Animal Welfare Information Center Newsletter, Summer 1994, Vol. 5, no. 2

Validation of In Vitro Methods: Regulatory Issues

by
Neil L. WIlcox, D.V.M., M.P.H.
Director, Office of Animal Care and Use
Center for Veterinary Medicine
U.S. Food and Drug Administration

Introduction
The FDA Mission
Drugs, Cosmetics, and Devices Defined
Regulations and Animal Use
Regulation of Cosmetics vs. Drugs and Medical Devices
Methods for Hazard/Safety Determination
Current Utility of In Vitro Methods
In Vitro Testing Methods and Regulatory Acceptance
In Vivo "Replacement"
The Tiered Approach to Substance Testing
Components of Methods Evaluation
Summary
References

Introduction

The U.S. Food and Drug Administration (FDA) is primarily concerned with public safety. To that end, the use of animals in toxicity testing has played an important role in hazard/safety determination for regulated products. FDA encourages the development of alternative methods to animal testing (e.g., in vitro tests) and is aware that many such tests are in various stages of evolution. New laws have been enacted that either ban the use of animals in testing for certain products or mandate the development and validation of alternative methods to animal testing. Research has resulted in much activity in the development of in vitro methods intended for use as screens, adjuncts, and replacements for current in vivo standards. For example, although technical progress in the development of non-whole animal testing methods has occurred, to date, no single test, or battery of tests, has been accepted by the scientific community as a replacement to the animal model currently used in ocular irritation testing, the Draize test. For replacement of the in vivo standard with in vitro tests, further research is needed to better understand the mechanisms of action of ocular irritants in vivo. Criteria for the validation and acceptance of in vitro methodologies intended to replace in vivo models need to be well defined; moreover, new risk assessment paradigms to analyze information generated by in vitro methods need to be developed. The international community should strive for harmonization based upon consistent, science-based standards, while pursuing improved methods intended to protect public health worldwide.

The FDA Mission

The mission of the FDA is to assure the American consumer that foods are pure and wholesome, safe to eat and produced under sanitary conditions; that drugs, medical devices, and cosmetics are safe and made from appropriate ingredients; and that labeling and packaging for these products are truthful and not deceptive. The authority for this mission is issued under the following laws: 1) Federal Food, Drug, and Cosmetic Act (FFDCA) (21 U.S.C. 301-392) and its accompanying regulations and the Fair Packaging and Labeling Act (FPLA) (15 U.S.C. 1451-1461), which apply to foods and drugs for humans or animals, cosmetics, and medical devices; 2) Sections of the Public Health Service Act (PHSA) relating to biological products for human use (42 U.S.C. 262-263) and control of communicable diseases (42 U.S.C. 264); and 3) The Radiation Control for Health and Safety Act, relating to electronic products which emit radiation, such as x rays, lasers, microwave ovens, and TV sets (42 U.S.C. 263b-263n).

Drugs, Cosmetics, and Devices Defined

A drug is an article intended for use in the diagnosis, cure, mitigation, treatment, or prevention of disease in humans and other animals and articles (other than food) intended to affect the structure or any function of the body of humans or other animals. It is the intended use which determines whether an article is a drug; therefore, foods and cosmetics may also be subject to the drug requirements of the law if therapeutic claims are made for them. The FFDCA defines cosmetics as articles intended to be applied to the human body for cleansing, beautifying, promoting attractiveness, or altering the appearance without affecting the body's structural function. A device is defined as any health care product that does not achieve any of its principal intended purposes by chemical action in or on the body or by being metabolized. The term "devices" also includes components, parts, or accessories of devices, diagnostic aids such as reagents, antibiotic sensitivity discs, and test kits for in vitro diagnosis of disease and other conditions (1).

Regulations and Animal Use

The FFDCA and the PHSA require manufacturers of certain consumer products to establish, before marketing, that such products meet the safety and effectiveness requirements of the law and are properly labeled. FDA regulations prescribe the type and extent of premarket testing that must be conducted, depending on the legal requirements applicable to the particular product and on the technology available to fulfill those requirements. Testing may include physical and chemical studies, non-clinical laboratory studies, and clinical tests.

Animal tests are required by FDA for drug products, vaccines, certain medical devices and electronic products, food and color additives, and new animal drugs. "Although the FFDCA does not require that cosmetic manufacturers or marketers test their products for safety, the FDA strongly urges cosmetic manufacturers to conduct whatever toxicological or other tests are appropriate to substantiate the safety of the cosmetics (2)." These tests should be state-of-the-art and generally represent consensus of the scientific community.

Regulation of Cosmetics vs. Drugs and Medical Devices

Cosmetics marketed in the United States, whether made here or imported, must comply with the FFDCA, the FPLA, and regulations issued under the authority of these laws. Unlike those products regulated by FDA that require premarket review prior to approval, there is no requirement for or against the use of animals in the substantiation of safety of the methods used by cosmetic manufacturers in testing their products. Ordinarily, a cosmetic comes under scrutiny only if a problem surfaces post-marketing. For example, if a product causes injury, such as severe ocular or dermal irritation or is otherwise shown to be deleterious to public health, the agency can require its withdrawal from the market. There appears to be a trend away from the use of animals in cosmetic testing, as many manufacturers join a growing group of those who claim to no longer use the animal model.

The application process for approval of human drugs may incorporate both in vivo and in vitro methods for toxicity testing; however, to determine efficacy or substantiate safety for products intended for use in humans, clinical trials are required for final approval of drugs and devices. Although not a regulatory requirement, the final product formulation in cosmetic testing is usually not marketed to the public until some form of limited human testing has occurred. The fundamental issue is that hazard determination and safety substantiation, although inextricably linked, are not the same. Safety is a relative concept and is achieved through a process of elimination. After all the evidence is considered, a decision is made based upon benefits of the proposed product compared to its risks.

Methods for Hazard/Safety Determination

For approximately 50 years, the rabbit has served as the model for eye and skin irritation testing (viz., the Draize tests). To date, salient issues have centered around "replacement" of the Draize eye and skin tests with in vitro methods. Unlike drugs and medical devices, where the product may not be marketed without regulatory approval, cosmetics are presumed substantiated for safety before marketing. If this is not the case, then the sponsor of the product must communicate this fact by placing a warning statement on the label. Since a warning statement on a cosmetic label is exceedingly rare, and most likely non-existent, the consumer regards cosmetics marketed in the United States as safe. In vitro methods play a significant role in the toxicological evaluation of raw chemicals, therapeutic drugs, medical devices, biologicals, and cosmetics; their current application is that of screening for toxicity, especially for moderate to severe irritants, primarily as a component of a tiered testing system that seems to differ considerably "in house" depending upon the company or government to whom one may speak.

Current Utility of In Vitro Methods

The use of in vitro methods as part of different ocular irritancy testing systems was recently demonstrated in a workshop organized by the Interagency Regulatory Alternatives Group (IRAG) titled "Workshop on Eye Irritation Testing: Practical Applications of Non-Whole Animal Alternatives." Two hundred people participated in the workshop where approximately 40 laboratories from around the world submitted 55 data sets representing 23 in vitro methods. Expert working groups were formed to review each of the major assay systems, and their summaries were presented during the workshop. While several salient topics relevant to the status of non-whole animal methods development and use were addressed during this workshop, a significant message was clear: many companies and some governments have established alternative testing systems to help evaluate certain chemicals and products for ocular irritation potential, and in vitro methods are an important part of those processes.

In Vitro Testing Methods and Regulatory Acceptance

Certain external influences (3,4) are driving change that will most likely result in in vitro methodologies occupying a more prominent place in toxicity testing. The cosmetic industry, in particular, is looking to Federal agencies for guidelines identifying regulatory acceptance criteria for submitting data generated from in vitro methods intended to at least partially replace some data that, heretofore, originated from in vivo testing. Although specific criteria for regulatory acceptance of in vitro models have not yet been published, validation of a proposed model may be considered an important criterion in this process.

Although validation of new methods is not a primary responsibility for regulatory agencies, validation by the scientific community may be considered pivotal to regulatory acceptance. However, "validation" of a method does not necessarily guarantee regulatory acceptance. Like pre-market approval of regulated products, the acceptance criteria of a proposed new method will largely be determined by the sponsor's claim for the test. As part of the review of a proposed in vitro method during the risk assessment process, new data need to be evaluated, and herein lies a formidable challenge; viz., new standards of data comparison need to be considered.

When an in vitro method is proposed to test for an in vivo response, such as dermal or ocular irritation, the qualitative data generated by the test as compared to the in vivo standard are imperative. In other words, what are the endpoints and how do they relate to the tissues evaluated by the Draize. For example, discussion often centers around the empirical vs. mechanistic approach. A correlative (empirical) method, which may successfully identify a severe dermal or ocular irritant early in the evaluative stages of testing, may be acceptable to the regulatory agency as a screen, but since it does not predict safety, would be inadequate as a replacement. However, for a method to successfully identify a severe irritant with an acceptable level of false positives (substantiate hazard/high sensitivity) and predict safety with a low incidence of false negatives (substantiate safety/high specificity), a full understanding of the mechanism by which the technique detects irritation in a specific tissue appears to be essential to replacement (3).

In Vivo "Replacement"

What, then, needs to be accomplished by the scientific community to advance the acceptance of new techniques? For true replacement of an in vivo with an in vitro model to successfully occur, research should focus on the mechanisms of dermal or ocular irritation in humans. It appears, therefore, that methods development targeted for replacement will need to successfully predict the presence or absence of irritation at the physiological, anatomical, biochemical, or molecular level of tissue pathology. Tissue repair, the reversibility of lesions, is an important facet in classification of substances; moreover, in the review of FDA-regulated substances, those products that cause irreversible damage to some tissues would be less likely to receive approval. Similarly, some products may not cause severe tissue damage or visible irritation upon exposure but may cause considerable discomfort or pain. Demonstrating such phenomena will be most difficult without in vivo modeling. These examples highlight formidable challenges that need to be addressed as issues germane to total replacement are identified and explored.

As we try to envision replacement of the animal model in the context of safety substantiation, several considerations clearly need to be addressed. Few would accept the simplistic notion of total replacement of an animal model, such as the Draize, with a single in vitro test. A risk assessment system that replaces an animal model will necessarily consist of a multidisciplinary approach that incorporates information from several sources in a systematic or tiered approach.

The Tiered Approach to Substance Testing

The first level will be that of reviewing information already known about the particular class in which the substance resides. It is important to note here that much of the historical information is derived from in vivo testing such as dose response relationships as well as toxicokinetic and toxicodynamic data. Other sources include structure activity relationships and known physical-chemical properties of the class from which the substance is derived. The next tier in the system may consist of a battery of assays, each measuring various mechanisms of dermal or ocular irritation. Finally, a decision point is reached; if there is sufficient evidence that the substance is a severe irritant, it may be classified. If insufficient information exists to classify, then further in vivo data are required. As methods are examined for their role in irritation testing, a standardized validation paradigm must be developed. A framework for such a model was reported in ATLA (6) from the CAAT (7)/ERGATT (8) Workshop and proposed by the Johns Hopkins University Center for Alternatives to Animal Testing (9).

Components of Methods Evaluation

As a validation paradigm is considered, the protocol and the data generated by the study are of particular importance for serious evaluation. The protocol for an in vitro method should clearly identify the in vivo endpoints, and the data generated from the test should provide information relevant to these endpoints. Once in vitro data have been accumulated, the standard to which they will be compared is extremely important. To that end, careful consideration must be given to the guidelines established for in vitro-in vivo data comparisons. Currently, this is established on a case-by-case basis with FDA-regulated products. Until a model for risk assessment based upon in vitro data is developed, there must be adequate means of comparing the results of our tests with known in vivo outcomes.

From a regulatory perspective, the following precepts should be considered as guidelines of the scientific process for validation testing of proposed methodologies:

Define the mechanistic relevance of the in vitro test endpoint to effects observed in vivo.
Determine the relationship between the known in vivo dermal or ocular irritation potential of the test substance and the in vitro test results.
Demonstrate the quality of the in vitro data by evaluating the method's protocol, intra-laboratory repeatability, inter-laboratory reproducibility, number and type of chemicals used, and adherence to GLP (10) standards.
Define potential uses and limitations of the alternative method including type or class of chemical to which it has application and how the data might be used for practical safety or hazard determinations (11).
For new molecular entities with no history of previous testing, animal data will have to validate any in vitro testing system for that class of compounds.

The following expands upon nomenclature germane to new methods development, validation, and regulatory acceptance:

Sensitivity is the ability of the proposed method to detect that proportion of those compounds tested that are truly positive as an irritant or toxicant. Although a high sensitivity is important in any risk assessment scheme, false positives in sensitivity testing error toward the conservative and, therefore, do not present the unfavorable consequences that may occur with false negatives in specificity testing.
Specificity is the ability of the proposed method to detect those compounds tested that are truly negative as an irritant or toxicant. False negatives in specificity testing mean truly positive substances fail to be detected, thus, allowing a potential toxicant to be classified incorrectly and inadvertently allowed for human/animal use. Clearly, a high specificity is extremely important from a regulatory perspective.
Predictive Value in screening tests is the probability that a positive test is truly positive and a negative test is truly negative.
Precision is the quality of being sharply defined or stated (e.g., the number of distinguishable alternatives from which a measurement was selected). An example of precision would be the standard deviation or comparing the standard deviation to the mean or the coefficient of variation (12).
Repeatability is the ability of the results of a testing method to perform consistently when conducted several times within a particular laboratory. The standard deviation may be employed as a measurement of precision when considering repeatability.
Ruggedness or rigor refers to the method's ability to achieve a suitable degree of repeatability in the sponsor's laboratory. Without ruggedness or rigor, a method would not be expected to perform adequately in different laboratories.
Reproducibility is the ability of the results of a testing method to perform consistently when conducted in different laboratory settings. The standard deviation may be employed as a measurement of precision when considering reproducibility.
Bias is the deviation of results or inferences from the truth or processes leading to such deviation. Bias may occur when any trend in the collection, analysis, interpretation, publication, or review of the data leads to conclusions that are systematically different from the truth. There are many sources of bias including flawed study design, data collection, statistical summary data, data analysis or interpretation, instrumental error, handling of outliers, and prejudice in study procedures that lead to one-sidedness in any facet of a study (13).

Summary

Many challenges face stakeholders now and in the future for risk assessment in the area of toxicity testing. The notion of replacement will vary considerably depending upon many variables. Particularly cogent at this time is the need to identify criteria for both the comparison of in vitro with in vivo data and regulatory acceptance for in vitro methods intended to replace the animal model in toxicity testing. The transition from comparing in vitro data to the animal standard to that of a "Gold Standard" is complex and will require the implementation of a novel risk assessment paradigm. Validation of in vitro methods needs to adhere to the scientific precepts of purpose and endpoint identification, correlative vs. mechanistic basis, intralaboratory reproducibility, interlaboratory repeatability, protocol standardization, technology transfer, chemical reference standardization, data base development and quality control/quality assessment through GLP-like standards. Finally, this necessarily requires an international effort to coordinate and harmonize the multifaceted issues in risk assessment for hazard/safety determination.

To obtain additional information, Dr. Wilcox may be contacted at 301-594-1798 (301-594-1830 FAX) or by writing to the Office of Animal Care and Use, U.S. Food and Drug Administration, Center for Veterinary Medicine, MPN-2, HFV-4, 7500 Standish Place, Rockville, MD 20855.

References

Requirements of Laws and Regulations Enforced by the U.S. Food and Drug Administration, U.S. Department of Health and Human Services, Public Health Service, Food and Drug Administration, DHHS Publication No. (FDA) 89-1115, 1989.
Ibid, DHHS Pub. No. (FDA) 89-1115, 56.
6th Amendment, EC Cosmetic Directive (93/35/EEC).
National Institutes of Health Revitalization Act of 1993, Title 11, Sec. 205; Title XIII, 1301: 27-28, 50-51.
Wilcox, N.L. (1992). The status of eye irritation testing: A regulatory perspective. Lens and Eye Toxicity Research 9(3&4): 259-271.
Balls, M., et al. (1990). Report and recommendations of the CAAT/ERGATT workshop on the validation of toxicity test procedures. Alternatives to Laboratory Animals, Reprinted from ATLA 18: 313-337.
The Center for Johns Hopkins University Center for Alternatives to Animal Testing.
European Research Group for Alternatives to Animal Testing.
Goldberg, A. M., et al. (1993). Framework for Validation and Implementation of In Vitro Toxicity Tests, Report of the Validation and Technology Transfer Committee of the Johns Hopkins Center for Alternatives to Animal Testing.
Code of Federal Regulations, Title 21, Part 58, Good Laboratory Practice for Non-Clinical Laboratory Studies.
Green, S., et al. (1993). Criteria for in vitro alternatives for the Eye Irritation Test. Food and Chemical Toxicology 31(2): 81-85.
Dictionary of Epidemiology, Second Edition, J. Last (Ed.), Oxford University Press, Inc., New York, N.Y., 1988.
Ibid, Dictionary of Epidemiology (1988).

This article appeared in the Animal Welfare Information Center Newsletter, Volume 5, Number 2, Summer 1994

Go to:
Contents, Animal Welfare Information Center Newsletter
Top of Document

The Animal Welfare Information Center
U.S. Department of Agriculture
Agricultural Research Service
National Agricultural Library
10301 Baltimore Ave.
Beltsville, MD 20705-2351

Phone: (301) 504-6212
FAX: (301) 504-7125
E-mail: awic@nal.usda.gov

Policies and Links

March 12, 1998
This page's URL is http://www.nal.usda.gov/awic/newsletters/v5n2/5n2wilco.htm