the ingredient, products containing it are subject to seizure.
Final Thoughts
While GRAS is more expedient and requires less investment, it should never be done “on the cheap” because the manufacturer puts his ingredient, his reputation and his customers’ products at risk. Also, because of negative reports by the U.S. General Accountability Office and others critical of the GRAS process, FDA will likely be more proactive and search out non-notified GRAS determinations with inadequate vetting.
References
1 Assumes specifications and stability data already generated.
2 If studies are not performed, this assumes supporting safety data already available in the scientific literature or, if the substance was commonly used in food prior to January 1, 1958, no testing data may be required.
3 Refer to footnotes 1 and 2.
4 If FDA determines the data supporting safety are inadequate, the notifier may be forced to withdraw and may resubmit once the data are generated and published.
5 Depending on how many tasks are performed simultaneously (i.e., the animal studies), the task time could be shortened.