12 Application of Computational Methods for the Safety Assessment …
235
12.1 Introduction
The Office of Food Additive Safety (OFAS) in the Center for Food Safety and Applied
Nutrition (CFSAN) of the United States Food and Drug Administration (US FDA)
regulates safe use of food ingredients. The term “food ingredients” includes food
additives and color additives used in food, substances classified as food contact substances (FCS; also known as indirect food additives), and substances classified as
generally recognized as safe (GRAS) as well as substances derived from bioengineering for their intended use in food [1]. OFAS review scientists assess chemical
identity, exposure information [2, 3], and toxicological data [1, 4] included in industry
submissions to support the safety of food ingredients under the intended conditions
of their use. OFAS review scientists also conduct independent literature searches to
find any additional information relevant to the reviews and analyses. Each evaluation
is performed to determine with reasonable certainty that the food ingredient is not
harmful when used as intended. In some instances, a postmarket safety evaluation
may be conducted for a certain food ingredient due to a change in exposure to the
consumer as a result of its intended use or due to the availability of new scientific
data that raises questions regarding its safety under the conditions of use. Under these
circumstances, OFAS review scientists would perform an updated safety assessment
based on contemporary methodology and guidelines.
As computational science advances and new data become available, integrating
predictive toxicology methods into safety assessments has become an agency-wide
priority. An example of this is the FDA’s “Predictive Toxicology Roadmap” [5]. This
roadmap identifies specific toxicological areas that could benefit from improved predictivity, such as modeling, to support decision making whenever there are data gaps.
This effort shows FDA’s commitment to developing and adopting new technologies,
including computational toxicology. At OFAS, the review scientists are exploring
the scientific utility of computational methods to assist with food ingredient safety
evaluations, particularly when scientific questions may not be directly answered by
experimental study data. In this chapter, we provide an overview of the underlying
concepts and applications of some of these methods for evaluating the safety of food
ingredients. These include:
(a) quantitative structure–activity relationship (QSAR) modeling to support
the safety evaluation of those substances whose exposure is less than
150 µg/person/day and have limited and insufficient toxicity data,
(b) toxicokinetic (TK) modeling and simulation to gain insight into the internal
exposure and modes of action of substances with intermediate to high toxicological potential, and
(c) bioinformatics approaches to evaluate potential allergenic effects from food proteins. Furthermore, we describe OFAS’s in-house food ingredient knowledgebase, Chemical Evaluation and Risk Estimation System (CERES) that contains
administrative, chemical, and safety data on a diverse set of food ingredients.
Lastly, we discuss current challenges for utilizing these computational methods
to their full potential for evaluating safety of food ingredients.
235
12.1 Introduction
The Office of Food Additive Safety (OFAS) in the Center for Food Safety and Applied
Nutrition (CFSAN) of the United States Food and Drug Administration (US FDA)
regulates safe use of food ingredients. The term “food ingredients” includes food
additives and color additives used in food, substances classified as food contact substances (FCS; also known as indirect food additives), and substances classified as
generally recognized as safe (GRAS) as well as substances derived from bioengineering for their intended use in food [1]. OFAS review scientists assess chemical
identity, exposure information [2, 3], and toxicological data [1, 4] included in industry
submissions to support the safety of food ingredients under the intended conditions
of their use. OFAS review scientists also conduct independent literature searches to
find any additional information relevant to the reviews and analyses. Each evaluation
is performed to determine with reasonable certainty that the food ingredient is not
harmful when used as intended. In some instances, a postmarket safety evaluation
may be conducted for a certain food ingredient due to a change in exposure to the
consumer as a result of its intended use or due to the availability of new scientific
data that raises questions regarding its safety under the conditions of use. Under these
circumstances, OFAS review scientists would perform an updated safety assessment
based on contemporary methodology and guidelines.
As computational science advances and new data become available, integrating
predictive toxicology methods into safety assessments has become an agency-wide
priority. An example of this is the FDA’s “Predictive Toxicology Roadmap” [5]. This
roadmap identifies specific toxicological areas that could benefit from improved predictivity, such as modeling, to support decision making whenever there are data gaps.
This effort shows FDA’s commitment to developing and adopting new technologies,
including computational toxicology. At OFAS, the review scientists are exploring
the scientific utility of computational methods to assist with food ingredient safety
evaluations, particularly when scientific questions may not be directly answered by
experimental study data. In this chapter, we provide an overview of the underlying
concepts and applications of some of these methods for evaluating the safety of food
ingredients. These include:
(a) quantitative structure–activity relationship (QSAR) modeling to support
the safety evaluation of those substances whose exposure is less than
150 µg/person/day and have limited and insufficient toxicity data,
(b) toxicokinetic (TK) modeling and simulation to gain insight into the internal
exposure and modes of action of substances with intermediate to high toxicological potential, and
(c) bioinformatics approaches to evaluate potential allergenic effects from food proteins. Furthermore, we describe OFAS’s in-house food ingredient knowledgebase, Chemical Evaluation and Risk Estimation System (CERES) that contains
administrative, chemical, and safety data on a diverse set of food ingredients.
Lastly, we discuss current challenges for utilizing these computational methods
to their full potential for evaluating safety of food ingredients.