The use of structure–activity relationship analysis in the food contact notification program

Abstract

Food contact substances (FCS) include polymers, paper and paperboard, and substances used in their manufacture, that do not impart a technical effect on food. Moreover, FCSs are industrial chemicals generally consumed at dietary concentrations (DC) of less than 1 mg/kg food (ppm), and more commonly at less than 0.05 ppm (50 ppb), in the daily diet. As such, many industrial chemicals have been analyzed for toxicological concern, some of which may share structural similarity with FCSs or their constituents, and the majority of these studies are available in the public domain. The DCs of these compounds lend themselves to using structure–activity relationship (SAR) analysis, as the available “expert systems” and use of analogs allows for prediction and management of potential carcinogens. This paper describes the newly implemented food contact notification (FCN) program, the program by which FDA reviews FCSs for safe use, the administrative review of FCSs, the SAR tools available to FDA, and qualitative and quantitative risk assessments using SAR analysis within the regulatory framework of reviewing the safety of FCSs.

Introduction

The Office of Food Additive Safety (OFAS), a program office located in the Food and Drug Administration’s (FDA) Center for Food Safety and Applied Nutrition, is responsible for ensuring the safe use of U.S. food ingredients and food packaging. In this capacity, OFAS administers the program that evaluates safety information in industry submissions for the use of various categories of food substances. In the past, these submissions were primarily petitions for food and color additives and generally recognized as safe substances.

Recently, Section 309 of the FDA Modernization Act of 1997 amended Section 409 of the Federal Food, Drug, and Cosmetic Act (or the Act) to establish a new process, referred to as the food contact notification (FCN) process, as the primary method of authorizing new uses of food additives that are food contact substances (FCS). Section 409(h)(6) of the Act defines an FCS as “any substance intended for use as a component of materials used in the manufacturing, packing, packaging, transporting or holding of food if such use is not intended to have a technical effect in such food.” The FCN process is described in Title 21 of the Code of Federal Regulations, Sections 170.100 through 170.106 (21 CFR 170.100-106). The information requirements for an FCN are addressed in 21 CFR 170.101.

The FCN process, and its predecessor the food additive petition (“petition”) process, have two important similarities. First, under both processes the submitter bears the burden of demonstrating that the intended use of the FCS is safe. Second, the safety standard is the same for all food additives as codified in Section 409(c)(3)(a) of the Act. Thus, the data and information in an FCN and petition are comparable; however, there are two main differences in the two processes. First, in contrast to the petition process, the FCN process will not result in a food additive regulation listing authorizing the use for any manufacturer of the FCS. Rather, an FCN for an FCS is effective only for the manufacturer identified in the FCN. Second, in contrast to the petition process which lacks a mandatory timeframe for review, an FCN becomes effective and the FCS may be legally marketed for the proposed use 120 days after acceptance of the submission, unless FDA concludes that the safety has not been demonstrated and so informs the notifier. As of March 2005, FDA has received and processed over 500 FCNs. This paper discusses the FCN administrative process as well as the use of structure–activity relationship (SAR) analysis in the safety evaluation of FCSs.

The OFAS utilizes a two-phase approach in the administrative and technical review of FCNs (Fig. 1). Phase 1 review, conducted during the first three weeks of initial receipt of the FCN, consists of an administrative review to ensure that the basic data and informational elements are present and that the submission meets the requirements set forth in the applicable statute and regulations. During this process, the consumer safety officer and technical reviewers assigned to the FCN examine it for regulatory, environmental, chemical, and toxicological completeness. This includes conducting literature searches and a preliminary assessment of the data to identify potential concerns. In addition, the SAR team, an OFAS team separate from that assembled for the particular FCN, also provides input regarding additional toxicology data on the FCS and any substances that might become components of food because of its use (i.e., constituents). A formal meeting is conducted 20–30 days after receipt to discuss the submission, identify deficiencies that inhibit going forward with the safety review, and discuss potential concerns that may arise during phase 2 review. If any deficiencies are noted, the FCN is considered incomplete and the notifier is asked by letter to address the deficiencies in a timely manner (categorized as “not accepted”). If the notifier adequately addresses all deficiencies in a timely manner, usually within 10 working days from the date of the deficiency letter, the FCN is considered complete (categorized as “accepted”). Conversely, if the deficiencies are not adequately addressed, the status of the FCN remains “not accepted.” The notifier is informed that the FCN is not acceptable in its present form and is encouraged to withdraw the FCN without prejudice to future filing. The 120-day review period begins on the day the complete FCN is received by FDA.

The review process enters phase 2 once the FCN has been “accepted” for review. Phase 2 review involves a detailed analysis of the safety data and other information contained in the FCN or otherwise obtained by FDA. If a safety issue arises during the phase 2 review, the notifier is given the opportunity to withdraw the FCN without prejudice to future filing; alternatively, in the absence of additional data, FDA may object to the FCN. If no safety issues arise during the phase 2 review, the FCN will become effective and the FCS may legally be marketed for the proposed use at the end of the 120-day review period.

The OFAS has found that a two-phase approach to the administrative and technical review is crucial in assuring that FCNs are reviewed in an efficient and effective manner. The OFAS SAR team plays a pivotal role in both the phase 1 and 2 reviews of FCNs.

The food additive anti-cancer, or Delaney, clause in Section 409 of the Act provides that “no additive shall be deemed safe if it is found to induce cancer when ingested by man or animal….” The Delaney clause applies to the food additive as a whole and not to the individual constituents of the food additive. Moreover, constituents are evaluated under the general safety standard using risk assessment procedures. This regulatory approach was articulated by FDA in what is often referred to as the Constituents Policy (U.S. Food and Drug Administration, 1982). Accordingly, if a constituent is the subject of a bioassay, the bioassay is reviewed to determine whether or not the constituent is a carcinogen (Gaylor et al., 1997). If the constituent is not a carcinogen, other toxicity data are considered in the safety assessment of that substance. On the other hand, if the constituent is determined to be a carcinogen, quantitative risk assessment procedures are used to determine whether the proposed use of the food additive would present an unacceptable level of risk (Gaylor et al., 1997, Kokoski et al., 1990, Lorentzen, 1984). The unit cancer risk (UCR, slope factor) derived from the linear, low-dose extrapolation (or other models for determining risk) is used in combination with the estimated daily intake [EDI, obtained by multiplying the dietary concentration (milligram of chemical in kilogram food) by the assumed daily intake of 3 kg food and dividing by the assumption that an individual weighs 60 kg, resulting in mg/kg bw/day] for the constituent to determine the upper-bound lifetime cancer risk. Guidance on the use of bioassay data in risk assessment is contained in the industry guidance document entitled “Preparation of Food Contact Notifications for Food Contact Substances: Toxicology Recommendations” available on the OFAS website (http://www.cfsan.fda.gov/~lrd/foodadd.html). Other methods for assessing the carcinogenic potential of constituents, such as short term genetic toxicity tests are used, as discussed below, to assess the potential carcinogenic risk in the absence of bioassays, as these assays are less expensive and time consuming and, thus, more likely to have been conducted.

An FCN may contain the results of several tests to address the genetic toxicity of the FCS or its constituents. Genetic toxicity testing is one of many issues discussed in the toxicology guidance document for FCSs (noted above). That guidance document discusses general safety testing recommendations, the relevance of various types of safety studies including genetic toxicity testing, as well as the evaluation of structural similarity of a substance to known toxicants. The toxicology guidance document also addresses the recommended toxicity tests to support “tiered” ranges of DCs. As part of this tiered approach, genetic toxicity studies are recommended at DCs >0.5 ppb, equivalent to an EDI of >1.5 μg/person/day, whereas subchronic studies as well as other endpoints are recommended at DCs >50 ppb.

A DC of 0.5 ppb is the lowest tier on our tiered testing scheme. Our analysis of toxicity data indicates that carcinogenicity is the pivotal endpoint at DCs <50 ppb. Statistical analysis of bioassays compiled in the Gold database (Gold and Zeiger, 1997) indicates that a threshold of concern of 0.5 ppb (DC) based on the carcinogenic potency profiles is adequate for protection of the public health (Rulis, 1992). This analysis served as the basis for FDA’s Threshold of Regulation (TOR) exemption policy under 21 CFR § 170.39 Threshold of regulation for substances used in food-contact articles (U.S. Food and Drug Administration, 1995).

To ascertain a substance’s potential to be a genotoxic carcinogen, OFAS recommends more than one test to allow for the evaluation of both mutagenicity and clastogenicity. As put forth in the toxicology guidance document, the recommended genetic toxicity tests at all DCs >0.5 ppb include tests for: (1) gene mutations in bacteria, (e.g., Ames assay), and (2) gene mutations in mammalian cells (e.g., in vitro mouse lymphoma tk^+/− assay) or cytogenetic damage in mammalian cells (e.g., in vitro detection of micronuclei or chromosomal aberrations). For DCs >50 ppb, FDA recommends an in vivo test for chromosomal aberrations in addition to the tests addressing the endpoints recommended at DCs >0.5 ppb.

It is reasonable to expect that the structural and physicochemical properties of a chemical are potential determinants of its toxicity. Although the FDA does not see a need to conduct genetic toxicity testing at DCs > 0.5 ppb, it is recommended that all available and reasonably applicable information on the potential carcinogenicity of a substance be discussed in the overall safety assessment. This information may include the relationship of the substance to other substances containing structural or functional similarity. It is important to note that this approach is entirely consistent with that recommended for food additives that are the subject of TOR exemption requests under 21 CFR § 170.39. Moreover, § 170.39(a)(1) specifies that the substance that is the subject of the request will be granted an exemption only if it has not been shown to be a carcinogen in humans or animals, and there is no reason, based on the chemical structure of the substance, to suspect that the substance is a carcinogen. This allows FDA to consider the carcinogenic potential of a substance whose structure suggests that it may be a potent carcinogen, even if the use of the substance would result in a DC ⩽0.5 ppb. Thus, a consideration of SAR analysis has always been an integral part of the safety assessment for food packaging. A consideration of structure alerts (SA) and the results of genotoxicity testing on a substance have recently been used to suggest higher thresholds for a TOR process for FCSs (Cheeseman and Machuga, 1997, Cheeseman et al., 1999).

SAs may be viewed as functional groups (or moieties or biophores) present within a substance that have been shown to be associated with biological activity. Biological activity includes interaction with macromolecules as in protein interaction (i.e., receptor binding) and, in the case of genotoxic carcinogens, DNA reactivity. With regard to genotoxic carcinogens, functional groups are identified on large molecules to determine which groups are capable of chemically interacting with DNA. Such fragments likely represent electrophilic moieties in the molecule. Thus, the presence or absence of an SA in an untested chemical (i.e., a chemical that has not been tested in a bioassay) would be an indicator of DNA reactivity or mutagenic activity and, hence, a predictor of carcinogenic potential. Various classification schemes have been developed, most of which are based on classification of either mutagens or carcinogens into broad categories, such as the schemes detailed in Cramer et al. (1978), Cheeseman et al. (1999), and FDA’s Redbook (Rulis et al., 1984). The OFAS has found that with regard to assessing carcinogenic potential of an untested substance, one of the most useful schemes is that developed by Ashby and Tennant (1991). The Ashby and Tenant classification scheme for SAs assesses the functional groups present in the untested substance. A list of functional groups associated with DNA reactivity (genotoxicity) is given in Table 1. Our list is based on Ashby’s composited “model structure” (Ashby and Paton, 1993) and a related functional group list compiled by Munro et al. (1996). We have found the combination of these two lists represents a more comprehensive overview of structurally reactive groups.