Agency Additional Correspondence Letter GRAS Notice No. GRN 000030
CFSAN/Office of Premarket Approval
February 20, 2001
1 Cliffstar Avenue
Dunkirk, NY 14048
Re: GRAS Notice No. GRN 000030
Dear Mr. Marciniak:
The Food and Drug Administration (FDA) is not filing the notice dated December 18, 2000, that you submitted on behalf of ShanStar Biotech, Incorporated(1) (ShanStar). FDA received the notice on January 22, 2001.
The subject of your notice is "cranberry-crospovidone extract, prepared from cranberry juice and crospovidone NF." The notice informs FDA of ShanStar's view that "cranberry-crospovidone extract, prepared from cranberry juice and crospovidone NF" is GRAS, through scientific procedures, for use in desserts or beverages served to patients in long term nursing facilities. The crospovidone component (which constitutes up to 60 percent of the ingredient) is present in "cranberry-crospovidone extract" at a greater amount than is the cranberry component (which constitutes up to 30 percent of the ingredient). Therefore, we are using the term "crospovidone-cranberry juice extract," rather than "cranberry-crospovidone extract" to describe the subject of your notice.
Your present notice is related to a GRAS Notice (GRN 000030) that ShanStar submitted in July 1999 in accordance with the agency's proposed regulation, proposed 21 CFR 170.36 (the GRAS proposal; 62 FR 18938; April 17, 1997; Substances Generally Recognized as Safe (GRAS)). FDA filed GRN 000030 on August 3, 1999, and responded to it in a letter dated December 20, 1999. In our response to GRN 000030, we concluded that GRN 000030 did not provide a basis for a GRAS determination and identified three issues that formed the basis for our conclusion. Likewise, your new notice does not provide a basis for a GRAS determination.
We note that your notice does not follow the specific recommendations of the proposed rule regarding the submission of a GRAS notice. For example, under the proposed rule a notifier would provide "a comprehensive discussion of, and citations to, generally available and accepted scientific data, information, methods, or principles that the notifier relies on to establish safety" (proposed 21 CFR 170.36(c)(4)(i)(A)). Rather than provide such a comprehensive discussion of the published scientific literature that you rely on, your notice merely provides the abstracts of this literature. As FDA discussed within the preamble to the proposed rule (62 FR 18949):
"the notifier should not cite published information unless the cited information bears directly on the GRAS determination. For example, a bibliography describing an exhaustive literature search about a notified substance is of limited or no value in supporting the common knowledge element of a GRAS determination if the relevance of the cited literature is not readily apparent or fully discussed. Moreover, such a bibliography would not, absent a discussion of the relevance of the material cited to the GRAS determination in question, fulfill the technical element of a GRAS determination."
For a notifier, the practical impact of this aspect of the GRAS proposal is that a GRAS notice must be a supported determination of GRAS status. This supported determination should include the notifier's basis for concluding that the conditions of use have been established to be safe and that there appears to be a consensus on that safety among qualified experts, rather than a collection of scientific articles or abstracts. For FDA, the practical impact of this aspect of the GRAS proposal is that the notifier would be showing why FDA need not review the primary scientific literature. FDA will not evaluate a submission that provides scientific abstracts but explains neither the relevance of each abstract nor the reasons why the collection of abstracts adequately establish the safety of the substance for its intended use.
As discussed in the GRAS proposal, FDA is accepting and administering GRAS notices in the interim period between the publication of the GRAS proposal and the publication of any final rule that may result from the proposed rule. Because the proposed rule is not yet final, FDA's acceptance and review of each submitted GRAS notice is voluntary. Because your notice does not provide the comprehensive discussion that we consider central to a GRAS notice, we are not filing your notice dated December 18, 2000 as a newly received GRAS notice. Rather we are considering it as additional correspondence regarding GRN 000030. In the remainder of this letter, we explain in more detail the specific reasons why your new notice does not provide a basis for a GRAS determination.
FDA's response to GRAS Notice 000030
In our letter dated December 20, 1999, we raised the following issues:
- Estimated dietary exposure
In GRN 000030, ShanStar described a broad use of crospovidone-cranberry extract in all foods and beverages. We explained that this was an issue because of the scientific principle that "the dose makes the poison." We explained the procedures that FDA uses to estimate dietary exposure, provided guidance documents that describe these procedures in more detail, and identified a contact who could answer questions that you have about estimates of dietary exposure.
- The crospovidone component of crospovidone-cranberry juice extract.
In GRN 000030, ShanStar stated that crospovidone, as listed in 21 CFR 173.50, is "declared GRAS" for use as a clarifying and stabilizing agent in beverage processing. ShanStar also cited a report of the Joint Expert Committee on Food Additives (JECFA; a joint committee of the Food and Agriculture Organization/World Health Organization), which evaluated several studies (including a 90-day feeding study in rats, a 180-day feeding study in dogs, a teratogenicity study, a pharmacokinetic study, and other short-term studies), and concluded that crospovidone does not represent a hazard to health for its use in food. ShanStar also relied on the conclusions of another firm (International Specialty Products (ISP)) regarding the GRAS status of a related substance, povidone, for use as a diluent in color additive formulations.
In our response to GRN 000030, we explained that your various conclusions about the regulatory status of crospovidone were incorrect in several respects. For example, in contrast to your intended use of crospovidone as a direct food ingredient, 21 CFR 173.50 requires that crospovidone be removed by filtration at the end of its application. In addition, the cited regulation is not based on common knowledge about crospovidone throughout the scientific community, and the safety studies that JECFA relied on in reaching its conclusion are not generally available to the public, e.g., through publication in the scientific literature. Because the generally available JECFA report is an opinion based on data and information that are not generally available, it does not satisfy the threshold requirement of the GRAS standard that the data and information relied on to establish safety must be generally available. Finally, the conclusions of ISP about the GRAS status of povidone are not relevant to ShanStar's product, which contains crospovidone.
- The cranberry component of crospovidone-cranberry juice extract
Cranberry has been found to contain significant amounts of mutagenic phenolic compounds, including quercetin (73-250 mg/kg fresh weight), myricetin (4.0-26.7 mg/kg), and kaempferol (0-2.7 mg/kg). Literature reports demonstrate that quercetin, myricetin, and kaempferol are mutagenic by the Ames test, as well as in a mammalian system. One of these mutagens, quercetin, was also reported to be carcinogenic in long term studies with laboratory animals. Although quercetin is consumed in the diet of individuals who consume cranberries, we estimated that the use of crospovidone-cranberry juice extract under the conditions of use described in GRN 000030 would increase dietary exposure to mutagenic phenolic compounds that are present in cranberries, including the quercetin component that is mutagenic and is a reported animal carcinogen, by as much as 50-fold.
Data and information that you present in your notice dated December 18, 2000
- Intended use and dietary exposure
Your new notice provides conflicting information about the intended use of crospovidone-cranberry juice extract. For example, the GRAS exemption claim states that crospovidone-cranberry juice extract will be used in desserts or beverages served to patients in long term facilities, but other sections (e.g., section 2.1) state that it is used in beverages for special dietary use, food for special dietary use and "regular beverages." In addition to the fact that the intended use is not clear, your notice presents no estimate of dietary exposure to crospovidone and cranberry extract. We do not know how qualified experts could evaluate safety if the dietary exposure is not clearly understood.
- The crospovidone component of crospovidone-cranberry juice extract
The main component of the crospovidone-cranberry juice extract is crospovidone (i.e., up to 60 percent of the crospovidone-cranberry juice extract). Thus, crospodivone is the major component of the substance that you intend to add to food. In your new notice, you inform FDA that much of the clinical data relevant to crospovidone is available to FDA in a drug master file. Importantly, proprietary data, which potentially could establish that the intended use of a food additive is safe, cannot establish that the safety of a food substance is generally recognized by qualified experts. We explicitly raised this point in our response to GRN 000030.
We note that you expect ISP to "apply for a broad scope GRAS for crospovidone in foods." For your information, neither ISP nor any other party could "apply for" GRAS status. Even if ISP submitted a GRAS notice, that notice would inform FDA of ISP's GRAS determination rather than apply for FDA's approval of GRAS status. At this time, you have not provided a basis for ISP's view that crospovidone is GRAS for any use.
You repeatedly cite 21 CFR 175.105 as evidence for your view that povidone and its related homo- and co-polymers are considered equivalent. However, 21 CFR 175.105 provides no support for this view. Under 21 CFR 175.105, povidone (vinyl pyrrolidone) may be polymerized with itself (to form a homopolymer) or polymerized with certain other monomers (to form a copolymer). FDA's approval of the use of vinyl pyrrolidone that is either polymerized with itself (as a homopolymer) or with the monomers listed in 21 CFR 175.105, for use behind a functional barrier, does not establish, and has no relevance to, a purported equivalence of povidone and crospovidone. Moreover, you rely on two letters from ISP, the manufacturer of crospovidone, regarding the GRAS status of a related substance, povidone, for use as a diluent in color additive formulations and as a processing aid in fish aquaculture. First, the fact that FDA approved the use of povidone as a color additive diluent in ink for marking food supplements or fruit provides no basis for concluding that povidone is GRAS for this or any other purpose. An approval would not have been necessary if this use was GRAS. Additionally, as we stated in our response to GRN 000030, the conclusions of ISP about the GRAS status of povidone are not relevant to ShanStar's product, which contains crospovidone.
- The cranberry component of crospovidone-cranberry juice extract
As we already mentioned, your notice lacks a comprehensive discussion of generally available and accepted scientific data, information, methods, or principles that you rely on to establish safety. In addition, your notice does not contain a comprehensive discussion of reports of investigations or other information that may be inconsistent with the GRAS determination (proposed 21 CFR 170.36(c)(4)(i)(B)), nor does it contain the basis for concluding that there is a consensus among experts about the safety of crospovidone-cranberry extract in light of the reports that we brought to your attention in our response to GRN 000030 (proposed 21 CFR 170.36(c)(4)(i)(C)). Your notice in no way explains how the voluminous abstracts that you included in your notice address the question that we raised about the mutagenicity of various phenolic compounds that are present in cranberries, including the quercetin component that also is a reported animal carcinogen.
Because we did not find your notice dated December 18, 2000, suitable for filing, we added it to the administrative file for GRN 000030. The new information that you have provided does not provide a basis for a GRAS determination. In accordance with proposed 21 CFR 170.36(f)(iii), a copy of the text of this letter is available for public review and copying on the Office of Premarket Approval's homepage on the Internet (http://www.cfsan.fda.gov/~lrd/foodadd.html).
Alan M. Rulis, Ph.D.
Office of Premarket Approval
Center for Food Safety and Applied Nutrition