Toxicological Principles for the Safety Assessment of Food Ingredients
The Food and Drug Administration (FDA) is the agency responsible for ensuring that food ingredients used in the U.S. are safe for all consumers. In 1982, in an effort to provide guidance concerning appropriate tests, the FDA issued Toxicological Principles for the Safety Assessment of Direct Food Additives and Color Additives Used in Food, commonly known as the Redbook.(7) The Redbook included detailed guidelines for testing the effects of food ingredients on mothers and their developing fetuses. Based on refinements in safety assessment and risk evaluation as well as expansion of knowledge concerning the metabolism and pharmacokinetics of food ingredients, the need to revise and update the 1982 document became apparent. In 1993, Redbook II in draft form(8) was made available for public comment. Since then, test end points and developmental landmarks have been refined. The latest proposed guidelines for developmental toxicity studies are provided here.
The Food and Drug Administration (FDA) is the agency responsible for ensuring that food ingredients used in the U.S. are safe for all consumers. In 1982, the FDA issued Toxicological Principles for the Safety Assessment of Direct Food Additives and Color Additives Used in Food.(7) Based on the color of its cover, the book quickly became known as the Redbook. The Redbook included detailed guidelines for testing the effects of food ingredients on mothers and their developing fetuses. The tests included a chapter on teratology/developmental toxicity studies as well as reproduction studies that spanned several generations. Guidelines for teratology/developmental toxicity studies are discussed here; guidelines for multigeneration studies are discussed in Chapter IVC9a.
Based on refinements in safety assessment and risk evaluation as well as expansion of knowledge concerning the metabolism and pharmacokinetics of food ingredients, the need to revise and update the 1982 document became apparent. In 1993, Redbook II in draft form was made available for public comment.(8) Changes in the chapter on reproduction and teratology/developmental toxicity guidelines were based on extensive literature review and public comments. In late 1996, current drafts of this and several other chapters of Redbook II were presented at a Redbook Update Symposium and the guidelines were compared with current draft guidelines from other national and international regulatory groups.(2) Since then, test end points and developmental landmarks have been refined. The latest proposed guidelines for developmental toxicity studies are provided here in Redbook 2000.
In a developmental toxicity study, the test substance is administered to pregnant animals at least from the day of implantation to the day prior to the day of expected parturition. A short time before the day of expected parturition, the pregnant females are euthanized, the uterine contents are examined, and the fetuses are removed. The fetuses are observed, preserved, and examined for skeletal and soft-tissue abnormalities.
The purpose of developmental toxicity studies is to evaluate the effects of test substances on developing fetuses that result from exposure of either parent prior to conception or to mothers during prenatal development. The adverse effects are as end points that may be used to evaluate the toxic potential of a test substance. The four major manifestations of an effect on the developing organism are: death, structural anomaly, altered or retarded growth, and functional deficiency. For many substances, these manifestations are related to dosage and to the developmental timing and duration of exposure. While high doses produce death, low doses that permit survival may produce malformed, growth retarded, or functionally deficient offspring.
III. Guideline for Developmental Toxicity Studies
The developmental toxicity test may be done as a stand-alone study, or may be part of a multigeneration reproduction study. If it is combined with a reproduction study, assessment of teratological effects may be performed on either the first or second generation, but it is usually performed on the last litter of the generation to maximize exposure to the test agent. As part of a multigeneration study, the fetuses may be exposed to the test substance from conception. In a stand-alone study, treatment must begin early enough to include organogenesis for the species used and should continue to the day prior to the expected day of parturition. This guideline may be used with substances given orally to the rat, mouse, hamster, and rabbit. If the test substance is believed to have the capacity to alter the rate of its own metabolism through induction of metabolizing enzymes or as a result of damage incurred by the liver, then consideration should be given to evaluating the teratogenic potential of the compound by using a separate study.
A. General Recommendations
The following recommendations are applicable to all FDA toxicity studies:
- Studies should be conducted according to Good Laboratory Practice Regulations (GLPs).(6)
- Animals should be cared for, maintained, and housed according to the recommendations contained in the Guide for the Care and Use of Laboratory Animals.(3)
- Healthy animals that have not been subjected to previous experimental procedures should be used. Generally, it is not possible to treat animals for infection during the course of a study without the risk of interaction between the treatment drug and the test substance. The females should not be pregnant and should be nulliparous.
- Test animals should be characterized by reference to their species, strain, sex, and weight or age.
- Animals should be assigned to control and experimental groups in a stratified random manner to minimize bias and ensure compatibility across experimental and control groups for statistical purposes. Each animal must be assigned a unique number.
B. Dose Range-Finding Study
A dose range-finding study is recommended to determine the most appropriate doses, unless suitable pharmacokinetic and metabolic data concerning the test substance are available prior to the start of the study. The dose range-finding study should preferably, but not necessarily, be done in pregnant animals. Comparison of the results from a trial study in non-pregnant animals and a main study in pregnant animals should establish whether the test substance is more or less toxic in pregnant animals than in non-pregnant animals.
C. Main Study
1. Experimental Animals, Species and Strain Selection
When pharmacokinetic and metabolic data or other information on the test substance suggest the most appropriate species for developmental toxicity testing, that species should be used. In the absence of such data, the preferred species are the rat and rabbit. These guidelines include information on the mouse and hamster in addition to the rat and rabbit. These animals are small, easy to care for, and have historically provided consistent results that can be extrapolated to human effects. The strains selected should have high fecundity and should be sensitive to teratogens and embryotoxins. Scaling of doses between species should be based on pharmacokinetic differences between them, unless precluded by differences in overt toxicity.
2. Animal Husbandry
Single housing of the animals is recommended, except during mating. Food and water should be provided ad libitum. The animals' diet should meet all nutritional requirements to support pregnancy in the test species. Special attention should be paid to diet composition when the test material itself is a nutrient, because such material may have to be incorporated into the diet at levels which may interfere with normal nutrition. Under these circumstances, an additional control group fed basal diet may be necessary.
3. Number, Sex, and Age
All test and control animals should be young, mature, primiparous, pregnant females of uniform age and size. A sufficient number of females should be used so that each test and control group consists of approximately 20 pregnant rats, mice, hamsters, or rabbits near term. These are the minimum numbers of pregnant animals for developmental toxicity testing. The objective is to ensure that enough litters are produced to permit effective evaluation of the teratogenic potential of the test substance.
4. Duration of Testing
The test substance should be administered daily throughout the treatment period. The minimum treatment period recommended for developmental toxicity studies is from implantation to Cesarean section one day prior to the expected day of parturition. In rats the approximate timing for this period includes days six through twenty; in mice, days six through eighteen; in hamsters, days four through fifteen; and in rabbits, days six through 29. Alternatively, treatment may be extended to include the entire period of gestation, from fertilization to the day of Cesarean section. If the developmental toxicity test is being conducted as part of a multigeneration reproduction study, the animals are dosed from before conception until they are necropsied. The presence of sperm in the vaginal lavage or the presence of a vaginal plug is considered day zero of gestation.
5. Route of Administration
The test substance or vehicle should be administered by the route that most closely approximates the pattern of human exposure (diet or drinking water). Oral intubation (gavage) may be appropriate in instances where human exposure is via a bolus dose or when it is essential for the animal to receive a specified amount of the test substance. Gavage may also be required when analysis of the agent in the diet is not possible, when the agent is not stable in the diet, or when the agent is not palatable. The maximum volume of solution that can be given by gavage in one dose depends on the test animal's size; for rodents, this should not exceed 1 ml/100 g body weight. If the test substance must be given in divided doses, all doses should be administered within a six-hour period, unless there is justification for increasing the duration of dosing. If the test substance is given by gavage, it should be given at approximately the same time each day, and the volume should be adjusted on a daily basis or every three days based on the animal's body weight. In diet and drinking-water studies, the amount consumed depends on each animal.
6. Mating Procedures
In-house mating of the animals is recommended. A sufficient number of males should be mated to ensure a large gene pool. Siblings should not be mated. Each male may be mated to either one or two females. The following morning, each female should be examined for the presence of sperm in the vaginal lavage or the presence of a sperm plug. The presence of sperm in the vaginal lavage or the presence of a vaginal plug is considered day zero of gestation (day zero of gestation in rabbits is the day insemination is performed or observed).
7. Control and Dosed Groups
Healthy animals should be assigned to test and control groups in a stratified random manner to minimize inter-group weight differences and ensure statistical comparability of relevant variables. The animals may also be assigned in a random procedure which results in comparable mean body weight values among all groups. At least three test groups and one control group should be used in the developmental toxicity study. All groups should be concurrent.
When the test substance is administered in a vehicle, the vehicle without the test substance should be administered to the control group at a volume equal to the maximal amount of vehicle given to any dosed group. If a vehicle or other additive is used to facilitate dosing, the effects on absorption, distribution, metabolism, or retention of the test substance should be considered, as well as alterations of toxicity due to effects on the chemical properties of the test substance. Effects of the vehicle on food consumption, water consumption, or nutritional status of the animals should also be considered.
If there are insufficient data on the toxic properties of the vehicle used in administering the test substance, a sham control group should also be included. If no vehicle is used, then the controls should be sham treated. In all other respects, the control group must be handled and maintained in a manner identical to that used with the groups given the test substance.
Unless limited by the physical or chemical properties of the substance, the high dose should induce some developmental and/or maternal toxicity but not more than approximately ten percent mortality. The high dose should not exceed five percent of the diet for non-nutritive additives. In dietary studies for macronutrient additives, the high dose should be based on nutritional effects rather than toxicological end points.
The low dose should not induce observable effects attributable to the test substance and should be set at a level which is expected to provide a margin of safety. The intermediate doses should be spaced to allow an arithmetic or geometric progression between the low and high doses. The addition of one or more groups is preferable to the use of large intervals between doses.
8. Maternal Toxicity and its Significance
End points which may serve as indicators of maternal toxicity include mortality, body weight, body weight gain, organ weights, feed and water consumption, clinical signs of toxicity, and gross or microscopic lesions. The calculation of a corrected mean maternal weight gain (difference between initial and terminal maternal body weight less the gravid uterus weight) may also be used as an index of maternal toxicity.
Various test substances have selective toxic effects on the male, the female, or the offspring, while other substances exhibit non-specific effects. When mother and offspring are adversely affected by a test substance, it can be difficult to determine if the developmental toxicity is mediated by maternal toxicity or occurs independently of it. Due to differences in metabolism, distribution, and elimination of the test substance, the sensitivity of the maternal system can vary significantly from that of the fetus. The response of the fetus can also differ markedly from that of the mother as a result of the developmental processes taking place that have no counterpart in the adult.
Developmental effects without maternal toxicity are commonly regarded as the most serious manifestations of toxicity, because their occurrence is thought to be the result of greater sensitivity of the developing organism. When developmental effects are found in the presence of maternal toxicity, the primary cause is often left to speculation. Without sufficient evidence to support the premise that developmental toxicity is always a secondary toxic effect in the presence of maternal toxicity, a default is needed. Developmental effects that occur in the presence of minimal maternal toxicity are thus considered to be evidence of developmental toxicity, unless it can be established that the developmental effects are unquestionably secondary to the maternal effects. In situations where developmental effects are observed only at doses where there is a substantial amount of maternal toxicity, then the possible relationship between maternal toxicity and the developmental effects should be evaluated in order to make a proper assessment regarding the toxicity of a test substance.
9. Clinical Observation and Examination of Dams and Fetuses
Throughout the study, each animal should be observed at least twice daily. The first observation should be a thorough clinical examination. The second may involve observing the animals through the cages. Observation times should be selected to permit detection of the onset and progression of all toxic and pharmacologic effects of the test substance and to minimize the loss of animals and organs/tissues. Relevant behavioral changes and all signs of toxicity, morbidity, or mortality should be recorded.
Dams should be weighed immediately before the first dose of the test substance is administered (usually on gestation day six for mice, rats, and rabbits; on gestation day four for hamsters), weekly until necropsy, and at the time of necropsy. If the test substance is given in the diet, weekly body weight is acceptable. If the test substance is given by gavage, body weights should be measured daily or at least every three days. At a minimum, weekly measurements of feed consumption should be made. Fluid consumption should be measured as appropriate. Any dam that shows signs of imminent abortion or premature delivery during the study should be necropsied on the date such signs are observed.
The test should be terminated approximately one day before the expected day of parturition (day 20 or 21 for rats, day 29 for rabbits, day eighteen for mice, and day fifteen for hamsters), when the dams should be subjected to gross pathologic examination. Immediately after the dams are killed, fetuses should be delivered by hysterotomy. Care should be taken to ensure that all fetuses (except those sacrificed before the end of the study) are delivered at approximately the same stage of fetal development. The intact uterus should be removed and weighed in order to calculate the adjusted body weight gain. The contents of the uterus should then be examined for embryonic or fetal deaths and for the number of live fetuses. For dead fetuses, it is usually possible to estimate and describe the approximate time of death in utero (early and late deaths). The number of corpora lutea should be determined for all pregnant animals.
The uterus of each dam that does not appear to be pregnant should be stained in a solution of sodium or ammonium sulfide(4) or other appropriate chemical to enhance the visibility of resorption sites. Evaluation of the females during Cesarean sections and subsequent fetal analyses should be conducted blind in order to minimize unconscious bias.
After removal from the uterus, the weight and sex of each fetus should be determined. The fetus should be examined externally, and all deviations from normal should be noted. Additional end points may be measured, such as the crown-to-rump distance of each fetus. The sex of rabbit fetuses should be determined by internal examination. Each fetus should be weighed individually, and the mean fetal weight per sex per group should be calculated.
Fetuses should be evaluated for skeletal and soft-tissue anomalies. For rodents, approximately one-half of the rodent fetuses should be preserved in Bouin's solution and sectioned by the Wilson serial section technique to evaluate alterations of the soft tissues.(9) The remaining fetuses should be prepared and stained for skeletal anomalies (Alizarin red stain for bone and optional Alcian blue stain for cartilage). The assignment to soft-tissue or skeletal examination should be done randomly or alternately. The alternation procedure is sometimes not followed when an abnormality is found which would be better observed by a different technique. For example, a specimen with an obvious skeletal defect would be prepared for skeletal examination. For identification of rodent bones, the atlas of Yasuda and Yuki (10) may be consulted. Alternatively, all rodent fetuses may be freshly dissected (1),(5) to discover soft-tissue abnormalities, then fixed and examined for skeletal anomalies.
Each rabbit fetus should be examined for both soft-tissue and skeletal malformations and variations. The bodies should be evaluated for soft-tissue anomalies by fresh dissection, followed by fixation and an examination for skeletal anomalies. Internal head structures should be evaluated in at least one-half of the fetal heads of rabbit fetuses. This evaluation should include at least the eyes, brain, nasal passages, and tongue.
When a developmental toxicity study is performed as a stand-alone study, there is no need to perform histopathology unless abnormalities are noted in the organs at the time of Cesarean section.
11. End Points Measured
Because the maternal animal is treated during gestation rather than the developing organism, data should be calculated as incidence per litter or as the number and percent of litters with particular end points. The degree of maternal toxicity may be useful in assessing the relevance of any embryotoxicity or fetotoxicity observed in the treated groups. Parameters used to measure maternal toxicity include body weight and adjusted body weight, feed and fluid consumption, daily clinical observations, and necropsy data such as organ weights.
If treatment is given throughout gestation, implantation may be affected. If, however, treatment begins after implantation, conception and implantation rates should be the same in control and treated groups. End points to be measured per litter should include the number of implantations, corpora lutea, live fetuses (and with separate sexes), dead fetuses, and resorbed fetuses. For litters with live fetuses, mean male and female body weights and the incidence per litter of all divergences from normal fetal development (skeletal and visceral analysis) should also be reported.
12. Analysis of Data
Values from control and test groups of animals should be compared statistically. The following techniques may be used, but others may be substituted if they are appropriate. Maternal body weights may be compared by analysis of co-variance after adjustment for initial body weight, and then analyzed by protected least significant difference tests. Fetal body weights may be evaluated using nested analysis of variance. Anomalies in litters may be compared by Fisher's exact test. Fetal survival and incidence of abnormalities per litter may be compared by analysis of variance after the data have been transformed by use of the Freeman-Tukey arc-sine transformation.
D. Reporting the Results of Developmental Toxicity Studies
Reports of all studies should contain the information required by the Good Laboratory Practice Regulations, including a copy of the study protocol and all amendments, absolute values for all parameters, complete data (individual pups) and tables of data summarized and analyzed by litter. Because the maternal animal is treated during gestation rather than the developing organism, data should be calculated as incidence per litter or as number and percent of litters with particular end points. The dosage rate of the test substance (doses) should be reported as mg/kg/day (milligrams of test substance per kilogram of body weight per day).
Problems commonly encountered in the review of developmental toxicity studies include insufficient numbers of pregnant animals per control or treatment group, non-random selection procedures, and statistical analyses of data on a per-fetus basis instead of a per-litter basis. Careful consideration of recommended guidelines and a review of protocols by the Agency before studies are conducted should help eliminate such problems.
Relevant historical control data may be used to increase the understanding of the study results. When used, historical data should be compiled and presented in an appropriate manner with additional information, such as dates of study, strain of animals, vehicle, and route of administration.
- Barrow, M.V. and Taylor, W.J. (1969). A Rapid Method for Detecting Malformations in Rat Fetuses. J. Morphol. 127:291-305.
- Collins, T.F.X., Sprando, R.L., Hansen, D.L., Shackelford, M.E., and Welsh, J.J. (1998). Testing Guidelines for Evaluation of Reproductive and Developmental Toxicity of Food Additives in Females. Int. J. Toxicol. 17:299-325.
- Institute of Laboratory Animal Resources. (1996). Guidelines for the Care and Use of Laboratory Animals. National Academy Press, Washington, D. C.
- Salewski (Koeln), V.E. (1964). Faerbermethode zum Makroskopischen Nachweis von Implantations Stellen am Uterus der Ratte. Naunyn-Schmeidebergs Archiv Pharmakol. Exper. Pathol. 247:367.
- Staples, R.E. (1977). Detection of Visceral Alterations in Mammalian Fetuses. Teratology 9:A37-A38.
- U.S. Food and Drug Administration. (1978). Good Laboratory Practice Regulations for Nonclinical Laboratory Studies. U.S. Code of Federal Regulations. Title 21 Part 8.
- U.S. Food and Drug Administration. (1982). Toxicological Principles for the Safety Assessment of Direct Food Additives and Color Additives Used in Food. U.S. Food and Drug Administration, Bureau of Foods, Washington, DC.
- U.S. Food and Drug Administration. (1993). Toxicological Principles for the Safety Assessment of Direct Food Additives and Color Additives Used in Food, Redbook 2000 (Draft). U. S. Food and Drug Administration, Center for Food Safety and Applied Nutrition, Washington, DC.
- Wilson, J.G. (1965). Methods for Administering Agents and Detecting Malformations in Experimental Animals. In Teratology Principles and Techniques (J.G. Wilson and J. Warkany, Eds.), pp. 262-277. University of Chicago Press, Chicago.
- Yasuda, M. and Yuki, T. (1996). Color Atlas of Fetal Skeleton of the Mouse, Rat, and Rabbit. Ace Art Co., Osaka, Japan.
Submit comments on this guidance document electronically via docket ID: FDA-2013-S-0610 - Specific Electronic Submissions Intended For FDA's Dockets Management Staff (i.e., Citizen Petitions, Draft Proposed Guidance Documents, Variances, and other administrative record submissions)
If unable to submit comments online, please mail written comments to:
Food and Drug Administration
5630 Fishers Lane, Rm 1061
Rockville, MD 20852
All comments should be identified with the title of the guidance.