Skip Navigation
Johns Hopkins Bloomberg School of Public HealthCAAT

Technical Report No. 5

The International Status of Validation of In Vitro Toxicity Tests

A Report of the CAAT/TCA Technical Workshop of June 16-20, 1991


The Johns Hopkins Center for Alternatives to Animal Testing (CAAT) and the Cellular Toxicology Committee of the Tissue Culture Association (TCA) organized a joint worshop on The International Status of Validation of In Vitro Toxicity Tests in conjunction with the World Congress on Tissue Culture held in Anaheim, California, June 16-20, 1991. The workshop involved 12 participants from academic, government, and industry as well as 17 observers. The discussions were divided into nine sections, each lead by one of the participants. The titles of the sections and the discussion leaders are listed in Table I. Following the meeting, a draft report was edited by John M. Frazier (The Johns Hopkins Center for Alternatives to Animal Testing) and circulated to participants for revisions and updates. This report reflects the updated documentation provided by the participants as of June, 1992. The final report was reviewed by several members of the Advisory Board of CAAT prior to publication.

The report consists of an Executive Summary prepared by John M. Frazier and individual reports submitted by the participants. The opinions expressed by the participants are their own and do not necessarily reflect the opinions of CAAT. It is hoped that these comments will further the process leading to an international consensus on validation and harmonization of new testing methodologies. Continuing discussions of these issues are important to assure international cooperation and sensitivity to regional concerns. CAAT and the Cellular Toxicology Committee of the TCA would like to thank The Procter & Gamble Company for partial support of the workshop and making the publication of the report possible.

International Status of Validation of
In Vitro Toxicity Test Validation
(1) Status of Validation Projects in the USA -- Jack Lipman
(2) Status of Validation Projects in Europe -- Bas Blaauboer
(3) Status of Validation Projects in Japan -- Makoto Umeda
(4) Needs for Chemical Banks -- Judson Spalding
(5) Needs for Data Banks -- Horst Spielmann
(6) Needs for Reference Laboratories -- Eugene Elmore
(7) Needs for Cell Banks -- June Bradlaw
(8) Needs for Scientific Panels -- Neil Wilcox and Leon Bruner
(9) Models for National/International Coordination and Facilitation of Validation -- Hugo van Looy


Prepared by:John M. Frazier, Ph.D.
Environmental Health Sciences Division
Toxicological Sciences
Johns Hopkins University
615 N. Wolfe Street
Baltimore, MD 21205

The workshop was divided into two major issues: (1) a review of the current status of ongoing validation studies, and (2) a discussion of the infrastructure components needed to facilitate validation activities in the future. It should be noted that the components of the validation infrastructure were previously proposed by Goldberg and Frazier (CAAT Newsletter: Vol. 8, No. 1).

Current Status of Validation Studies

Four papers were presented (Gettings, Lipman, Blaauboer, and Umeda) describing the current validation activities in the United States, Europe, and Japan. The number of active projects is impressive and includes projects supported by trade organizations (e.g. Soap and Detergent Association [SDA], Cosmetic, Toiletry, and Fragrance Association [CTFA], Pharmaceutical Manufacturers Association [PMA]), government organizations (e.g. Commission of the European Community [CEC], ZEBET), commercial organizations (e.g. Advanced Tissue Sciences, In Vitro International), and other organizations (e.g. Fund for the Replacement of Animals in Medical Experimentation [FRAME], MIEC). These activities have produced some positive results. One example is the Organization of Economic Corporation and Development (OECD) considerations of the British Toxicology Society Fixed Dose Procedure as an in vivo refinement and reduction in the classical LD50 test. However, the various activities have also highlighted many of the problems which still exist concerning the design and conduct of in vitro test validation projects. Some of the critical issues are:

  1. How do we establish the "goodness" of a test for a particular purpose -- linear (nonlinear) parameter correlations, rank correlations, concordance, some other measure?
  2. Do we compare in vitro performance to in vivo animal or human data? How do we rationalize the choice?
  3. How do we select the chemicals to be used in validation studies?
  4. How do we keep track of all the data being generated in academic, industrial, and government laboratories?

These are a few problems that we have not yet been able to resolve.

In spite of the failure to resolve major issues, it is clear that in vitro testing systems have made significant progress in contributing to product development and chemical safety evaluations. Many corporations have established in vitro toxicity testing laboratories and are using information provided by internally validated methods in toxicological evaluations. Thus, the weight of evidence seems to be in favor of an increasing role of in vitro models in safety/hazard assessments.

Validation Infrastructure

A wide range of discussions relating to desirable components of a validation infrastructure were presented. The major components discussed were:

  1. chemical banks;
  2. data banks;
  3. cell banks;
  4. reference laboratories; and
  5. Scientific Review Panels.

The general consensus of the workshop participants was that these components would greatly facilitate validation activities and enhance the compatibility of data from one validation project to the next.

A major stumbling block to progress, as pointed out by several discussion leaders, is the lack of well-defined funding sources to support these activities. One reason for this situation is the splintering of interests and authority. Commercial interests are diverse and often do not recognize their common governmental authorities frequently have a narrow perspective of the problems when viewed within the constraints of their governing legislation. The situation is constantly evolving and sensitivity to these issues is growing as exemplified by the U.S. Interagency Regulatory Alternatives Group (IRAG) activities as well as the new proposal by the CEC, a center for validation. All participants agreed, that it is important to promote these infrastructural activities and maintain international communications to assure a priori harmonizations of new methodologies, thus avoiding potential regulatory barriers to international trade.



Prepared by:Stephen D. Gettings, PhD, D.A.B.T.
The Cosmetic, Toiletry and Fragrance Association
1101 17th Street, N.W., Suite 300
Washington, DC 20036

Several commentators (17, 3) have suggested that validation is an ongoing process and that the criteria used for assessing the outcome of validation activities are essentially defined by variables such as the current stage of development of the tests at: (i) the time that a particular stage of the validation process was completed; (ii) the particular nature of the test materials selected; and (iii) the particular use for which the tests were intended. For these reasons, Flint has suggested adoption of the term "evaluation", as first used by Gettings & McEwen (28), to describe the various elements (biological function; correlation with in vivo toxicity; practical usefulness) of the validation process (17).

Scientific criteria for the validation of in vitro toxicology tests have been described by Frazier (19), but have yet to gain widespread acceptance, particularly (and importantly) by national and international regulatory agencies. Although some commentators (17, 19) have implied that there must be a strong mechanistic basis for the implementation of tests (even when a strong correlation between measured in vitro activity and known in vivo toxicity has been demonstrated), the report and recommendations of the CAAT/ERGATT Workshop of the Validation of Toxicity Test Procedures (3) recognizes that the acceptable degree of mechanistic similarity of an in vitro procedure to an existing animal model is conditional upon the ultimate use ("the specific decision making process") for which it is deemed appropriate.

The degree of predictivity that will satisfy the intended application of a given in vitro procedure must thus be established as an essential part of any validation exercise. Predictivity is established by validation exercise. Predictivity is established by evaluating the degree of correlation between measured in vitro activity and known toxicity in vivo, so-called "correlation with reference classification" (17). Given these "definitions", several ongoing and recently completed projects in the United States are described. The approach (particularly with regard to statistical correlation of in vivo and in vitro data) varies considerably from project to project. Although several sources were consulted for purposes of this brief review, the following examples serve only to illustrate the extent of evaluation activities in the United States and should by no means be considered a definitive listing of all projects currently in progress or recently completed. Similarly, no attempt has been made to compare the results of studies using nominally similar test protocols, nor to review the relative performance of different test methods.

Trade Association Programs

Extensive information is available on the evaluation programs conducted by U.S. trade associations, and these have been reviewed previously (26). The intention of both the Soap and Detergent Association (SDA) and the Cosmetic, Toiletry, and Fragrance Association (CTFA) is merely to provide their respective industries with preliminary information on the performance of a series of potential alternatives to the Draize eye irritation test (neither association has, as yet, directed its attention to dermal irritation). Neither program addresses validation, although many elements of the design and conduct of the two programs might usefully be incorporated into any validation effort.

SDA Program. The Soap and Detergent Association (SDA) evaluation of candidate nonanimal tests is focused on the determination of the eye irritancy potential of cleaning products and ingredients. The SDA effort began with a pilot study (Phase I) involving 14 test systems and eight test materials (4). Factors which contributed to test selection were: (i) available data on applicability to detergent ingredients; (ii)mechanistic relevance; and (iii) feasibility. Based upon their performance in Phase I, a subset of nine tests were put forward for further evaluation in Phase II. Since alkalinity (or acidity) is an important characteristic of many cleaning products, the tests materials in this second phase contained a number of alkaline products, and some acidic and basic chemicals; furthermore, the representation of various types of detergent products and ingredients were increased by the addition of 15 new test materials for which Draize data were available. As a check on reproducibility of test results, several Phase I samples were provided to investigators in Phase II. All of the materials were provided as blind samples. The results suggested the possibility that a combination of in vitro and alkalinity data might yield useful estimates of eye irritancy, even though alkalinity alone is inadequate (5). The SDA evaluation is now continuing (Phase III) with a still smaller subset of six test methods (cell protein accumulation assay, CAM-Vascular Assay, corneal epithelial plasminogen activator assay, Neutral Red Uptake Assay, SIRC Assay, Tetrahymena Motility Assay) and 30 more test materials. As in Phases I and II, Phase III is an intralaboratory assessment; evaluation of test performance is based upon comparison of correlation coefficients of the association between log10 9(in vitro) vs. log10 (in vivo) data, and Spearman rank correlation. The reference data for Phase III will include low volume rabbit eye irritancy data, as well as standard Draize test data. The test materials have been chosen to represent a range of detergent products and ingredients, and include a set designed to provide information on the alkalinity boundary between irritant and nonirritant as a function of surfactant concentration. In total, 36 new test materials have been added; this brings the total number of test materials in Phases I, II, and III of the SDA Program to 59.

CTFA Program. The CTFA Evaluation of Alternatives Program is essentially a multicenter intralaboratory assessment of several potential in vitro alternatives to the Draize eye irritation test, conducted under blind conditions, using a limited set of identical coded test materials representative of the types of formulations manufactured and marketed by the cosmetic and personal care product industry (27). The program design contains many of the elements of interlaboratory assessment recommended by the joint CAAT/ERGATT Workshop on the Validation of Toxicity Test Procedures (3). Coding/distribution of test materials and collation/statistical analysis of data is assigned to an independent research laboratory (Columbus Division, Battelle Memorial Institute), which does not participate in the actual performance of either the in vivo or in vitro tests. The choice of test material in each phase is consistent with the objectives of the program, namely to determine the limitations and effectiveness of in vitro tests over a range in irritancy exhibited by cosmetic/personal care product formulations. Individual test materials are generic formulations selected as representative of a variety of product types. A different category of test materials is being tested in each phase of the program. Test materials supplied to participating laboratories are individually coded and subject to stringent quality control (30). Specialized procedures for handling and distribution of test materials receipt and analysis of data (including development of specific computer software), have been implemented. Formulae and stability information are retained on file by the program director at CTFA. The identity of each test material sample shipped to investigators is known only to the Program Manager at Battelle.

Statistically, the CTFA program is designed as a correlational analysis of Draize primary eye irritation test data with comparative data from a selection of in vitro tests (16). Intralaboratory reproducibility of both in vitro and in vivo data will be assessed in each phase. Although interlaboratory variability (of in vitro test performance) is not evaluated, this does not preclude comparison of the relative performance of each assay. Test performance is evaluated by correlation of in vivo (Maximum Average Draize score; [MAS]) and in vitro score (29, 16). Initially, assays are ranked in order of their ability to statistically separate test outcomes when compared to MAS values. The relationship between in vitro and in vivo test scores for those assays which exhibit relatively low scatter and shown to have least amount of discordance with the Draize procedure in terms of ability to discriminate between test material irritation potential, are subsequently further analyzed by regression analysis (29, 30). Performance of the assays is thus based on the magnitude of the 95% prediction intervals of plots of in vitro versus in vivo response.

In Phase I, ten representative hydro-alcoholic, personal-care formulations were subject to the Draize primary eye irritation test and 25 in vitro assay protocols. Of the assays evaluated, six (EYTEX™ MPA, HET-CAM I Assay, Neutral Red Release Assay, HET-CAM II Assay, CAM-Vascular Assay, Pollen Tube Test) were shown to have the most agreement with the Draize test (30). Thirty individual protocols (essentially 14 types of in vitro endpoint) have been evaluated in Phase II of the CTFA Program; 18 oil/water emulsion formulations (representing such products as sunscreens, hair conditioners, and cleansing creams) were selected as test materials (31). Although the results have yet to be fully analyzed, preliminary data suggests that the correlation of in vitro results with in vivo data for the materials evaluated in Phase II is not as good as that for the materials evaluated in Phase I. CTFA plans to evaluate approximately 35 individual protocols (encompassing over a dozen types of endpoint) in Phase III. Twenty-five surfactant-containing formulations (e.g., shampoos, facial cleansers, shower gels) will be investigated. In a parallel program, CTFA and participating member companies are evaluating the Low Volume Eye Test (LVET) as an alternative to the Draize test and as the basis of comparison for the in vitro alternatives investigated in Phase I, II, and III of the CTFA Evaluation Program. The LVET differs from the Draize test in the volume of material instilled in the animal eye and in the method of instillation. A single dose of 10 L of the respective test material is instilled in the eye, rather than the 0.1 mL dose administered in the Draize test. The material is instilled directly on the cornea of the eye rather than in the everted lower lid of the eye. The irritation scoring scale for the low volume test is identical to that for the Draize, although additional emphasis is given to the number of days required for the eye to recover (i.e., days to clear). These procedures are intended to more closely simulate accidental human eye exposure to potentially irritant materials. A report on the results of comparing the LVET with Draize and in vitro test data using Phase I test materials is anticipated shortly (32).

Other Programs

In addition to the well-publicized programs organized by the Soap and Detergent Association (SDA) and the CTFA, a number of other projects conducted in corporate laboratories, or in association with independent contract laboratories and purveyors of commercial tests are currently in progress (it is difficult to document the evaluation projects conducted on a strictly internal basis by individual companies and corporations). None of these projects have received as much public scrutiny as the SDA or CTFA programs, and (in the past) details have been difficult to obtain. However, the results of several evaluations are now beginning to be presented at symposia, e.g., much of the information presented in this communication was presented at the 10th Anniversary CAAT Symposium, Johns Hopkins University, Baltimore, MD (April 1992). It is anticipated that this trend will continue, and that the results of such studies will eventually appear in published literature.

Ocular Irritation Tests

The most readily obtainable information on validation/evaluation pertains to the in vitro prediction of ocular toxicity. Over thirty different protocols have been identified, encompassing 18 types of in vitro endpoints (20), many of which are being evaluated in the CTFA and SDA programs described previously. The following descriptions pertain to independent (nonaligned) evaluations of in vitro performance which generally employ considerably greater numbers of test materials than the trade association-sponsored programs, and which are generally of specific and particular interest to the individual companies engaged in the evaluation. For ease of presentation, individual projects have been grouped under headings which, in general terms, describe the various types of in vitro endpoints which are being evaluated.

Physical/Chemical Models -- The EYTEX™ System (33, 38) is a commercially available in vitro ocular irritation test developed by National Testing Corporation (now In Vitro International, Inc., formerly Ropak Laboratories, Irvine, CA). The rationale behind the test system is based upon the premise that opacity, or precipitation of proteins in the cornea, is an important component of acute ocular toxicity observations in vivo. The basis of the test is an aqueous (synthetic, i.e. nonanimal) protein matrix, the degree of opacity of which is reported to be proportional to the irritancy potential of the test substance evaluated.

The EYTEX™ is involved in several intra- and interlaboratory validation/evaluation exercises in the United States. Studies have been, or are being, conducted by cosmetic/personal care, pharmaceutical, consumer product, petrochemical, agrochemical, and chemical companies to assess the transferability and reproducibility of the methods, as well as to establish correction with in vivo data (V.C. Gordon, personal communication.) As an example, the EYTEX™ system has been used to evaluate 42 adult and baby shampoos (13), in a study conducted by Helene Curtis, Inc. The sulfate surfactants are both ethoxylated and nonethoxylated, whereas the primary components of the baby's shampoo's were branched chain ethoxylated alkyl and amphoteric surfactants. All shampoos were diluted 1:10 in deionized water prior to testing. The EYTEX™ scores were used to establish irritation classification scheme (minimal, mild, moderate, severe); accordingly, in vitro/in vivo irritation classification correlation for adult and baby shampoo data was 87% and 100% respectively, and correlation of the Draize scores to the EYTEX™ scores was statistically significant (r2=0.90).

As a second example, in a joint publication (37) Avon Products, Inc., and In Vitro International Inc., report the results of 465 cosmetic product formulations and raw ingredients which were evaluated with the EYTEX™ system to determine eye irritation potential. The test samples included over 30 different product types and represented a wide range of eye irritancy. All the EYTEX™ protocols available at the time of the study (STD, MPA, RMA, AMA) were used. Samples were evaluated double-blind and the EYTEX™ data correlated with rabbit eye irritation data obtained from Avon's historical records. Kruszewski  et al claim that a positive agreement of EYTEX™ with the in vivo assay was demonstrated by an overall concordance of 80%. Assay error was 20%, of which 18% was claimed due to an over-estimation of sample irritancy (false positives) and 2% was attributed to underestimation (false negatives). The investigators noted that overestimation error in this study was due in part to the inability of the protocols to accurately classify test samples with very low irritation potential. The report cites 100% sensitivity and 85% predictability, which the authors ascribe to the efficiency of EYTEX™ in identifying known irritants; however, a specificity rate of 39% showed the EYTEX™ assay to be weak in discerning non-irritants (the authors claim that the EYTEX™ protocols used in this study were not designed to identify nonirritants).

In contrast to the first two examples, Bruner et al (9) evaluated the EYTEX™ System using a set of 17 surfactant-containing materials and found a low correlation (r=0.29) with the Low Volume Eye Test as the basis of comparison with irritation in vivo.

Mammalian Cytotoxicity Assays -- Numerous in vitro cytotoxicity assays have been proposed as potential alternatives to the Draize eye irritancy test (10). A study by Kennah et al (35) evaluated the use of in vitro cytotoxicity data for predicting the ocular irritancy potential of 24 chemicals (six surfactants, seven alcohols, four ketones, four acetates, and three aromatics). BALB/c 3T3 cells were grown overnight, then exposed for 30 minutes for at least four different concentrations of each chemical. Linear regression analysis of the log concentration versus percentage of control growth was used to calculate the concentration of toxicant that inhibited the normal growth rate by 50% (GI50). The rank ordering of cytotoxicity (based upon GI50 values) was surfactants > aromatics > alcohols > ketones or acetates. The larger molecular weight representative of each series (i.e., 2-ethyl-1-hexanol for alcohols) had lower GI50 values than those of the lower molecular weight substances. The GI50 values were then directly calibrated against in vivo ocular irritancy quantitated as percentage corneal swelling following exposure of rabbits to the same test chemicals. A significant linear correlation between cytotoxicity and ocular irritancy was established only for surfactants and alcohols. For acetates, ketones, and aromatics there was little correlation. Kennah et al (35) attributed overall poor correlation between cytotoxicity and ocular irritancy to differences in mechanisms of irritancy, and concluded that the lack of correlation illustrates that in vitro cytotoxicity data cannot necessarily be used to predict the ocular irritancy potential of a broad spectrum of chemicals. Similarly, researchers at Merck Sharp & Dohme have evaluated the use of cytotoxicity assays in vitro as an alternative to predicting ocular irritation potential in animals (45). Three different measures of cytotoxicity -leucine incorporation into protein, MTT dye reduction, and neutral red uptake -- were measured in a presumed target cell (corneal epithelial cells from rabbit), as well as in a nontarget cell (V79 Chinese hamster lung fibroblasts). An IC50 value was determined for each endpoint in one or both target cells for a series of 27 commercially available compounds and 56 in-house materials from a variety of chemical classes (carbonitriles, imidazoles, substituted benzenes, aromatic acids, peptides, phenols, esters, etc.). Analysis of the data by Spearman p rank correlation and Pearson's correlation indicated that none of the endpoint-target cell combinations accurately predicted in vivo irritation potential. The MTT dye reduction endpoint gave the best overall correlation coefficient below -0.5; thus the authors of the report concluded that measurement of cytotoxicity is of limited value as an alternative assay for the classes of materials studied.

One further example of the evaluation of in vitro cytotoxicity assays involves a collaborative study between Avon Products, Inc. and the Fund for the Replacement of Animals in Medical Experiments (FRAME). This project, which is currently in progress, will evaluate three different cytotoxicity assays using specific mouse fibroblast cell lines for their ability to predict ocular irritation potential in the rabbit. The three assays are the Avon Neutral Release assay, the FRAME Neutral Red Release Assay, and the Kenacid Blue assay. All rabbit eye irritation data is obtained from Avon's historical records. The test samples constitute 50 personal-care formulations of which ten are hydro-alcoholic and 40 are surfactant-containing. The Neutral Red Release assays incorporate a one minute treatment of target cells with dilutions of test samples to generate a NRR50 value. The NRR50 is the concentration of test substance which causes a 50% loss of pre-loaded Neutral Red from cells, in comparison with untreated (PBS) control cultures. The Kenacid Blue assay incorporates a 72 hours exposure of target cells to dilutions of test samples to generate an ID50. The ID50 is the concentration of test substance which causes a 50% inhibition of cell growth as determined by staining with Kenacid Blue. It is anticipated that this data will help develop the most efficient use of these in vitro systems (K. Renskers, personal communication).

Bacterial Cytotoxicity Assays -- The Microtox Assay (Microbics Corporation, Carlsbad, CA) estimates toxicity by measuring changes in the light output of luminescent bacteria (Photobacterium phosphoreum) after exposure to a potentially toxic material (6). A collaborative study between Avon Products, Inc. and Microbiological Associates, Inc. used the Microtox assay to assess the toxicity of 20 cosmetic products (provided by Avon) which contained 4 - 40% surfactants (41). Twenty-nine individual constituent surfactants and nine preservatives were also tested. Dilutions of the test materials were made in a high salt buffer (the bacteria are marine organisms) and incubated with the bacteria in a temperature controlled photometer. Measurements of light output were made immediately before exposure and five minutes later. A ration of light intensity pre- and post-exposure was plotted against the concentration of test material, and the EC50 (the amount of test material required to reduce light output by 50%) was calculated. Draize scores (24 hour, Avon historical data) were then compared to the EC50 for each test material. In general, the Microtex assay was able to correctly predict the toxicity of the surfactant-containing products at either end of the Draize scale, i.e., for very mild or very irritating materials; it was not very predictive for those products with ocular irritation scores (MAS) between five and 20. Slightly better estimations were found using the individual surfactants. The preservatives were the least well predicted with both overestimations and underestimations of irritancy.

Membrane Permeability Assays -- The transepithelial permeability assay has been evaluated by Johnson & Johnson Consumer Products, Inc. as a model to predict the ocular irritation potential of surfactant-based formulations. The corneal epithelium forms a relatively impermeable barrier to aqueous solutions. On the basis of the presence of tight junctions in monolayers of Madin-Darby canine kidney cells on microporous filters, it has been proposed that this test system may simulate the surface of the cornea (40), and thus changes in the permeability to sodium fluorescein following a 15 minute exposure to dilutions of test formulations were used to calculate EC50's for 28 products. EC50 values were compared with corneal reactions from historical rabbit eye data and gave a correlation coefficient of 0.77. By selection of an EC50 of 2% as cutoff, products could be classified as slight-to-moderate (negative) or moderate-to-severe (positive) irritants. Based on this criteria, the assay had a reported sensitivity of 1.00, a specificity of 0.87, a positive predictive value of 0.87 and a negative predictive value of 1.00.

Chorioallantoic Membrane Assays -- There are a number of variants of the chorioallantoic membrane (CAM) assay, all of which use freshly fertilized hens' eggs as test medium (the chorioallantoic membrane is the vascularized respiratory membrane that surrounds the chicken embryo developing inside the egg). Differences in how the CAM is exposed to test materials and in how the response is evaluated can have a serious input on the correspondence of the CAM with in vivo test results (1). The CAM response to test materials is dose-dependent, thus the frequency of positive responses (vascular hemorrhaging, ghost vessels or capillary injection) is plotted against test material concentration and subjected to probit analysis to obtain an RC50 value (concentration at which 50% of the treated eggs show a positive response). The RC50 value is then used to compare and run test materials and as a basis of comparison with in vivo data or with PC50 values obtained from the standard CAMVA to the standard CAM using 19 surfactants or surfactant-based formulations (2). Two approaches were used, namely predictability (i) with regard to FHSA classification of materials, and (ii) to Draize score. When using the FHSA binary classification, the distribution of compounds into irritant vs. nonirritant classes lends itself to evaluation of the criteria of sensitivity, specificity, and predictive value (12). Both CAM assays were reported to have a sensitivity of 100%, i.e., neither procedure resulted in false-negative responses. The higher specificity of the CAMVA (90%) than the standard CAMVA (50%) is a reflection of the reduction in the number of false positives when CAMVA method is used. Similarly, the predictive value of the CAMVA was higher (90%) than that of the CAM (64%). The CAMVA and the standard CAM were also compared for their ability to rank test materials in the same order that they would be ranked based on Draize scores. Average maximum Draize scores (the average of the highest Draize score obtained at any reading from each of the six rabbits tested) were used to rank samples tested in the in vivo assay; CAMVA and standard CAM results were ranked using RC50 and PC50 values, respectively. For this limited population of samples the CAMVA had a Spearman rank correlation coefficient of 0.7 when compared with in vivo data, compared to 0.6 for the standard CAM (i.e., the CAMVA had a positive correlation with Draize scores and gives a higher positive correlation to the actual Draize score than the standard CAM).

Multi-Endpoint Evaluations -- Mary Kay Cosmetics, Inc., have evaluated eleven in vitro assays (3T3 Neutral Red, 3T3-LDH, Skin2 Full Skin Thickness-MTT, Skin2 Full Skin Thickness-Neutral Red, Testskin-PGE2, CAMVA, Microtox, EYTEX™, and Tetrahymena thermophile motility), using as many as 22 individual surfactants per assay (M. Rozen, personal communication). The surfactants represented the four main surfactant categories (non-ionics, cationics, anionics, and amphoterics) and a broad range of ocular irritation potential (essentially nonirritating to severely irritating). In vivo ocular irritancy was based on historical data provided by vendors. For technical reasons, not all surfactants could be used for the evaluation of each method. The fewest numb evaluated was 14 by EYTEX™ whilst the maximum number was 22 in both the Microtox and 3T3-LDH assays. Based on statistical performance (regression analysis of in vivo vs. in vitro data and sensitivity-specificity analysis), four methods (Skin2-MTT, CAMVA, Microtox, and 3T3-NR) were selected for further evaluation with an additional 15 surfactants. Studies using Sking2-MTT, CAMVA, and Microtox have been completed; studies on 3T3-NR are still in progress. Specificity/sensitivity analysis for the combined first and second sets of surfactants for the three completed assays are 86%/92% (n=34), 81%/80% (n=36), and 90%/56% (n=37), respectively.

Seven in vitro assays were evaluated by the Procter & Gamble Company to determine if any were useful as screening procedures in ocular safety assessment (9). In vivo ocular irritation scores for the materials were obtained from existing rabbit Low Volume Eye Test (LVET) data. Seventeen test materials (chemicals, household cleaners, hand soaps, dishwashing liquids, shampoos, and liquid laundry detergents) were tested in each assay. Significant correlations between in vivo and in vitro data for all the tests except EYTEX™ (silicon microphysiometer, r= -0.87; luminescent bacteria toxicity test, r=-0.91; neutral red assay, r=-0.85; Tetrahymena thermophile motility assay, r=0.78; total protein assay, r=-0.86; EYTEX™ System, r=0.29) were reported. In contrast the irritation classifications provided by a seventh assay, the bovine eye/chorioallantoic membrane assay, did not correspond with the actual in vivo irritancy potential of the test materials. The results of this study suggested it may be possible to classify materials into broad irritancy categories with some of the assays, thereby allowing their use as screens in the ocular safety assessment process (7). The authors conclude that, for the types of materials tested, the silicon microphysiometer and the neutral red assay appear to be the most promising assays likely to be included in a tier testing process (11).

Twelve companies (Abbott Laboratories, Bristol-Myers Squibb, Eli Lily and Co., SmithKline Beecham Corp, S.C. Johnson & Son, Inc., Syntex (USA) Inc., Warner-Lambert Co., Microbiological Associates, Inc., G.D. Searle & Co., The Upjohn Co., Merck & Co., and Hoffmann-LaRoche, Inc.) are participating in a collaborative study to evaluate currently available alternative models with a set of generally unrelated compounds (24). Several tests have been selected for evaluation: the Bovine Corneal Opacity/Permeability (BCOP), EYTEX™ (In Vitro International, Inc.), Neutral Red Uptake Bioassay in Normal Human Epidermal Keratinocytes (Clonetics Corporation), MTT Assay using TESTSKIN™ Living Dermal Equivalent (Organogenesis Inc.), Microtox (Microbics Corp.), and the Chorioallantoic Membrane Vascular Assay (CAMVA), as well as a computer-based structure-activity relationship model (TOPKAT, Health Designs, Inc.). Thirty-seven test compounds were selected on the basis of historical in vivo irritation potential, pH, and solubility from among synthetic intermediates isolated during the manufacture of pharmaceutical and chemical products. To assure technical consistency and evaluation, all tests were conducted at a single location (Merck Sharp and Dohme Laboratories in France for the BCOP test, and Microbiological Associates, Inc. for the remaining tests) using coded samples. The results were pooled and analyzed relative to physical properties and in vivo activity. Among the assays evaluated, the BCOP test showed the most promise in predicting the irritation class of the test compounds, although investigators concluded that the CAMVA test and SAR analysis using TOPKAT may have application in predicting nonirritancy, subject to some limitations.

Dermal Irritation Tests

Although less information is available on ongoing evaluations of  in vitro dermal irritation tests, the results of several studies utilizing a variety of endpoints have recently been reported.

Physical/Chemical Models -- In addition to the study on EYTEX™ discussed earlier, Avon Cosmetics, Inc. and In Vitro International, Inc. also report that a total of 133 cosmetic and personal care product formulations and raw ingredients were evaluated with the SKINTEX™ System (Upright Membrane Assay [UMA] and Alkaline Membrane Assay [AMA], where appropriate) to determine the potential to predict in vivo responses from rabbit skin irritation studies and human patch tests (39). Test samples included raw ingredients and products from 14 different classes, and represented a range of  in vitro dermal irritancy from nonirritant to moderately irritant. SKINTEX™ results were compared with rabbit dermal irritation data (single occlusive patch) obtained from Avon historical records. The authors of the report claim that a positive correlation between SKINTEX™ results and the in vivo assay was demonstrated by an overall concordance rate of 83.5%. Assay error was reported as 16.5%, of which 7.5% was due to an overestimation of sample irritancy (false positives), and 9.0% was attributed to underestimation (false negatives). SKINTEX™ data for a subset of 68 products and raw ingredients were additionally compared with human dermal irritation data (single insult patch also obtained from Avon historical records). A positive correlation between SKINTEX™ results and the human assay was demonstrated by an overall concordance rate of 88.2%. Assay error (11.8%) was attributed to an overestimation of sample irritancy (false positives).

Mammalian Cytotoxicity Assays -- The Procter & Gamble Company has evaluated in vitro models for prediction of human skin irritation (42, 43). Normal human epidermal keratinocyte (NHEK) cultures (Clonetics, San Diego, CA) were treated with dilutions of product ingredients and formulations, and the extent of cell damage was assessed by uptake of the vital dye neutral red. Cell damage correlated with human skin patch data for ingredient chemicals, except for acids and alkalies, however, NHEK responses did not correlate with in vivo skin irritation for surfactant-contaminating product formulations. As a result, Osborne and Perkins report that the NHEK assay has been incorporated into a tiered testing process for preclinical skin irritation assessments for chemicals that are compatible with the assay. Elements of the assessment process include (i) analysis of historic data on the test substance of interest and similar substances; (ii) characterization of physico chemical characteristics that may indicate irritancy potential, such as buffering capacity; and (iii) determination of the effects of compatible test substances in the in vitro assay. The NHEK assay allows determination of the similarity of new materials to benchmark standards and aids in the selection of doses for clinical studies.

Reconstituted Human Skin Models -- Due to the limited ability of the NHEK assay in their hands to predict the irritancy of acids, alkalies, product formulations, and undiluted aqueous incompatible substances, the Procter & Gamble Co. have also evaluated human skin equivalent cultures. These cultures contain key structural elements of skin, including dermal fibroblasts in a collagen matrix of stratified epidermal keratinocytes. The advantage of these cultures is that they contain an air-interface, so that test substances can be applied topically in a manner that mimics in vivo exposure. A battery of endpoints was developed to measure responses to prototyped ingredients and formulations in skin equivalent cultures grown on a nylon mesh (Skin2 from Advanced Tissue Sciences, La Jolla, CA). The endpoints measure cytotoxicity (neutral red and MTT vital dye staining and lactate dehydrogenase release) and inflammatory mediator (prostaglandin E2) release. Initial experiments with prototype surfactants, acids, and alkalies indicated a promising correlation between responses of the Skin2 model and in vivo human skin irritation. Based on this approach, a tissue equivalent assay for aiding ocular irritation assessments for undiluted aqueous incompatible substances, such as powders and solids, was developed (44). This method measures the rate of cell damage produced by substances applied topically in partially stratified Skin2 cultures. These cultures have stromal and epithelial components that act as in vitro counterparts to eye structures (cornea and conjectiva) that are important targets in ocular irritation. The tissue equivalent assay allowed placement of 75 test substances into broad irritancy categories of innocuous-slight, slight-moderate, and moderate-severe, and is being used in a tiered testing process for ocular irritancy assessments.

A second example of a reconstituted human skin model is the living skin equivalent (Organogenesis Inc., Cambridge, MA). Living skin equivalents (LSE) are co-cultures of human dermal fibroblasts in a collagen lattice with overlying layers of stratified human epidermal keratinocytes. LSE were used to evaluate the relative dermal irritation potential of selected chemicals (including organics, alcohols, surfactants, and water-insoluble materials) using a variety of endpoints (25). Time and dose-dependent changes in cellular viability, the release of the pro-inflammatory mediators, (prostaglandin E2 and the interleukin-1-alpha) and tritiated water penetration through the LSE were evaluated. Irritants from different chemical causes were rank ordered using several criteria including the effective concentration which inhibited 50% (EC50) of thiazolyl blue conversion (MTT) at a constant exposure time, and the time required to inhibit 50% MTT conversion (ET50) at a set concentration. The extent to which the barrier function of the LSE was damaged, as assessed by water penetration through the LSE, was also used to grade the irritation potential of different test chemicals. The correlation between in vitro rankings and in vivo dermal irritation scores was dependent on chemical dose and the in vitro assay method. In the case of anionic surfactants, Gay et al report good agreement (based upon correlation coefficients) between ET50 scores, rates of water penetration through the LSE, and the ability of these irritants to induce erythema and dryness in human skin (25).

Similarly, the evaluation of the skin irritation potential of petroleum-based compounds in a reconstituted human skin model has also been reported (36). In a collaborative study between ARCO and Thomas J. Stephens & Associates, Inc., seven petroleum-based test materials were evaluated for their skin irritation potential using full thickness reconstituted human skin models (Skin2 Model 1300 from Advanced Tissue Sciences, Inc., and Living Skin Equivalent from Organogenesis, Inc.). Because of their high vapor pressure and low solubility in tissue culture media the evaluation of volatile test materials in in vitro test systems can be problematic, hence test materials were applied to (undiluted) the epidermal side of the tissues and incubated in polyethylene bags designed to minimize cross-contamination and the loss of volatile constituents. Endpoints were determined as MTT reduction, and the presence of LDH, PGE2, and IL-1a in spent media from treated and untreated tissues. Data were correlated with Draize Primary Dermal Irritation Index scores using criteria (specificity/sensitivity) described by Cooper et al (12). Results showed that all endpoints measured (with the exception of IL-1A in Skin2 Model 1300) approximated the skin irritation potential of the test materials. The authors of the report note that further refinement of the dosing chamber may result in a novel method for evaluating volatile test materials in tissue culture.


The obvious conclusion which may be drawn from this limited review of current U.S. "validation" activities is that a broad diversity of approach (both with regard to selection of test materials and test methods, and to statistical methods of evaluating in vitro test performance) has been adopted. Adoption of appropriate criteria for assessing in vitro test performance appears destined to be particularly controversial (18). Development and evaluation of in vitro eye irritation tests continues to be the focus for most investigators, although there now appears to be significant interest in developing in vitro dermal irritation tests. The approach to the evaluation of in vitro dermal irritation tests appears to be mimicking the historical approach which has evolved for in vitro ocular tests.

In September 1988, representatives of government, industry, academia, and the general public gathered to participate in the Joint Government-Industry Workshop on Progress Towards Non-Animal Alternatives to the Draize Test. The workshop was sponsored by CPSC, EPA, FDA, SDA, CTFA, CSMA, and CMA. The objective of the workshop was to provide a forum for interested parties to present data and discuss policy regarding the development and validation of alternatives to in vivo ocular irritation testing procedures. Several important areas of consensus were developed. In particular, it was agreed that: (i) in the absence of an extensive quantitative human database, rabbit ocular irritation data should be used to confirm irritancy classification; and (ii) development and validation of tests or combinations of tests for specific classifications of chemicals would lead to a more rapid acceptance and implementation of in vitro tests than would occur if universal validation of in vitro tests were required. Both of these consensus findings were incorporated in the two industry-sponsored programs (SDA, CTFA), conceived and initiated in the latter part of the 1980s, and are characteristic of all programs described in this review.

Chemicals selected for inclusion in any validation or evaluation study must be well-documented with respect to their anticipated in vivo toxicity. If a particular test chemical or formulation is to be included, then (whenever possible) the toxicity database for that chemical must be compatible with the other chemicals or formulations selected since complications in the interpretation of results may arise if toxicity data are not consistent. To satisfy such criteria, it may be necessary to develop independently a comprehensive and compatible database for test materials selected for any scientifically rigorous evaluation/validation study (19). Ocular irritancy has traditionally been evaluated using the method described by Draize (14); consequently many studies comprise an in vivo phase (Draize test) and an in vitro phase (involving a number of alternative tests), and all test materials undergo parallel testing (30). In contrast, other studies have relied upon the availability of historical data where such information has been well-documented (see several examples described in this review). One notable exception to the practice of relying exclusively on correlation with Draize data, is CTFA's evaluation of in vitro test performance based upon comparison with the Low Volume Eye Test (LVET). The evaluation will also result in a useful comparison of Draize test and LVET data on the same set of test materials. The LVET has been claimed to better predict the human response to ocular irritation than the traditional Draize method, and to be more humane (34, 21, 22). Correlation of in vitro data with the LVET has also been described by Bruner et al (9).

Although CAAT/ERGATT have recommended that tests should be validated with a diverse selection of chemicals to which the test might be applied, and that such a selection should include chemicals of different structural groupings, physical forms, physiochemical properties, mechanisms of action and spectra of broad biological activities (3), programs for the evaluation of in vitro alternatives to the eye irritation tests have most often focused on series of structurally or functionally related compounds. As was noted at the Joint Government-Industry Workshop in 1988, development and validation of test methods for specific classifications of materials would, in all likelihood, lead to a more rapid acceptance and implementation of in vitro tests than would occur if universal validation were required (for example, the detergent industry must evaluate granular, alkaline solids while these substances do not concern the paint or cosmetics industries). It was predicted that each industry would identify unique aspects about the spectrum of chemicals and products of specific and particular interest, and that such an approach would lead to a unique validation program for that industry. Thus, the limited diversity of test materials selected for inclusion in many of the studies described herein is acknowledged, but in each case is considered appropriate to the intended use of the tests. These studies have generally been used to correlate ranking of irritation potential in vitro to ranking of irritation potential in vivo. Based upon some of the disparate results reported here (i.e., from studies using nominally similar assays to evaluate different classes of test materials, but with a greater or lesser degree of success) it is apparent that further studies may be required to evaluate in vitro tests for other types of materials and for other applications. It is not yet possible to use currently available tests to categorize the irritation potential of individual compounds without regard to product type or chemical class.

Although it has been implied that there must be a strong mechanistic basis for the implementation of alternative tests even when a strong correlation between measured in vitro activity and known in vivo toxicity has been demonstrated (17), the Report and Recommendations of the CAAT/ERGATT Workshop on the Validation of Toxicity Test Procedures (3) recognizes that the acceptable degree of mechanistic similarity of an in vitro procedure to an existing animal model is conditional upon the ultimate use ("specific decision making process") for which it is deemed appropriate, namely:

Screening test: mechanistic similarity is not necessary if the empirical correlation/predictivity criterion is met.

Adjunct test: mechanistic relatedness is desired, but not required.

Replacement test: mechanistic similarity for replacement tests is generally required, but exceptions could exist.

Although (in the long term) the development and validation of mechanism-based tests are of the utmost importance (8), under certain circumstances in vitro test methods are being used as supplementary procedures for purposes of safety substantiation, particularly in the cosmetic/personal care industry. It is important to realize that safety decisions are not based solely on the outcome of single toxicity tests, whether they be in vivo or in vitro (23). Safety tests are undertaken only after characterization of physical/chemical properties, exhaustive review of existing data, and only when outstanding safety questions still remain. In the cosmetic and personal care industry, where new products often constitute minor formulation changes based upon an existing product (for which adequate safety data may already exist), such outstanding questions may sometimes be adequately addressed by comparing a new formulation with an existing formulation based on performance in an in vitro test (so-called adjunct testing). Under such conditions it may not be necessary to obtain an absolute determination of in vivo toxicity to assure safety of the product, and thus utilization of an in vitro assay with a demonstrated correlation with in vivo response may be sufficient. Presumably an in vitro test with demonstrable mechanistic similarity is more likely to correlate better with an in vivo assay, but if a sufficient and reliable degree of correlation exists between in vivo and in vitro data, then lack of such an attribute may not necessarily preclude its use (albeit under limited and well-defined circumstances) (15). Similarly, it is difficult to envisage why (whilst desirable) a screening test should necessarily have demonstrable mechanistic similarity to events in vivo. If sufficient degree of correlation exists, lack of mechanistic similarity may not necessarily preclude its use as a screen since (by definition) a screen is merely a preliminary to more definitive investigations. Clearly however, in situations where little or no information is available on an ingredient or formulation (and where a more definitive toxicological approach is needed), proven mechanistic similarity becomes significant, if not paramount.

In summary, a diverse approach to the eventual validation of alternative test methods (as exemplified by the development of in vitro eye irritation test methodology) has thus far served the science of in vitro toxicology well. However, given the extent of information on test performance now becoming available, perhaps the time has come to assess what has already been achieved in the field with a view to deciding future directions and an approach to "true" validation. Significant decisions must be reached with regard to the degree of acceptance of "correlative"f tests (i.e., those with no apparent mechanistic relevance), appropriate use of in vivo data (e.g., maximum average Draize score [MAS] or reference classification [FHSA irritant/nonirritant]), and adoption of appropriate statistical methodology. The challenge is to incorporate the best elements of past diversity into a future, unified, and concerted approach to validation.


  1. Bagley, D.M., Rizvi, P.Y., Kong, B.M., & De Salva, S.J. (1991). Factors affecting use of the hen's egg chorioallantoic membrane as a model for predicting eye irritation potential: I. Journal of Toxicology Cutaneous & Ocular Toxicology 10: 95-104.
  2. Bagley, D.M., Rizvi, P.Y., Kong, B.M., & De Salva, S.J. (1991). Evaluation of the vascular components of the chorioallantoic membrane assay as a model for eye irritation potential: II. Journal of Toxicology Cutaneous & Ocular Toxicology 10: 105-113.
  3. Balls, N., Blaauboer, B., Brusnick, D., Frazier, J., Lamb, D., Pemberton, M., Reinhardt, C., Roberfroid, M., Rosenkranz, H., Schmid, B., Spielmann, H., Stammat, A-L., & Walum, E. (1990). Report and recommendations of the CAAT/ERGATT Workshop on the validation of toxicity test procedures. ATLA 18: 313-338.
  4. Booman, K.A., Cascieri, T.M., Demetrulias, J., Driedger, A., Giffith, J.F., Grochoski, G.T., Kong, B., McCormick, W.C. III., North-Root, H., Rozen, M.G. & Sedlak, R.I. (1988a). In vitro methods for evaluating eye irritancy of cleaning products Phase I: preliminary assessment. Journal of Toxicology Cutaneous & Ocular Toxicology 7(3): 173-185.
  5. Booman, K.A., De Prospo, J., Demetrulias, J., Driedger, A., Griffith, J.F., Grochoski, G.T., Kong, B., McCormick, W.C. III, North-Root, H., Rozen, M.G., & Sedlak, R.I. (1988b). The SDA alternatives program: comparison of in vitro data with Draize test data. Journal of Toxicology Cutaneous & Ocular Toxicology 8(1): 35-49.
  6. Bulich, A.A. (1986). Bioluminescence assays. In Toxicity Testing Using Micro-Organisms, Vol. 1, (eds. G. Bitton & B.J. Dutka) pp. 57-74. Boca Raton: CRC Press, Inc.
  7. Bruner, L.H. (1992a). Alternatives to the use of animals in household product and cosmetic testing. Journal of the American Veterinary Medical Association, 200: 669-673.
  8. Bruner, L.H. (1992b). Ocular irritation. In In Vitro Toxicity Testing, Applications to Safety Evaluation (ed. J.M. Frazier) pp. 149-190. New York: Marcel Dekker, Inc.
  9. Bruner, L.H., Kain, D.J., Roberts, D.A., & Parker, R.D. (1991a). Evaluation of seven in vitro alternatives for ocular safety testing. Fundamental and Applied Toxicology 17: 136-149.
  10. Bruner, L.H., Shadduck, J., & Essex-Sorlie, D. (1991b). Alternative methods for assessing the effects of chemicals in the eye. In Dermal and Ocular Toxicology, Fundamentals and Methods (ed. D.W. Hobson) pp. 585-606. Boca Raton, FL: CRC Press, Inc.
  11. Bruner, L.H., Miller, K.R., Owicki, J.C., Parce, J.W., & Muir, V.C. (1991c). Testing ocular irritancy in vitro with the silicon microphysiometer. Toxicology In Vitro 5 (4): 277-284.
  12. Cooper, J.A., Saracci, R., & Cole, P. (1979). Describing the validity of carcinogen screening tests. British Journal of Cancer 39: 87.
  13. Decker, D. & Harper, R., Evaluation of the EYTEX™ System for use as a predictor of ocular irritancy of shampoos. In In Vitro Toxicology: 10th Anniversary of CAAT (Alternatives Methods in Toxicology, Vol. 9, ed. A.M. Goldberg) (In Press). New York: Mary Ann Leibert, Inc.
  14. Draize, J.H., Woodard, G. & Calvery, H.O. (1944). Methods for the study of irritation and toxicity of substances applied topically to the skin and mucous membranes. Journal of Pharmacology and Experimental Therapy 82: 377-390.
  15. FDA (1989). U.S. Food and Drug Administration Statement to the Maryland Governor's Task Force to Study Animal Testing (April 17, 1989).
  16. Feder, P.I., Lordo, R.A., DiPasquale, L.C., Bagley, D.M., Chudkowski, M., Demetrulias, J.L., Hintze, K.L., Marenus, K.D., Pape, W.J.W., Roddy, M.T., Schnetzinger, R., Silbber, P.M., Teal, J.J., Weise, S.L., & Gettings, S.D. (1991). The CTFA Evaluation of Alternatives Program. An evaluation of in vitro alternatives to the Drazie primary eye irritation tests. (Phase I) Hydroalcoholic formulations; (Part 1) Statistical methods. In Vitro Toxicology 4 (4): 231-246.
  17. Flint, O.P. (1990). In vitro toxicity testing: purpose, validation and strategy. ATLA 18, 11-18.
  18. Flint, O.P. (1990). In vitro test validation: A house built on sand. ATLA 20: 196-198.
  19. Frazier, J.M. (1990). Scientific criteria for validation of in vitro toxicity tests. OECD Environment Monographs, No. 36.
  20. Frazier, J.M., Gad, S.C., Goldberg, A.M. & McCulley, J.P. (1987). A critical evaluation of alterntaives to acute ocular irritation testing. In Alternative Methods in Toxicology, Vol. 4. New York: Mary Ann Liebert, Inc.
  21. Freeberg, F.E., Griffith, J.F., Bruce, R.D. & Bay, P.H.S. (184). Correlation of animal test methods with human experience for household products. Journal of Toxicology Cutaneous & Ocular Toxicology 3 (1): 53-64.
  22. Freeberg, F.E., Nixon, G.A., Reer, P.J., Weaver, J.E., Bruce, R.D., Griffith, J.F. & Sanders, L.W., III (1986). Human and rabbit eye responses to chemical insult. Fundaments and Applications in Toxicology 7: 626-634.
  23. Gad, S.C. (1991). Use of ocular and dermatotoxicologic test data in product development and in the assessment of chemical hazrads. In Dermal and Ocular Toxicology, Fundamentals and Methods (ed. D.W. Hobson) pp. 369-404. Boca Raton: CRC Press, Inc.
  24. Galer, D.M., Curren, R., Gad, S.C., Gautheron, P., Leong, B., Miller, K., Sargent, E., Sina, J., & Sussman, R.G. A 10-company collaborative evaluation of alternatives to the eye irritation test using chemical intermediates. In In Vitro Toxicology: 10th Anniversary of CAAT (Alternative Methods in Toxicology, Vol. 9, ed. A.M. Goldberg) (In Press). New York: Mary Ann Liebert, Inc.
  25. Gay, R., Swiderek, M., Nelson, D., & Ernesti, A. The Living Skin Equivalent as a model in vitro for ranking the toxic potential of dermal irritants. Toxic. In Vitro (In Press).
  26. Gettings, S.D. (1991). The current status of in vitro test validation (evaluation) in the United States. ATLA 19: 432-436.
  27. Gettings, S.D. (1991). The CTFA Evaluation of Alternatives Program: Objectives and Overview. In Proceddings of the 2nd CTFA Ocular Safety Testing Workshop: Evaluation of In Vitro Alternatives (eds. S.D. Gettings & G.N. McEwen), pp. 2-6. Washington: The Cosmetic, Toiletry, and Fragrance Asoociation.
  28. Gettings, S.D. & McEwen, G.N. (1990). Development of potential alternatives to the Draize eye test: The CTFA Evaluation of Alternatives Program. ATLA 17: 317-324.
  29. Gettings, S.D., DiPasqulae, L.C., Bagley, D.M., Chudkowski, M., Demtrulias, J.L., Feder, P.I., Hintze, K.L.., Marenus, K.D., Pape, W.J.W., Roddy, M., Schnetzinger, R., Silber, P.M., Teal, J.J., & Weise, S.L. (1990). The CTFA Evaluation of Alternatives Program. An evaluation of in vitro alternatives to the Drazie primary eye irritation test. (Phase I) Hydroalcoholic formulations; a preliminary communication. In Vitro Toxicology 3: 293-302.
  30. Gettings, S.D. Bagley, D.M., Demetrulias, J.L., DiPasquale, L.C., Hintze, K.L., Rozen, M.G., Teal, J.J., Weise, S.L., Chudkowski, M., Marenus, K.D., Pape, W.J.W., Roddy, M.T., Schnetzinger, R., Silber, P.M. Glaza, S.M., & Kurtz, P.J. (1991). The CTFA Evaulation of Alternatives Program. An evaluation of in vitro alternatives to the Draize priimary eye irritaion test. (Phase I) Hydroalcoholic formulations; (Part 2) Data analysis and biological significance. In Vitro Toxicology 4 (4): 247-288.
  31. Gettings, S.D., Bagley, D.M., Chudkowski, M., Demetrulias, J.L., DiPasquale, L.C., Galli, C.L., Gay, R., Hintze, K.L., Janus, J., Marenus, K.D., Muscatiello, M.J., Pape, W.J.W., Renskers, K.J., Roddy, M.T., & Schnetzinger, R. (1992). Development of potential alternatives to the Draize Eye Test: The CTFA Evaluation of Alternatives Program. Phase II: Review of materials and methods. ATLA 20: 164-171.
  32. Gettings, S.D., Demetrulias, J.L., Feder, P.I., Glaza, S.M., Lordo, R.A., & Hintze, K.L. A comparison of in vitro assay, Low Volume Eye Test (LVET), and Draize Primary Eye Irritation Test data. (Phase I) Hydroalcoholic formulations (In preparation).
  33. Gordon, V.C. & Bergman, H.C. (1987). EYTEX™, an in vitro method for evaluation of ocular irritancy. In In Vitro Toxicology -- Approaches to Validation (Alternative Methods in Toxicity, Vol. 5., ed. A.M. Goldberg) p.p 293-302. New York: Mary Ann Liebert, Inc.
  34. Griffith, J.F., Nixon, G.A., Bruce, R.D., Reer, P.J., & Bannan, E.A. (1980). Dose-response studies with chemical irritants in the albino rabbit eye as a basis for selecting optimum testing conditions for predicting hazard to the human eye. Toxicology and Applied Pharmacology 55: 501-513.
  35. Kennah, H.E., Albulesum, D., Higret, S. & Barrow, C.S. (1989). A critical evaluation of predicting ocular irritancy potential from an in vitro cytotoxicity assay. Fundaments and Applications in Toxicology 12: 281-290.
  36. Koschier, F.J., Roth, R.N., Stephens, T.J., Spence, E.T., Kasbarian, H. & Duke, M.A. Evaluation of the skin irritation potential of petroleum based compounds in a reconstituted human skin model. In Vitro Toxicology: 10th Anniversary of CAAT (Alternative Methods in Toxicology, Vol. 9, ed., A.M. Goldberg) (In Press). New York: Mary Ann Liebert, Inc.
  37. Kruszewski, F.H. (1991). The EYTEX™ System - A review of current methodology. In Proceedings of the 2nd CTFA Ocular Safety Testing Workshop: Evaluation of In Vitro Alternatives (eds. S.D. Gettings & G.N. McEwen) pp. 80-93. Washington, D.C.: The Cosmetic, Toiletry, and Fragrance Association.
  38. Krsuzewski, F.H., Hearn, L.H., Smith, K.T., Teal, J.J., Gordon, V.C. & Dickens, M.S. (1992). Application of the EYTEX™ system to the evaluation of cosmetic products and their ingredients. ATLA 20: 146-163.
  39. Kruszewski, F.H., Dickens, M.S., Gordon, V.C., & Renskers, K.J. Application of the SKINTEX™ System to the evaluation of cosmetic products. In Vitro Toxicology: 10th Anniversary of CAAT (Alternative Methods in Toxicology, Vol. 9, ed., A.M. Goldberg) (In Press). New York: Mary Ann Liebert, Inc.
  40. Martin, K.M. & Stott, C.W. The transepithelial permeability assay as in vitro assay for predicting ocular irritation of surfactant formulations. In In Vitro Toxicology: 10th Anniversary of CAAT (Alternative Methods in Toxicology, Vol. 9, ed., A.M. Goldberg) (In Press). New York: Mary Ann Liebert, Inc.
  41. Miller, K.A., Stahl, R.W., Marchesani, F.S., Harbell, J.W., Wallace, K.A., Kruszewski, F.H., Renskers, K.J. & Curren, R.D. A toxicity evaluation by the microtox assay of twenty surfactant-containing products and their constituent surfactants and preservatives. In In Vitro Toxicology: 10th Anniversary of CAAT (Alternative Methods in Toxicology, Vol. 9, ed., A.M. Goldberg) (In Press). New York: Mary Ann Liebert, Inc.
  42. Osborne, R. & Perkins, M.A. (1991). Evaluation of human skin cell cultures for in vitro irritancy testing. In Vitro Toxicology: Mechanisms and New Technology (Alternative Methods in Toxicology, Vol. 8, ed., A.M. Goldberg) pp. 317-324. New York: Mary Ann Liebert, Inc.
  43. Osborne, R. & Perkins, M.A. (1992). A strategy for development of alternative methods for skin irritation testing. Food and Chemical Toxicology (In Press).
  44. Perkins, M.A., Roberts, D.A. & Osborne, R. (1992). Human skin cell cultures for in vitro skin and eye irritancy assessments of neat test materials. The Toxicologist 12: 296.
  45. Sina, J.F., Ward, G.J., Laszek, M.A., & Gautheron, P.D. (1992). Assessment of cytotoxicity assays as predictors of ocular irritation of pharamceuticals. Fundamental Applications in Toxicolgy 18: 515-521.



Prepared by:Jack M. Lipman, PhD
Hoffmann-La Roche, Inc.
340 Kingland Street
Nutley, NJ 07110

Virtually all pharmaceutical companies involved in drug development have or are planning resource commitment to in vitro toxicity testing. Two of the major goals of this effort are: first, to improve our understanding of the mechanisms of drug induced toxicity and second to achieve the 3Rs (reduction, refinement, and replacement) of animal usage. Clearly the need to evaluate all test systems for both intra- and interlaboratory variation is a key step, only superseded in importance by the development of a specific assay system itself.

This report will describe two such initiatives currently underway within the pharmaceutical industry.

The first was sponsored by the Pharmaceutical Manufacturing Associations: In Vitro Toxicology Task Force and consists of participants from over 13 pharmaceutical companies.

This study is a collaborative effort to evaluate an in vitro muscle irritation assay. The original in vivo test measured the muscle irritation potential of parenterally administered products by the injection of test solution into the sacrospinalis muscle of the rabbit. Irritation was determined by measurement of creatine phosphokinase levels in the serum and gross/microscopic pathology of the injection site.

The goal of this study is to evaluate the assay's ability to predict potential human irritancy, develop an interlaboratory database to evaluate the assay's utility as an adjunct to in vivo testing and assess inter/intralaboratory variability of the assay.

The assay design was to use protocols and procedural information in such a manner as to comply with GLP standards for the conduct and reporting of toxicology studies. The assay system used L-6 rat skeletal myoblasts and all compounds were tested in a blind manner. After the cells were plated (24 hour) they were exposed to the compounds (n=12) as supplied in their prepackaged clinical formulations in serum free media at six concentrations. Cell extracts were made and CPK, ALT, and LDH enzyme levels measured by each of the participating laboratories. These data next underwent statistical analysis and are currently being finalized for presentation and publication.

The second validation study was begun by the In Vitro Alternative Tests Task Force which consists of representatives from over 13 different companies. This project has set as its goal the evaluation of in vitro alternatives to the eye test. For uniformity, the actual testing was conducted at a contract laboratory.

The design and potential success for this program rests upon the criteria for compound choice. It should be noted that a constant flaw in the attempt to validate tests is the inability to correlate data with in vivo data, thereby showing intra- and interreliability, but not showing if the test will in fact replace either the animal model, the actual human experience or both. This study attempts to overcome this by using the following criteria for compound selection:

  1. Each participant will supply approximately 3-6 compounds, no more than two can be in the same structure/synthetic series.
  2. Compounds must have a pH between 3 and 11.
  3. Compounds that are insoluble in water will be limited to one participant. Compounds which form good suspensions will be considered. A compound is considered insoluble if solubility is below 10 mg/ml.
  4. Raw Draize scores must be available for compounds tested.
  5. Compounds will be single entity chemicals, not mixtures, and not pharmaceutically active.
  6. Highly toxic compounds (LD50 < 50 mg/kg) are not included.
  7. Each participant will not submit more than two compounds in each irritation category if testing five or fewer compounds.
  8. Each irritation category (nonirritant, mild, moderate, severe) will be represented by not less than 10 compounds in the complete set.

Even though the above criteria sound simple, it should be realized that it is extremely difficult to find compounds that meet all of these requirements.

For this validation study the test chosen, whenever possible, represented commercially available systems. This allows: a) the assay to be rapidly introduced into new laboratories without adaptation from literature sources, b) for a consistent source of assay components. The test systems chosen included:

  1. The Bovine Corneal Opacity/Fluorescein Penetration Assay
  2. Neutral Red Cytotoxicity (Clonectics Corp.)
  3. EYTEX™ (In Vitro International)
  4. Testskin MTT Assay (Organogenesis, Inc.)
  5. CAM-VA
  6. Microtox (Microbics Corporation)
  7. TOPKAT (Computer Analysis of Structure-Activity Relationships)

Upon completion of the analysis of the compounds, a task force will be formed to tabulate and analyze the results for presentation and publication in a peer-reviewed journal. It is hoped that this project will be completed by the fall of 1992.

The pharmaceutical industry has a long standing commitment to developing and validating in vitro tests to be used as both an adjunct to in vivo testing and with sufficient validation possibly replace the  vivo test completely. Initiatives as described above are important steps in achieving these goals.



Prepared by:Bas J. Blaauboer, PhD
Research Institute of Toxicology
University of Utrecht
P.O. Box 80.176, NL-3508TD, Utrecht
The Netherlands

The need for a reduction in the use of animals in toxicological risk assessment has started a number of developments directed towards changes in strategies for toxicity testing. Before these alternative test methods can be incorporated into such strategies, the tests must be validated. This process may consist of a number of steps. After the development of a test system, the performance of the method needs to be evaluated in an intralaboratory assessment, the reproducibility must be determined in an interlaboratory assessment and reference data in a database on the system must be collected. After evaluation of all data the test can be considered for acceptance as part of a risk evaluation strategy (1, 2).

In Europe, alternative test methods have been validated or are undergoing this process. These activities comprise methods for the replacement of the LD50 test or other acute toxicity testing, the Draize eye irritancy test, tests on teratology, and organ-specific toxicity. For a number of these validation programs, it is now possible to evaluate the success of the validation strategy and to assess the degree of acceptance (3; see Table 1).

Table 1: Validation Studies in Europe (1990)





Acute toxicityBTSValidated4, 5, 7
Acute toxicityBGAValidated6
Cytotoxicity vs acute toxicityFRAMEValidated8, 9, 11
Cytotoxicity vs acute toxicityItalyValidated10
Cytotoxicity vs acute toxicityMEICIn progress12, 13
Eye irritancyBGAIn progress14
Eye irritancyOPALIn progress15
Eye irritancyRCC-NOTOXAssessed in laboratory of origin16
Developmental toxicityInternational studyIn progress17
Hepato(cyto)-toxicityFranceIn progress18

During the last few years, a number of alternative test methods have been the subject of validation studies or are now undergoing this process. In Europe, these activities can be regrouped in various fields of interest. Some of the more important activities will be mentioned here, without having the pretention of giving a complete overview of all validation studies and related programs in Europe. The aim is to discuss the chosen strategies and the implications for the refinement, reduction or replacement of animal experiments, rather than giving the results of the individual validation studies. This is the prerogative of the scientists responsible for these programs.

One area is the replacement of the classical LD50 test for classification purposes. Such a scheme was the Fixed Dose Procedure, proposed by the British Toxicology Society (BTS; 4). This method consists of a stepwise testing of a limited number of dose levels which are related to the limits used in classification schemes for chemicals. If the result on one dose level gives the required information, no further testing is needed. Another feature of the method is that emphasis is laid on signs of toxicity rather than on death (5). The procedure was validated in a study in more than 30 laboratories in 11 countries and was sponsored by the Commission of the European Communities and carried out under the patronage of the OECD. The result was that over 80% of the chemicals classified equivalently to their classification on the basis of the classical LD50 test, with an average of only 15 animals used per chemical. In Germany another scheme was developed, in which only about five animals per compound were used, but the endpoint is still the exposure related death of the animal (6). The methods were discussed at a symposium held in Brussels in 1989 (7) and it was accepted by regulatory authorities within the EC. Acceptance will also be considered by the OECD. There may be some debate on the size of the dosage groups in the FDP. In the validation program following the BTS proposal, this size was ten animals per group. If this number can be reduced to five, acceptance of this procedure will lead to a considerable reduction and also to a refinement of animal experiments. Considerable efforts were also made on the validation of in vitro methods to predict acute toxicity of chemicals. In the United Kingdom, FRAME (Fund for the Replacement of Animals in Medical Experiments) compared the cytotoxicity of a great number of chemicals with their in vivo toxicity (8, 9). Similar programs were also carried out in Italy (10). The conclusion of these studies was that it is not possible to predict in vivo acute toxicity on the basis of results of one particular in vitro method. However, for many types of compounds the correlation between in vitro toxicity determined in a number of cytotoxicity tests and the in vivo lethal dose was good. In the case that the results of cytotoxicity measurements for groups of chemicals are ranked according to in vivo toxicity data, it will be possible to compare the results of unknown compounds with a number of reference compounds (11).

A different approach was chosen by the Scandinavian Society for Cell Toxicity in the MEIC program (Multicenter Evaluation of In Vitro Cytotoxicity Testing). In this program participating laboratories apply their own test systems to a set of chemicals which were selected on the basis of the availability of data on human toxicity, including lethal or toxic doses and related blood concentrations, target organs, and biokinetic parameters (12). An interlaboratory assessment of a method or a battery of methods will be a later part of the program. However, this initiative can result in the recommendation of batteries of standard tests that can give specific answers to certain types of toxic reactions in man (13).

Another area in which several validation programs were or are being carried out is the in vitro measurement of eye irritancy as a replacement for the Draize test. In Germany an assessment of the HET-CAM test (hen's egg test at the chorioallantoic membrane) in combination with cytotoxicity determinations is being carried out in twelve laboratories (14), and similar studies are in progress in France (15). In The Netherlands the CAM test, combined with a test on isolated bovine eyes, was used in an interlaboratory assessment using over 145 chemicals, and this combination proved to be a good basis for classification of these chemicals (16). Elsewhere, validation of a system consisting of isolated rabbit eyes is proceeding. Additional information can be obtained with the incorporation in these approaches of the EYTEX™ system in which the effects on protein modification is measured. The conclusion from these efforts seems to be that in the future the classical Draize test will be employed only to assess the absence of an effect for compounds that had no effect in a number of in vitro test systems. For chemicals that are strongly positive in one or more of the alternative systems it can be considered pointless to ask for an additional in vivo test. This strategy will lead to considerable reduction in animal suffering; here refinement, reduction, and in many cases replacement are all involved.

Several European laboratories are involved in a validation program on the micromass assay for the determination program of teratogenic effects of compounds, developed by Flint and Orton (17). This test employs primary cultures of rat embryo midbrain and limb bud cells to determine the effects on cytotoxicity and organ-specific differentiation markers. The study is being carried out in twelve laboratories under the supervision of Flint and the results are to be expected in the near future.

Methods to assess specific organ toxicity have been developed for a number of organs, including liver, skin, and kidney. In France a multi-laboratory study on the hepatotoxicity of compounds is being carried out (18). Programs on validation of neurotoxicity testing are presently being planned.


The examples of validation studies mentioned above show that in Europe a considerable effort is given to this subject. Fortunately, these activities are not restricted to our continent and many studies are being carried out in cooperation with laboratories in other countries as well. Discussions on the scientific basis of validation studies as well as on the acceptance of well-validated alternative test methods do take place at an international level. In this context it is of importance to stress the need for internationally agreed acceptance of test methods. Therefore, if validation studies are carried out under the supervision of international bodies such as the OECD, and in Europe the EC, this must be considered an advantage for the eventual acceptance by regulatory authorities. Some of the studies mentioned in this overview are being carried out on a national scale, and it should be encouraged that these programs, that are sometimes similar, will seek cooperation.

It has been considered that the ideal number of institutions taking part in an interlaboratory assessment study to be four or five. If the acceptance of alternative methods is promoted by the inclusion of laboratories from more countries this may be the more pragmatic approach. Acceptance of alternative methodologies will be facilitated if the evaluation of validation studies are published in the international scientific literature. It would be of great importance if review panels would be formed or agreed upon by the scientific community and by international regulatory bodies.

It is very encouraging to see many activities in the field of validation of alternative methods, since this is a prerequisite for their acceptance and application in risk assessment of chemicals. For many scientists working in the field of alternatives for animal testing these developments will never go fast enough. However, in a number of cases there is more progress than many of us had hoped some years ago. The question now seems not to be if we convince regulatory agencies to accept nonanimal methods by how and when this will be achieved.


  1. Frazier, J.M. (1990). Scientific criteria for validation of in vitro toxicity tests. OECD Environmental Monographs No. 36, OECD, Paris, 68.
  2. Balls, M., Blaauboer, B., Brusnick, D., Fraizer, J., Lamb, D., Pemberton, M., Reinhardt, C., Roberfroid, M., Rosenkranz, H., Schmid, B., Spielmann, H., Stammati, A.-L. & Walum, E. (1990). Report and recommendations of the CAAT/ERGATT workshop on validation of toxicity test procedures. ATLA 18: 303-337.
  3. Balls, M. & Clothier, R. (1989). Validation of alternative toxicity test systems: lessons learned and to be learned. Molecular Toxicology 1: 547-559.
  4. British Toxicology Society (1984). Special report: A new approach to the classification of substance and preparations on the basis of their acute toxicity. Human Toxicology 3: 85-92.
  5. Van den Heuvel, M.J., Dayan, A.D. & Shillaker, R.O. (1987). Evaluation of the BTS appraoch to the testing of substances and preparations for their acute toxicity. Human Toxicology 6:279-291.
  6. Kayser, D. (1989). Alternative test procedures to the classical LD50 test. LD50 testing and classification schemes - the possibilities for change. Commissions of the European Communities, Brussels. pp. 65-69.
  7. Van den Heuvel, M.J. (1989). A fixed-dose procedure as a replacement for the classical LD50 test: the UK initiative. LD50 testing and classification schemes - the possibilities for change. Commission of the European Communitites, Brussels. pp. 71-82.
  8. Knox, P., Uphill, P.F., Fry, J.R., Benford, D.J. & Balls, M. (186). The FRAME multicentre project on in vitro cytotoxicity. Food and Chemical Toxicology 24: 457-461.
  9. Clothier, R.H., Hulme, L.M., Ahmed, A.B., Reeves, H.L., Smith, M., & Balls, M. (1988). In vitro cytotoxicity of 150 chemicals to 3T3-L1 cells assessed by the FRAME kenacid blue method. ATLA 16: 84-95.
  10. Mazziotti, I., Stammati, A.-L. & Zucco, F. (1990). In vitro cytotoxicity of 26 coded chemicals to HEp-2 cells: a validation sutdy. ATLA 17: 401-406.
  11. Clothier, R.H., Hulme, L.M., Smith, M. & Balls, M. (1989). A comparison of the in vitro cutotoxicities and acute in vivo toxicities of 59 chemicals. Molecular Toxicology 18: 571-577.
  12. Bondesson, I., Ekwall, B., Hellberg, S., Romert, L., Stenberg, K. & Walum, E. (189). MEIC, a new international multicenter project to evaluate the relevance to human toxicity of in vitro cytotoxicity tests. Cell Biology and Toxicology 5: 331-347.
  13. Ekwall, B., Bondesson, I., Castell, J.V., Gomez-Lechon, M.J., Hellberg, S., Hogberg, J., Jover, R., Ronsoda, X., Romert, L., Stenberg, K. & Walum, E. (1989). Cytotoxicity evaluation of the first ten MEIC chemicals: acute lethal cellular assays and by oral LD50 tests in rodents. ATLA 17: 82-100.
  14. Spielmann, H., Gerner, I., Kalweit, S., Moog, R., Wirnsberger, T., Krauser, K., Kreiling, R., Kreuzer, H., Lüpke, Miltenburger, H.G., Müller, N., Mürmann, P., Pape, W., Siegemund, B., Spengler, J., Steiling, W., & Wiebel, F.J. (1991). Interlaboratory assessment of alternatives to the Drazie eye irritation test in Germany. Toxicology In Vitro 5: 539-542.
  15. Blein, O., Adolphe, M., Lakhdar, B., Cambar, J., Gubanski, G., Castelli, D., Contie, C., Hubert, F., Latrille, F., Masson, P., Clouzeau, J., Le Bigot, J.F., De Silva, O. & Dossou, K.G. (1991). Correlation and validation of alternative methods to the Drazie eye irritaion test (OPAL project). Toxicology In Vitro 5: 555-557.
  16. Van Erp, Y.H.M. & Weterings, P.J.J.M. (1990). Eye irritancy screening for classification of chemicals. Toxicology In Vitro 4: 267-269.
  17. Flint, O.P. & Orton T.C. (1984). AN in vitro asay for teratogens with cultures of rat embryo midbrain and limb bud cells. Toxicology and Applied Pharmacology 76: 383-395.
  18. Fautrel, A., Chesné, C., Guillouzo, A., De Sousa, G., Placidi, M., Rahmani, R., Braut, F., Pichon, J., Hoellinger, H., Vinetézou, P., Mecion, C., Cordier, A., Lorenzon, G., Genicourt, M., Vannier, B., Fournex, R., Pelous, A.F., Bichet, N., Goy, D. & Cano, J.O. (1991). A multicenter study of acute in vitro cytotoxicity in rat liver cells. Toxicology In Vitro 5: 543-547.



Prepared by:Makoto Umeda, PhD
Yokohama City University
Kihara Institute for Biological Research
Nakamura-cho 2-120-3
Yokohama, 232 Japan

Validation activities involving alternatives/in vitro methods in Japan have not been as active and well-organized as elsewhere, except in the area of chromosome aberration tests. Several reasons can be proposed for this slow reaction. Actions of animal rights activists, "Japan Green Federation" and "Antivivisection Association", are relatively small scale, and the Japanese societies are not so seriously concerned with them. There are many animal toxicologists but only a few tissue culture specialists engaged in toxicological research and testing. Regulatory agencies in Japan are very conservative.

Meanwhile, Dr. Sugahara, an emeritus professor of Kyoto University, noticed the importance of alternative tests for animal experiments, and started a survey and organized a study group on this subject 10 years ago. The group developed to organize the Japanese Society of Alternatives for Animal Experiments (JSAAE) in 1989, only two years ago. Recently, the Japanese Society of Toxicological Science (JSTS) also initiated activity on organ toxicity using tissue culture and pleiotropic approaches to toxicity evaluation. Therefore, real systematic activities on alternative tests have just begun in Japan.

Organizations which are involved or interested in in vitro alternatives for toxicity testing are listed in Table 2.

Table 2: Organizations Involving or Interested in In Vitro Alternatives

Name of OrganizationDescription
Ministry of Health & WelfareGovernment
National Institute of Hygiene SciencesNational Institute belonging to Ministry of Health & Welfare
Japan Health Sciences FoundationSatellite organ of Ministry of Health & Welfare
Science and Technology AgencyGovernment
Institute of Physical & Chemical ResearchAn extradepartmental body of Science and Technology Agency
Agency of Industry, Science and TechnologyGovernment institute belonging to Ministry of International Trade and Industry
Japanese Society of Alternatives for Animal ExperimentsScientific Society
Japanese Tissue Culture AssociationScientific Society
Japanese Society of Toxicological ScienceScientific Society
Japanese Society of Dermal Materials and DevicesScientific Society
Japanese Society of BiomaterialsScientific Society
Japan Pharmaceutuic Manufacturers AssociationScientific Society
Japanese Cosmetic Industry AssociationScientific Society

Background Factors

In order to obtain reliable and relevant data using in vitro methods, background factors need to be established. One of the important activities related to the background factors is to standardize tissue culture materials. Recently, MEM, plastic dishes, and plastic freezing tubes were standardized as JIS (Japanese Industrial Standards). Other tissue culture materials such as clean benches, filtration sets for sterilization and so on are on the way to standardization.

The second type of activity is the establishment of cell banks. Several cell banks are established in Japan. The activities in ATCC are far beyond those in the cell banks in Japan. However, Japanese researchers can now utilize standard cell lines supplied from those cell banks. A committee set up in the Japanese Tissue Culture Association (JTCA) endeavors to promote the cooperation of these banks. They attempt to use the same protocols for quality control and to establish a universal data bank. That are also discussing such issues as how to cope with the ethical problems.

The third type of activity is to publish practical methodology books. Groups in the Japanese Environmental Mutagen Society (JEMS) published an atlas book (1) and a data book (2) on chromosome aberrations. A group in JTCA is also preparing a book on toxicity tests.

GLP assignment is another factor for standardizing testing activities. Many industrial laboratories are now GLP certified for the chromosome aberration test. Thus, the data obtained in these laboratories can be evaluated on an equivalent basis.

Some organizations, such as the Institute of Physical and Chemical Research (Riken), are trying to begin activities as a data bank, but are still on a small scale. We have no recognized organization of chemical banks or reference laboratories for alternative tests for animal experiments.

Present Status of In Vitro Toxicity Testing in Japan

In Japan only the chromosome aberration test is validated and other tests need further study for validation (Table 3).

Table 3: Present Status of In Vitro Toxicity Testing in Japan

Name of StudySponsorName of Test EvaluatedNumber of Chemicals TestedStatus of ProjectPublication
Chromosome aberration testJEMS1AtlasVarious chemicalsValidated1
Data bookVarious chemicals--781Validated2
Test on injection drugsMHW2Toxicity on HeLa cellsInjections drugs--13Intralab.3
MHW2Toxicity on rat muscle cellsInjection drugs--11Intralab.4
MHW2Toxicity on HeLa and CHL cellsInjection drugs--13Intralab.5
Draize TestJSAAE3Cytotoxicity on cultured cells (NK-assay)Detergents--52
STA4Cytotoxicity on SQ-5 cells at ED1Detergents--10Intralab.Ohno et al
Cytotoxicity on HeLa and SIRC cellsDetergents--12Intralab.Ohno et al
Biomaterials guideline for chemical and biological evaluation for dental materialsMHWCytotoxicity on L-929 cellsDental materials--1
Versailles projects on advanced materials and standardSTACytotoxicity on L-929 cellsBiomaterials--2Intralab.13
Studies on the estab. on toxicity testing program for med. device and biomaterialMHWCytotoxicity on BLAB/3T3, L-929, and V79Rubber accelerators--5
Polyurethane containing nibber accelerators or antioxidants-10
Rubber antioxidants-5
Application of stress responseJSAATCytotoxicity on cultured cellsMetals--12
Tests for anticancer drugsJSAATCytotoxicity using mutant cellsAnti-cancer drugs--8Intralab.20
Tests for anticancer drugsCytotoxicity on SQ-5 cellsAnti-cancer drugs--6Intralab.21
Genotoxicity testJSAATPremature chromosome condensation methodRadiations--7
CytotoxicityJSAATCytotoxicity on insect cellsChemicals--13Intralab.24
Neuronal antestetized drugsJSAATNeuronal antestetization using houseflyGlutamic acidIntralab.25
Establishment of new culture methodJSAATIntralab.26
Test for steroid hormonesJSAATHen's fertile egg testSteroid hormones--2Intralab.27
EmbryotoxicityJSAATToxicity test on chick and rat embryosCyclophos-phamideIntralab.28

1 JEMS: Japanese Environmental Mutagen Society
   2 MHW: Ministry of Health and Welfare
   3 JSAAE: Japanese Society of Alternatives for Animal Experiments
   4 STA: Science and Technology Agency

This table was completed by the help of Dr. M. Watanabe, Div. Radiat. Biol. Med., Yokohama City Univer., Dr. A. Sato, Dept. Biomat. Sci., Fac. Dentistry, Tokyo Medical and Dental Univ., and Dr. T. Ohno Cell Bank, Inst. of Physical and Chemical Res. (Riken).

  1. Chromosome Aberration Test

    The government adopted the chromosome aberration test for many regulations on safety evaluation of various types of chemicals as a screening test for carcinogenic substances. There are two reasons for this early adoption about this test. One is the people's serious concern about dangerous environmental mutagens since the 1950s, because of the occurrence of several mass disasters such as alkyl mercurial, PCB, cadmium or pesticide poisonings, and severe public hazards such as air and water pollution. Another reason is the relationship between worldwide trade and chemical safety. OECD led the standardization of the safety evaluation of chemicals, especially carcinogenic substances, in order to avoid trade impediment. While JEMS positively and actively supported and helped the standardization and validation of the test, the validation process, however, was not systematic, but empirical. Especially, the laboratories at the National Institute of Hygienic Sciences played a fundamental and important role in the process. As already mentioned, several background factors such as publication of an atlas, GLP assignment, and so on was provided in relation to this test.

    My impressions of this test, as a member of the government regulatory committee, are that protocols proposed by Japan and other countries were a little different, and we found it difficult to judge their results. Concentration of test chemicals, treatment period, and concentration of S9 are different. Moreover, CHL cells are usually used in Japan, but V79 cells, CHO cells or human lymphocytes are used in other countries. V79 and CHO cells sometimes have an elevated level of background aberrations. My opinion is that the establishment of a worldwide standard protocol is necessary.

    Another important observation is that, although all the participating laboratories have GLP assignment, still there are some laboratories with low quality. They report various background aberrations, or lower or higher aberrations in positive controls. Therefore, the necessity of inspection of laboratories is recommended for the sake of improving some laboratories.

  2. Toxicity Test on Injection Drugs

    In Japan there was a high occurrence of quadriceps femoris contracture about 30 years ago. One of the causes is suggested to be the repeated intramuscular application of some injection drugs to infants. Safety evaluation of injection drugs was an important issue. Because the injury seemed to arise from a direct cytotoxic effect of the muscle tissue, the use of cell culture cytotoxicity test was proposed to be applicable for detecting potentially injurious preparations. The study group organized under the guidance of the Ministry of Health and Welfare during late 1970s and 1980s adopted an in vitro toxicity test as one of the useful screening and adjunct tests for injection drugs, as well as the hemolytic test and measurement of serum creatine phosphokinase.

    Creatine phosphokinase is a muscle-specific enzyme. After injection of drugs to rabbits, serum creatine phosphokinase increased to reaching a maximum level after 24 hours. decreased gradually, and reached the normal level again after seven days. The magnitude of the serum response was shown to be well correlated to the muscle injury. After thorough examination and discussion, the advisory group decided to request that the pharmaceutical companies include animal experiments for the final marketing preparations. Animal experiments were thought superior at the time.

    We conducted many experiments on alternatives for injection drugs (3-5). Experimental methods using 96 well microplates are useful for measuring cytotoxic effects (5). Another trial involved the cultivation of muscle cells. When newborn rat muscle culture was treated with some drugs, myotubes were severely affected although the concomitantly growing fibroblasts looked almost intact (4). This means that in some drug tests using muscle cell culture is relevant.

    According to these results, in vitro cytotoxicity test can be satisfactorily applied to the toxicity evaluation of injection drugs. Cytotoxicity assay using strain cells can be utilized as a screening test, and that muscle cells can be used as an adjunct test. Recently, it is reported by the Pharmaceutical Manufacturers Association (USA) of a method based on the leakage of creatine phosphokinase from cultured muscle cells (6). The method is surely meaningful as a possible replacement test.

  3. Alternatives to Draize Test

    Cosmetics industries in Japan have also realized that they cannot continue to use the Draize test for ocular irritation testing. Many in vitro trials have been made (7-11), and the accumulating evidences are now ready for the extensive validation of interlaboratory assessment. The Science and Technology Agency has just initiated a validation study.

  4. Biomaterials

    The safety of dental and other biomaterials is also a very important issue. Studies sponsored by Ministry of Health and Welfare are in progress to establish the guidelines of biological evaluation for dental materials or toxicity testing programs for medical devices and biomaterials. Meanwhile, the Science and Technology Agency continues the Versailles projects on advanced materials and standards (12-16).

  5. Other Tests

    Many studies on alternatives using cultured cells are ongoing. They include a cell toxicity test using the stress response (17, 18), computer-simulation (19) m toxicity tests for antitumor drugs (20-21), genotoxicity tests (22, 23), and others. As to organ- or tissue-specific toxicity tests, cultivation of liver parenchymal cells maintaining their specific character has succeeded. Skin, kidney or nervous cells, and tissues can be cultivated constantly. Organ culture of chick or rat embryo tissue has been developed for the examination of teratogenic substance (25-28).

    These trials are well on their way. Reports on these research projects are presented and discussed at the meetings of JTCA, JSAAE, JEMS, and JSTS. These tests, however, are at the stage of the establishment of procedure, and need more study for systematic validation.

  6. Transformation Test and Promoter Test

    Animal carcinogenicity tests require much time and money, and the Ames and chromosomal aberration tests serve as screens. In vitro transformation tests which have been reported to have direct relevance to carcinogenesis can be proposed as candidates for adjunct or replacement tests for carcinogen detection. In 1984 a working group met in Lyon and discussed molecular and cellular mechanisms of cell transformation, biological similarities between in vitro and in vivo carcinogenesis, and the feasibility of screening environmental carcinogens by in vitro transformation assays using two established cell lines, BALB/3T3 and C3H/10T1/2 cells (29). The working group identified the methodological and technical problems associated with these assays, prepared recommendations for practical procedures in test performance and scoring morphologically transformed foci. The major aims of this working group were to identify the problems pertinent to transformation assay procedures and to try to improve the assay protocol, so that cell transformation could be used reliably for screening environmental carcinogens. It was hoped that use of these recommended protocols would facilitate international validation of the results of the cell transformation assay.

    In spite of the expectation of the working group, however, the methods have developed little during the past eight years. There are still problems in the cell transformation assay, even using these well-exploited cell lines. There are several points which need attention. Careful cell maintenance is needed. Serum is critical. Serum appropriate for transformation assay must be selected before the experiments. Transformation rate is rather low, and, therefore, relatively large number of dishes must be prepared. This makes this assay method laborious and very expensive among various in vitro tests. The experimental method itself is rather simple, and can detect carcinogenic promoters (30-31). This latter fact is a very important advantage of this test.

    It is generally accepted that application of the cell transformation assay to carcinogen screening is very important. Transformation assay using appropriate cell lines and advantageous medium must be a high priority candidate test for validation.

Future Directions for Japanese Validation Activities

From the background mentioned above, the following items must be considered for Japanese validation activities.

  1. Promoting an understanding of the necessity for alternative testing methods among scientists and the general public.
  2. Education of more in vitro toxicologists in order to increase the population of toxicologist with appropriate skills.
  3. Need for further improvement of background factors.
  4. Promotion of collaboration among scientists, which include governmental scientists, toxicology experts, statisticians, tissue culture specialists, and industrial scientists.
  5. Consideration of the establishment of database, chemical bank, and so on.
  6. Periodic inspection of testing laboratories.

In order to be started along the right lines of validation work, the following steps are recommended. Initially, working group including governmental and toxicological scientists must be set up. Their task is to survey what type of alternative experiments can be validated. Then, after selecting one or two suitable methods, they must try to raise funds from governmental or industrial sources. This phase is indispensable. Next, the study group must be set up. Here, the tissue culture specialists engaged in research with the topic cell culture methods must be included. Statisticians are also to be included. Only after a thorough validation project headed by toxicologists and statisticians is completed, will acceptance be attained.


  1. Mammalian Mutagenicity Study Group, Japanese Environemtnal Mutagen Society. Atlas on Chromosome Aberration Induced by Chemicals. Asakura Shoten, Tokyo, 1988.
  2. Ishida, M., Jr. (1987). Data book of Chromosomal Aberration Test In Vitro, Revised Edition. Life-Science Information Center, Tokyo.
  3. Umeda, M., et al. (1977). Cytotoxicity of injection drugs on cultured cells (in Japanese). Igakuno Ayumi 100: 447-453.
  4. Gorai, I., et al. (1978). Cytotoxicity of injection drugs on cultured cells (in Japanese). Yokohama Igaku, 29: 85-94.
  5. Saotome, K. et al. (1989). Cytotoxicity test with simplified crystal violet staining method using microtitre plates and its application to injection drugs. Toxicology In Vitro 3: 317-321.
  6. Ratner, M. (1990). Toward using in vitro toxicology in the drug approval process. Bio/Technology 8:1248-1249.
  7. Watanabe, M., Watanabe, K., et al. (1988). In vitro cytotoxicity test using primary cells derived from rabbit eye is useful as an alternative for Draize testing. Alternative Methods in Toxicology 6: 285-289.
  8. Watanabe, M., Watanabe, K., et al. (1989). Use of primary rabbit cornea cells to replace the Draize rabbit eye irritancy test. Toxicology In Vitro 3: 329-334.
  9. Torishima, H., Arakawa, H., et al. (1990). Application of normal epidermal keratinocytes in serum-free medium as alternative to the Draize ocular irritatiing test. Alternative Animal Test Experiments 1: 20-26.
  10. Okubo, T., Kiraiwa, K. et al. (1990). In vitro cytoxicity test using rabbit conjuectiva, rabbit cornea and HeLa cells as alternatives for the Draize eye irritation test. Alternative Animal Test Experiments 1: 2-9.
  11. Itagaki, H., et al. (1991). An in vitro alternative to the Draize eye-irritation test: Evaluation of the crystal violet staining method. Toxicology In Vitro 5: 139-143.
  12. Sato, A., et al. (1990). Studies on the standardization of cytotoxicity testing for dental materials. Tissue Culture in Dentistry 27: 33-35.
  13. Tateishi, T., et al. (1990). Standardization of biocompatibility test procedure by cell culture. Journal of Mechanical Engineering Lab 44: 166-172.
  14. Nakamura, A., et al. (1990). Correlations among chemical constituents, cytotoxicities and tissue responses: in the case of natural rubber latex materials. Biomaterials 11: 92-94.
  15. Ikarashi, T., et al. Comparative studies of the toxicity of natural rubber latex materials in cell culture and in vivo implantation test. Journal of Biomedical Materials Research, submitted.
  16. Tsuchiya, T., et al. Comparative studies of reference standard materials in various cytotoxicity tests and in vivo implantation test. BIOMAT 91, in print.
  17. Takahashi, T., Inada, Y., et al. Sysnthesis of heat shock protein during the development of thermo-tolerance in mouse sarcoma 180 cells in vivo and in vitro, submitted.
  18. Watanabe, M., Suzuki, K. et al. Differential heat sensitivity of normal and transformed Syrian hamster embryo bells in confluence. International Journal of Hyperthermia, in press.
  19. Yonehara, Y., Aoyama, T. et al. (1990). An attempt to develop computer-simulated alternatives to in vivo LD50 assay. Alternative Animal Test Experiments 1: 41.
  20. Utsumi, H., Ito, A., and Watanabe, H. (1989). Establishment of wild-type and acatalasemic mouse cell lines. In Proceedings of the 3rd Annual Meeting of JSAAE, Yokohama, pp. 124-127, 1989.
  21. Ohno, T. et al. In submission.
  22. Watanabe, M., Suzuki, M., et al. Comparison of the efficiency of mutants and/or transformants and chromosome damage induced in SHE cells by various LEF radiation. In Vitro Toxicology, in press.
  23. Watanabe, M. Suzuki, M., et al. Radiation-induced chromosome damage in G1-phase cells as breask in premature chromosome condensation (PCC) and its biological meaning. Journal of Radiation Research, in press.
  24. Yanagimoto, Y., Sato, K., Mitsuhashi, J. (1990). Difference in effects of rotenone on insect cell lines. Altern. Animal Test Exp. 1: 38.
  25. Abe, T., and Takahashi, Y. (1990). Investigation of mass testing method for neuronal anesthetization using housefly. Alternative Animal Test Experiments 1: 38.
  26. Negami, A., and Tominaga, T. The growth and differentiation of human endometrial carcinoma cell line (NTK-1) -- new culture methods using an extra-cellualr matrix. Alternative Animal Test Experiments 1: 55-56, 100.
  27. Nishigori, H., Mizunuma, M., and Iwatsuru, M. (1990). Hen's fertile egg screening test (HSET): Determination of androgenic activity of steroid and LD50 of drugs. Alternative Animal Test Experiments 1: 40-41.
  28. Nakamura, T., and Inomata, T. Use of chick embryo in screening for embryo and rat fetuses.
  29. IARC/NCI/EPA Working Group: Cellular and molecular mechanisms of cell transformation and standardization of transformation assays of established cell lines for the prediction of carcinogenic chemicals: Overview and recommended protocols. Cancer Research 45: 2395-2399, 1985.
  30. Sakai, A. and Sato, M. (1989). Improvement of carcinogen identification in BALB/3T3 cell transformation by application of a 2-stage method. Mutation Research 214: 285-296.
  31. Umeda, M., et al (1989). Promotional effect of lithocholic acid and 3-hydroxyanthranilic acid on transformation of x-ray-initiated BALB/3T3 cells. Carcinogenesis 10: 1665-1668.



Prepared by:Judson W. Spalding, PhD
National Toxicology Program
NIEHS, Building 101, South Campus
111 Alexander Drive
Research Triangle Park, NC 27709

I will assume that "chemical bank" refers to a central chemical repository where chemicals can be accumulated and stored under proper and controlled conditions. This chemical bank will then serve as the official source for chemical distribution to the investigators (laboratories) where the validation exercises are to be performed.

  1. A central chemical repository is desirable for the following reasons:
    1. Large amounts of single batches of chemicals to be tested can be stored properly and the source verified.
    2. A chemical bank facilitates and ensures that all validation laboratories receive the same chemical from the same lot/batch. These chemicals should be sent UNDER CODE each with a separate "aliquot number" which is randomly selected. Initially, it may be advisable to include several known active and inactive substances which are identified as such, e.g., not coded.
    3. The chemical bank, as a repository, should be able to verify the purity of each chemical and/or identify the major components of any mixtures that is used. (The question of a requirement for the high purity, >99%, of any particular substance can be argued from two sides. If the product used in commerce is a mixture or of technical grade, you should test. Others believe you should test the pure substance or the pure forms of the major components of a mixture).
    4. Ideally, the chemical bank (repository) will become the highly credible source of standard and/or reference chemicals that other investigators will want to use as positive controls or to use as standards for developing other in vitro models.
    5. It is obvious that a data management system will be required.
  2. The chemicals (substances) first used in the validation exercise should be of a limited number and distributed UNDER CODE. The group of chemicals should include positive (toxic) agents of high, medium, and low potency (toxicity) and several inactive substances. For the first exercise, the chemicals should be such that their physical characteristics do not interfere with test performance; e.g., the substances should be soluble within the expected toxic range (0-100% survival), of low volatility, should not react with the culture dishes (plastic), and should not effectively alter the pH of the culture system over the toxicity range employed. The chemicals should not be a problem in the initial validation exercise. The fact that they will be CODED is all of the "mystery" that is required.

It is important to keep in mind that the purpose of the "validation process" is to evaluate the biological assay system and the practitioners of the assay in such a way that the "critical elements" of the test system can be identified as soon as possible. The objective should be not only to establish the effectiveness of an assay in giving positive results with "known toxicants", but more importantly, to determine the assay's effectiveness at discriminating between known active and inactive chemicals.

The test results for any specific chemical/biological system must be demonstrated to be REPRODUCIBLE. Results from at least two completely different experiments must be in agreement. In validation exercise, all of the results from experiments performed should be reported and should include the data from experiments that were not in agreement.

It is very important to begin to look forward to building a database of chemicals/substances so eventually a set of test results can be submitted to a computer-based intelligence system for structure-activity analysis.

Some Comments On:

Important Characteristics of an In Vitro Test System

  1. In the beginning, it may not be desireable to require that a strict standard protocol be followed for each in vitro test system. If the chemicals selected are those for which there should be an expected and consistent response, it is better to let a standard protocol evolve, so that the "critical elements" required for performing the assay can be identified. It should be noted, that for any particular assay, the standard protocol will not always identify an "active substance" if the protocol is rigidly followed.
  2. It is important to identify the factors that are inherently characteristic of the particular biological system being used: e.g., the factors associated with the husbandry of a cell system that may impact on the variability in expression of the assay endpoint or contribute to the spontaneous incidence rate of the assay endpoint
  3. Positive (active) and negative controls must be run in every experiment in order to validate the response of each individual test performance of the biological system being used.
  4. In our experience with the validation of in vitro mammalian cell transformation assays, the development of criteria for measuring chemical activity was an evolving and ongoing process for some time. You can and should expect that the process for developing protocols and criteria for measuring test results will be an ongoing exercise.
  5. You cannot expect that a single in vitro cell system is going to be adequate for optimally discriminating between all active and inactive substances. However, you should expect that one or several in vitro systems will correctly discriminate between active and inactive chemicals among certain classes of substances. It is very important to define and recognize the limits of any in vitro test system.
  6. An in vitro assay system must demonstrate some acceptable level of sensitivity (detection of active substances), specificity (correctly identifying inactive substances), and accuracy (percent of all substances identified correctly). These data are obtained from retrospective studies of chemicals (substances) which have been previously defined as being active or inactive in an acceptable in vivo animal assay.
  7. Finally, an in vitro assay system must predict with an acceptable degree of accuracy the activity of a set of substances that will be subsequently evaluated in the most appropriate in vivo assay.



Prepared by:Horst Spielmann, PhD
P.O. Box 330013
D-100 Berlin 33, Germany

In 1990 two international workshops what were held in Europe on the "Validation of Toxicity Test Procedures" (1, 2) as well as the "Second Report of the FRAME Toxicity Committee" (3) recommended unanimously that the establishment of an International Reference Chemical Data Bank is a matter of urgency. Moreover, the bank should:

  1. Provide open-access listings of scientifically-selected chemicals, and
  2. Be backed by toxicological data reviews, safety advice, and a source of chemicals of known purity, for use in validation studies.

The objective of collecting toxicological data from animals and humans is to provide a basis for classifying chemicals with respect to specific biological endpoints, which will serve as a reference against which the performance of tests developed to predict the endpoints can be judged. In practice in vivo toxicological studies provide two forms of reference data: data generated in animals and in humans.

Animal Data

Over the past 50 years, a considerable amount of toxicological data has been generated in laboratory animals on a wide range of chemicals and mixtures, usually with the aim to predict effects in humans and not to be used as references in toxicological validation studies. It is, therefore, not surprising that recent attempts to assess the completeness of these toxicological profiles on existing chemicals, have shown that the data are not comprehensive for most chemicals and that, for many chemicals, little or no data exist (1, 2).

Many of the in vivo test methods are complex since many factors are influencing both exposure to the chemical and the expression of toxicity. These factors, together with differences in assessment and interpretation of toxicity, have lead to the generation of diverse data (1).

Furthermore, not all data generated in toxicological studies are available for review, since much of this information has not been considered suitable for publication, e.g. negative data or confidential data from industry as well as contract laboratories. Therefore, even the most complete literature search will have to be supplemented by additional inquires that unpublished data be made available (1, 2).

Human Data

Human toxicological data are providing another important source of information that validations studies have to take into account. However, the validity of human toxicological data is highly variable, primarily because they are collected for a variety of purposes. Consequently, it is not easy to use them in a uniform fashion in a reference database. Therefore, schemes currently being developed for assessing human toxicological data and making them more available to validation studies should be welcomed and supported by government and funding agencies (1-3). Furthermore, information toxicologists should be encouraged to investigate ways of integrating the information obtained in experimental and human toxicology, and incorporating these data into the validation process for reference classification of chemicals, whenever this is feasible (1).

Whenever it is desirable and practicable, methods for the collection, collation, evaluation, and expression of experimental, human and veterinary toxicological data should be standardized in ways which have been agreed upon by, and are acceptable to, the toxicological community at large (1).

Information on Validation Studies

In order to avoid unnecessary duplication and to avoid the omission of promising new in vitro methods, public announcement should be made when the decision to plan and undertake a validation study is reached. Readily accessible in vitro toxicology data banks should be developed and supported to ensure that all information concerning test methodology and test validation, including detailed protocols and results obtained in validation studies, can be made freely available to all interested parties. Such data banks should be set up both within individual countries and/or within regional coordination centers (e.g. CEVMA, the proposed EEC Centre for Alternatives to Animal Experiments in Ispra, Italy). INVITTOX, the ERGATT/FRAME in vitro toxicology data bank has established (4) a validation study registration scheme, and lists of studies in progress will be published in ATLA at regular intervals.

Table 4: Current In Vitro Toxicological Data Banks

Project Name and SponsorPurpose of Data Bank (Hardware)Area of ApplicationContact PersonReference
INVITTOX ERGATT/CEC & FRAME (est. 1988)Collection of in vitro methods (PC)Research & testing pharm, tox validationKrys Ungar c/o
FRAME 34 Stoney
Nottingham NG1
ATLA 16, 323-343, 1989
Register of Cytotoxicity
Private & Acad. Science (est. 1988)
300 chemicals mammalian cytotoxicity IC50 data & acute LD50 (book)Research & testing pharm., toxWilli Halle
Inst. Drug
Research Berlin
Beiträge Wirkst.
Froschg. 32, 1988
ATLA '91 in press
Japan Res. Group Alt. to Anim. Testing (est. 1989)
In vitro tests experimental data (PC)Research & testing pharm, toxY. Yamada
Natl. Inst. Radiol.
Sci. Japan
Alt. Anim.
Testing & Exper. 1, pg. 42, 1990
German Fed. Health Office BGA (est. 1989)
In vitro tests methods and experimental data (PC)Research & testing pharm, tox bact, immun validationH. Spielmann
ZEBET BGA, Berlin, Germany
Bundesgeshblatt 28, 1989
Technical Database Services (est. 1990)
In vitro tox.
testing database
ongoing validation projects (PC)
Research & testing pharm, tox validationM. Green
Technical Database Services
New York
CAAT Newsletter 1991
MEDLINE US Natl. Lib. Med.Biomedical literature databank (online)Research & experimental unspecificWorldwide accessLocal inform. centers
US Natl. Lib. Med.
Toxicology literature data bank (online)Research & experimental unspecificWorldwide accessLocal inform. centers
Health Sci. Inc.
Predic. tox. endpoints from chem struct. (PC)Testing & research
Commercial productHealth Sci. Inc.
Bethesda, MD


  1. Balls, M., Blaauboer, B., Brusnick, D., Frazier, J., Lamb, D., Pemberton, M., Reinhardt, C., Roberfroid, M., Rosenkranz, H., Schmid, B., Spielmann, H., Stammati, A.L. and Walum, E. (1990). Report and recommendations of the CAAT/ERGATT workshop on the validation of toxicity test procedures. ATLA 18: 313-337.
  2. Balls, M., Botham, P., Cordier, A., Fumero, S., Kayser, D., Koeter, H., Joundakjian, P., Lindquist, N.G., Meyer, O., Pioda, L., Reindhardy, C., Rozemond, H., Smyrniotis, T., Spielmann, H., van Looy, H., van der Venne, M.T., and Walum, E. (1990). Report and recommendations of an international workshop on promotion of the regulatory acceptance of validated nonanimal test procedures. ATLA 18: 339-344.
  3. Animals and alternatives in toxicology: present status and future prospects (The Second Report of FRAME Toxicity Committe). ATLA 19: 116-138 (1991).
  4. Warren, M., Atkinson, K. and Steer, S. (1989). Introducing INVITTOX: The ERGATT/FRAME in vitro toxicology data bank. ATLA 16: 332-341.



Prepared by:Eugene Elmore, PhD
National Institute for the Advancement of In Vitro Sciences
4199 Campus Drive, #550
Irvine, CA 92715

In vitro validation studies have generally been performed by testing, commercial or university laboratories. These studies have been funded by both private industry, governmental agencies or government or private funded universities or agencies. The standards established for validation have varied from study to study and agency to agency. In vitro validation studies have demonstrated that no matter how hard you try to ensure that procedures are duplicated between laboratories, unique differences are usually present that may bias the data. Reference laboratories would provide standards, guidelines and training to hopefully minimize the bias and permit comparability and standardization between laboratories.

Reference Laboratory Location and Capabilities

Several reference laboratories should be located in the United States, Europe, and Japan. Each laboratory should be selected based on its experience in the development and validation of in vitro systems for evaluating toxicity. Such experience would permit the rapid evaluation of new assays for potential incorporation into the validation study. Laboratories would each perform common, standard assays in addition to unique assays.

Standard performance criteria for assays are important in establishing conditions for laboratories participating in validation studies. Reference laboratories would perform standardized in vitro assays upon request to evaluate test agent and assay response and to evaluate cell lines/strains from different laboratories to ensure uniformity of assay responses. In addition, an individual reference laboratory would perform assays that are not performed routinely in other laboratories. This unique capability would be used in establishing new systems and developing guidelines for future validation. The responses obtained in the reference laboratory would be used to develop draft protocols.

Each laboratory would have capabilities to bring assays on line to provide detailed training for new testing laboratories. By applying information gained in the performance of new assay procedures, the variability observed in the initial assay response could be assessed and potential assay problems identified prior to validation. The reference laboratory would work with the assay developer and other outside laboratories to resolve problems identified during the initial evaluation. Validation protocols could be finalized and outside laboratories trained (as appropriate) for performing the new assay during validation.

Once the assay is validated, the reference laboratory would train new laboratories in the performance of specific assays and recertify existing laboratories to ensure consistency of performance. Reference laboratories would evaluate evolving methodologies for incorporation into the assay protocols.

Each laboratory should be funded to maintain sufficient staff levels to meet routine assay evaluations, training, and certification needs. Reference laboratories with expertise on the intricacies of validation, standardization, and assay performance. Databases, which will be developed for each assay, can be shared with an online database, and hard and soft copy backup will be maintained at the individual laboratories.

The following specific functions should be considered for reference laboratories:

  1. Provide assay development and optimization.
  2. Development and validation of protocols.
  3. Generate standardized protocols for use by all testing laboratories.
  4. Routinely retest standards for each assay to determine reproducibility.
  5. Collect and maintain a database for standard agent responses for each assay.
  6. Establish performance and scoring criteria for each assay including: positive controls, negative controls, assay response criteria, and valid assay criteria.
  7. Provide on-site training in individual laboratories for specific protocols.
  8. Provide good laboratory procedure (GLP) and quality assurance training.
  9. Provide safety protocol reviews and laboratory safety awareness training (individual laboratory assessment).
  10. Interact with cell and chemical banks to provide ongoing testing for new lots of test agents and cells.
  11. Coordinate the evaluation and certification of new laboratories.
  12. Provide consultation for special problem agents, i.e., difficult to test agents, solubility problems, etc.
  13. Provide support and expertise as requested to the scientific panels and in vitro validation coordinators.

Reference laboratories would play a key role in the practical aspects of in vitro validation. Reference testing laboratories would provide for the development, standardization, and performance of individual assays, and generate standard protocols and operating procedures to permit the routine utilization of the assays by other laboratories, thereby permitting the development of an assay database for use in establishing valid assay criteria. The reference laboratories would generate response data for coded chemicals that would be used by other laboratories to ensure that the standards produced meet the expected guidelines. Reference laboratories would also provide training for basic procedures, safety, and assay performance expectations.

The establishment of reference laboratories would provide critically needed guidelines and support for the validation process and would speed the acceptance of validation in vitro assays at the regulatory agency level.



Prepared by:June Bradlaw, PhD
Division of Toxicological Sciences
U.S. Food and Drug Administration
8301 Muirkirk Road
Laurel, MD 20708

At the 1975 working conference on Toxicity Testing In Vitro convened by the Committee on Carcinogenesis, Mutagenesis, and Toxicity Testing In Vitro of the Tissue Culture Association, the working group on standardization concluded that there was a need to: (1) "employ a reference cell line to serve as a competitive biological standard, a yardstick to be used in measuring the biological response of other cell lines," (2) establish standardization based on the best-characterized human diploid cell system available, (3) recommend reference sources of cell lines that are well-characterized for distribution as reference or standard cultures for toxicity testing. Thus, the need for a cell bank or cell repository for use in in vitro toxicity testing schemes was identified and justified by a scientific panel over 15 years ago (1).

In our current efforts toward validation of in vitro toxicity tests, the time is at hand to consider the mechanism by which such a cell bank or repository is established and operated at the national and international levels. For successful implementation, critical questions must be addressed including: (1) Should the repository be located at a central point? (2) What cell lines should be included in the collection? (3) How will the cell lines be characterized and maintained? (4) How will the acquisition and distribution of reference stock cell cultures proceed? and (5) How is this project to be funded?

In the United States, many universities and institutions have small individual cell culture collections. However, the two main cell culture repositories are: the American Type Culture Collection (ATCC) in Rockville, MD and the Cornell Institute for Medical Research in Camden, NJ which maintains the NIGMS Human Genetic Mutant Repository and the National Institute on Aging (NIA) Cell Culture Repository (2). Outside the United States, two cell culture collections are maintained in Japan, one in Braunschweig, Germany and another collection in south Britain. Ideally, a cell repository maintained for international use would facilitate the objectives required of validation activities. However, the transport of biological materials between countries may meet with various difficulties.

A precedent for the acquisition and distribution of stock cultures for a specific objective has already been established. For example, a specific collection of cell stocks for the production of vaccines are maintained at the ATCC through a cooperative agreement funded by the Food and Drug Administration and the World Health Organization (WHO). FDA and NIH diploid human cell lines plus hundreds of WHO cell bank seeds stocks were specially characterized in detail by several laboratories, tested, frozen, stored, and distributed throughout the world by ATCC. Conditions of the agreement include: storage and distribution of stock ampoules, backup freezer units for preservation of stocks, expansion of cell populations at various time intervals, recovery, and recharacterization of representative ampoules to monitor for viability and contamination and a description of cell stocks (3).

The establishment of a cell culture repository to facilitate validation activities appears to be justified. It may be appropriate to first initiate a small pilot project with a well-characterized cell line, maintained under standardized conditions to evaluate the process of cell stock acquisition and distribution. Participating laboratories will have the advantage of receiving ampoules of the same seed stock. The special collection could be expanded as needed and could include primary and/or low passage cell lines from target organ tissue in the future. However, the latter objective would require a different set of standards because cells would be limited unless a cell pool is considered. Several cell lines widely used in a number of cytotoxicity assays include: Hep G2 (human hepatoma cells), L929 (mouse fibroblasts), SIRC (rabbit corneal cells), MDCK (dog kidney cells), and a number of human diploid fibroblast cell lines. The type of standardization required for these representative cell stocks prior to distribution to laboratories include: (1) a complete screen for adventitious agents; (2) defined culture conditions for cell maintenance and specific culture conditions that might be helpful for use in particular assays, such as use of low serum or serum-free methods; (3) population doubling time under defined conditions; and (4) metabolic activation potential of the cell population where appropriate. It will be extremely important to make the cell stocks available to both nonprofit and for-profit organizations at reasonable costs and without royalty fees.

For primary or early passage mammalian cells, including human cells from target organ tissue, cell banking on a large scale is less likely for many of these target cells. However, both nonprofit and for-profit companies are making such cell model systems available. In order to meet the growing demand for primary and early passage cells in validation studies, a cell repository might consider establishing an appropriate set of standards. With human tissue for example, certain information is important including: donor and/or cell pool data, a screen for infectious agents and target tissue origin for primary cell cultures. Critical information might include: method of isolation, medium, serum and supplements, passage number, use of antibiotics, and screens for adventitious agents. Cell characterization will become an increasingly important component to standardization of primary or early passage cells. Issues such as cell type verification with cell-specific markers, maintenance of cell-specific function, and population doubling capacity of the cell become critical to defining the suitability of the differentiated cell for use in toxicity testing protocols.

The Tissue Culture Association (TCA) has an established committee that considers problems related to "Cell Standardization". Many cell culture experts could provide input on the subject of standard cell characteristics of a specific differentiated cell type.

Such information helps to establish the credibility of the cell system proposed for use in validation efforts.


  1. Berky, J., and Sherrod, P.D. (eds.) In Vitro Toxicity Testing. 1975-1976. Philadelphia, Franklin Institute Press 1977.
  2. Greene, A.E., Mueller, S.N., Mulivor, R.A. and Hay, R.J. (1989). Cell Repositories. In Vitro Cell & Development Biology 24: 855-856.
  3. Hay, R.J. (1988). The seed stock concept and quality control for cell lines. Analytical Biochemistry 171: 225-237.



Prepared by:John Yam, PhD, and
Leon H. Bruner, DVM, PhD
The Procter & Gamble Company
P.O. Box 398707
Cincinnati, OH 45239

Independent peer review is a generally accepted process for the evaluation of the quality of scientific research. This principle also applies to the scientific quest for viable alternative methods to animal testing, and is consistent with the Second Report of the FRAME Toxicity Committee on alternatives in toxicology which recommends that:

"Before the formal acceptance and incorporation of new methods into regulatory toxicology is proposed or considered, the results of a validation study should be considered by at least one independent assessment panel."

What are the benefits of an independent scientific panel? The purpose of the panel is to provide an impartial peer review of a validation exercise. Independent peer review inevitably leads to better science. Good peer review helps assure that the study was designed properly, the procedures were executed properly, the data was interpreted appropriately, and the conclusions drawn are supported by the data. The members of the scientific panel also could be an important resource providing guidance during the study. Importantly, peer review with endorsement from the independent scientific panel can provide the basis for acceptance of a method or battery of methods by the scientific and regulatory communities.

How should a panel be selected? Since independent assessment can take place at various stages of methods development and validation, there is no set rule. Flexibility is important. The panel members should be competent and independent. Ideally, the panel could include members with diverse backgrounds, allowing each to provide unique contributions. For example, a panel for a major validation project may include people with expertise in toxicology, cell biology, statistics, chemistry, and government regulations. However, it is not necessary that the whole panel encompass all these scientific areas. The membership also need not be fixed throughout the validation exercise. Appropriate experts could be invited to join the panel during any part of the study to address specific issues as they arise. The members could come from industry, academia, public, and government sectors. It also would be best that they not be directly involved in the validation activity under review. Depending on the scope of the validation project, the panel could be international or national.

When should the panel be assembled, before or after a validation activity? Scientific panels should be formed prior to the start of any major validation project. This way the panel can help define the goals of the study, review the hypothesis tested and assays studied, evaluate the test materials used, examine the quality of the in vivo data available, critique the designs of the study early, and help set criteria for interpretation of the results. This will ensure that the study is performed properly the first time, and that the criteria are not changed later to force-fit the data.

To ensure the effectiveness of the review process, it is critical that the charge of the panel be clearly explained at the beginning of the study. This will provide the panel members with a clear understanding of what they are expected to complete. At the end of the validation project, the panel should assess the quality of the experiments, data collection procedures, data analyses, and conclusions drawn. At the end of the review, the panel must respond specifically to each of the charges given, write its conclusions, and offer its recommendations for acceptance or the need for additional studies. Their report would then be used as support for additional evaluation or for regulatory acceptance of the methods.


  1. Animals and alternatives in toxicology: Present status and future projects. The Second Report of the FRAME Toxicity Committee. ATLA 19: 116-138, 1991.



Prepared by:Neil Wilcox, DVM, MPH
US Food and Drug Administration
Center for Veterinary Medicine
Room No.7-57, HFV-4
5600 Fishers Lane
Rockville, MD 20850

On behalf of the Food and Drug Administration, I would like to convey appreciation to the Johns Hopkins University, Center for Alternatives to Animal Testing and the Tissue Culture Association, for the opportunity to participate in this workshop. Establishing dialogue is imperative to achieving meaningful progress toward and developing the notion of international harmonization of validation of in vitro toxicity tests.

Approximately 30 officials in the federal government representing several agencies and units within those agencies were contacted for background information pursuant to this presentation. A sincere attempt was made to identify the current status of the validation of in vitro alternatives in those agencies which either regulate or participate in toxicity testing. The individuals contacted represent diverse expertise and vast experience in the Food and Drug Administration (FDA), Environmental Protection Agency (EPA), Consumer Product Safety Commission (CPSC), Center for Disease Control (CDC), Office of Assistant Secretary for Health (OASH), National Institutes of Health (NIH), National Institute of Environmental Health Sciences (NIEHS), and the National Toxicology Program (NTP). The cooperation and assistance received were gratefully appreciated.

The role of a regulatory agency in the federal government, such as the Food and Drug Administration (FDA), in the validation/acceptance of new testing methodologies is necessarily complex. The basic mission of the regulatory agency is that of public safety. Protecting the public from potential hazards that may occur from the use of therapeutic drugs, medical devices, foods, and cosmetics as mandated by the Federal Food, Drug, and Cosmetic Act and related laws is the responsibility of the FDA. Similarly, other agencies such as the Environmental Protection Agency (EPA), Consumer Product Safety Commission (CPSC), and Department of Transportation (DOT) have congressional mandates which charge them with specific regulatory duties in the interest of public safety.

In the regulatory process, the evaluation of safety is conducted on proposed products by various testing methods within the purview of their intended use. Although neither development nor validation of proposed new testing methodologies is the specific responsibility of federal regulatory agencies, there are circumstances under which either the development or validation of certain alternatives may be necessary. Historically, the in vivo model represents the standard to which most new, non-whole animal methodologies will be compared.

The policy issue, therefore, arises as to the role of the federal government in the "validation" process. Formal participation of the regulatory agency in the "alternatives" arena may occur more commonly in the role of "acceptance" of a proposed in vitro methodology that has been previously validated by the scientific community. Although it may be emphasized at this juncture that the federal regulatory agency does not necessarily have a determinate role in validation, it does not obviate the importance for the agency to directly participate in the process. Informal participation may take the form of commenting or reviewing data as well as providing views regarding the adequacy of various testing procedures as they proceed through the validation process. The review and acceptance of scientific innovations through the peer-review process is generally the sine quan non to recognition and evaluation by federal regulatory agencies.

In the context of this overview, it would be inappropriate to make the presumption that a "scientific panel" would be established in a regulatory agency as part of a scientific and policy review of a proposed in vitro model submitted for evaluation. Such a notion may or may not be applicable in a regulatory agency depending on the defined role and authority of the proposed panel. The intent of this brief is to outline a general paradigm illustrative of the administrative, scientific, and policy expertise emblematic of the regulatory review. No attempt will be made to identify specific scientific criteria necessary for validation.

Currently, a proposed new test is examined on a case-by-case basis by the federal agency to which its application may be appropriate. For this discussion, the "needs of a scientific panel" will be presented in a generic framework with reference to substantive issues requisite to the review of an alternative methodology. The "needs" at each level will vary depending on several determinate factors. The following outline is illustrative of a regulatory review process as it might progress in the evaluation of a proposed, unapproved in vitro test method. This generic approach is utilized to avoid the difficult task of composing a comprehensive list of the scientific disciplines that may be necessary in the evaluation of testing methodologies intended for validation and/or acceptance in the regulatory process.

  1. Application for Approval
    • Guidelines available advising principle investigator of the basic requirements for the submission of a proposed, unapproved alternative methodology.
    • Appropriate forms for submission completed.
    • Application submission -- forms evaluated for completeness.
    • Application returned for additional information or recommended for classification.
  2. Methodology Classification
    • Initial application review to evaluate purpose, rationale and relevance.
    • Proposed testing methodology endpoint identification and evaluation.
    • Databank search and review.
    • Application returned for clarification or recommendation for continued review.
  3. Review Coordination and Management
    • Policy issues identifies.
    • Research needs assigned and prioritized.
    • Necessary scientific expertise, support staff, technical expertise, and materials identified.
    • Time frames for reviews assigned.
  4. Intra- and Inter-Laboratory Testing
    • Protocol for proposed testing methodology reviewed.
    • Needs for intra-laboratory testing assessed.
    • Criteria for inter-laboratory.
    • Laboratory parameters analyzed including reliability, transferability, and reproducibility.
    • Parallel studies conducted where appropriate.
    • Scientific percept of blinding incorporated where possible.
  5. System Analysis
    • Chemical analysis including identification of calibration and standardization chemicals.
    • Test system analysis for appropriateness and compatibility for interaction with testing chemicals.
    • Mechanisms of action analyzed.
    • Dose selection for appropriateness.
    • Structure-activity analysis.
  6. Data Management and Analysis
    • Statistical evaluation for reliability, sensitivity, specificity, and correlation factor.
    • Comparative analysis with standard.
    • Assumptions and biases identified and evaluated.
  7. Final Analysis
    • Review data on all facets of review process.
    • Recommendation for appropriate action.
    • Policy decisions based upon final recommendation.
    • Data referred to data bank.
    • Review by General Council where appropriate.
  8. Re-evaluation
    • Follow-up quality assurance evaluation of approved testing methodologies.


Regulatory agencies within the United States are responsible for public health and safety. The promulgation and administration of regulations based upon laws enacted by Congress constitute the authority for this responsibility. Each federal agency is charged with the evaluation of multiple chemicals, products, and devices with potential hazards of varying degrees. Policy decisions have to be made on the relative safety of these entities within a continuum of constantly changing criteria delicately balanced between law, scientific data, and societal demands.

Each regulatory agency must make these decisions within the purview of their applicable laws, regulations, guidelines, and policies. The existing approved standards for evaluating proposed products have generally passed the test of time and have been extremely successful. However, the undeniable future trend is that of technological advancements in the non-whole animal (e.g. in vitro) testing arena.

New in vitro testing methodologies will most likely offer several advantages over the traditional in vivo models. Although the total replacement of animals in most testing methods remains a major obstacle, significant progress is being made in other testing areas such as reduction and refinement. Scientific consensus remains of major importance in the validation and acceptance of proposed, new testing methodologies. The underpinnings for agreement within the scientific community should be sustained within the province of scientific principles and not empiricism. All participants including research scientists, educational leaders, private industry, regulatory sector, and the international community need to work in harmony by adhering to this idealism. The basic principles of reliability, reproducibility, and transferability must provide the foundation for meaningful progress to occur as the notion of "validation" continues to be developed in the "alternatives" arena.



Prepared by:Hugo van Looy, PhD
OECD Environment Directorate
2 Rue Andre Pascal
Paris 75775, CEDEX 16 France


Under this title we discuss how validation of an in vitro test can be promoted at various stages in the development of the test.


The many parties interested in in vitro methods (e.g. laboratories in industry, academia, and government; scientific bodies; regulatory agencies) need access to a simple but complete system for identifying who is doing what. It is not information on large-scale (interlaboratory and national or international) validation exercises which has to be provided, although such information can be included. Furthermore, it is not information found in the open literature which has to be provided, although again it may be useful to include such. What is mainly needed is early-on information on who is developing new in vitro tests and information on who is using in vitro tests (whether they are used for in-house screening or for some other purpose). The purpose of such an information system is to create possibilities for cooperation in validation and for pooling validation data.

A Proposed Awareness System

Envisaged is not a resource-intensive data bank like ERGATT/FRAME INVITTOX which (although it includes promotion of general awareness as an essential element) contains full details, protocols, etc. Rather, the system would be a pointer-system used in interlinked networks with central contact points. Large countries would in all likelihood have one central contact point and small countries may prefer to make a grouping round a central contact point located in one of them. Alternatively only three central contact points covering respectively North America, Europe, and Japan could be established.

Laboratories using or developing in vitro tests would submit to a central contact point information on a format with a set of fields. These could include:

  • laboratory submitting the information (contact person);
  • test used or under development (endpoint, organism or cell type);
  • type of activity: development, validation, implemented;
  • status (ongoing, terminated, planned);
  • chemicals tested;
  • literature published, if any.

Central contact points would periodically exchange updated information and complement their own information. Periodically the information could be disseminated to users in PC-readable form.

Depending on resources available, the central contact point could activate the system by picking up multiple entries or similarities in inputs and make users aware of these.

An operational awareness system of this sort becomes self-facilitating by allowing users with similar interests to establish contacts and to engage in cooperation. The awareness system would also give an indication of what types of test are gaining acceptance.


With a view towards later regulatory use of certain in vitro tests, selection of candidate tests for validation is an essential element. This could be achieved in the context of the awareness system if one or more scientific panels could be associated with the central contact points. The panels would identify priorities for validation. Obviously, there are many other options for selection.

Coordination and Facilitation of Validation Studies

Initiatives for validation studies have been taken and will be taken by a variety of bodies like:

  • centers for promotion of alternatives: FRAME, MEIC, etc.
  • professional associations: CTFA, SDA, etc.
  • regulatory bodies: CEC, etc.

These bodies have the size and resources needed for coordinating validation studies.

In facilitation of a validation study, two aspects are to be considered: information and funds. In respect to information the means are:

  • newsletters from centers for promotion of alternatives (CAAT, FRAME, MEIC, etc.);
  • scientific journals;
  • regular information flows in organizations (WHO-IPCS, OECD, EC, EPA, etc.);
  • congresses (Toxicological Forum, etc.).

An operational awareness system of the type described above could also alert scientists worldwide on planned validation.

A more difficult issue, not surprisingly, is funding. In the absence of an institutional mechanism, the problem of funding needs to be solved on a case-by-case basis.



Prepared by:George C. Becking, Ph.D.
International Program on Chemical Safety (IPCS)
World Health Organization
P.O. Box 12233, NIEHS
Research Triangle Park, NC 27709

General Remarks

My colleague Hugo van Looy, from the Organization of Economic Cooperation and Development (OECD), has identified several excellent informatics-related activities which would be most useful in facilitating international collaboration and avoid duplication during the validation of in vitro tests. Although useful databases will be developed, they will do little in the way of generating internationally accepted data. Many elegant national studies can be run, but the results may or may not be accepted internationally without the coordination efforts of an international group perceived as being scientifically neutral.

Besides information sources we need active laboratory exercises internationally, if the in vitro tests are to be used in chemical safety programs worldwide. However, Dr. van Looy in the last paragraph of his excellent paper touches on the most crucial issue -- FUNDING. Countries and donor agencies must agree that in vitro tests have the potential for playing a significant role in health and environment protection. They then must make a long-term commitment to support this conclusion with funds.

Proposed Mechanism for International Coordination of Validation Studies

To ensure the acceptance by more than one country of new testing methods or the results generated from such tests in other countries, international validation studies are essential. The following model is based upon the mechanism developed by the International Programme on Chemical Safety (IPCS) and used to coordinate international collaborative studies ranging from eight to close to 100 laboratories. Test studied ranged from neurobehavioural in vivo tests to in vitro and in vivo short-term genotoxicity assays. I believe it could serve as a model for the international coordination of validation studies of in vitro toxicity tests.

IPCS Model

  1. Tests are coordinated by an international scientifically-based organization (OECD, CEC, WHO/IPCS, etc.) perceived as neutral regarding use of in vitro tests.
  2. A small steering group (5-6 experts) is appointed by the coordinating agency to provide technical guidance.
  3. The work is carried out by a qualified lead institution (e.g. NIEHS, EPA, CAAT, University of Uppsala, etc.) This institution works closely with the coordinating agency and the Technical Steering Group. It must have the appropriate expertise, but more importantly be committed to the program and be able to allow staff adequate time to carry.
  4. Participating laboratories are chosen, as was the steering group, primarily on the basis of expertise and secondarily on geographic distribution. Laboratories working with IPCS carried out the tests within their own budgets.
  5. Funding was obtained by IPCS from National Institutes, governments, and intergovernmental groups (CEC). It was only sufficient to pay the costs of the steering group meetings (five meetings), a final workshop for the participating laboratories, and publication costs. Total costs to IPCS for the in vivo short-term tests (97 laboratories in 16 countries) was approximately US $250,000 (1985). It is hard to estimate the cost per laboratory if all in-laboratory expenses had been paid, but a conservative estimate would be an additional 2.5 to 3.0 million dollars. It should be noted that IPCS had no trouble obtaining participating laboratories without paying testing costs until in vivo behavioral tests were suggested. For these tests in-house expenses were estimated at over $125,000 and very few institutions were willing to accept this additional burden.

I believe the IPCS model and that being used in the MEIC (Sweden) evaluation study should be compared. Perhaps an appropriate model could be developed which would facilitate additional test validation of in vitro toxicity assays worldwide.

Workshop Participants -- Present

Bas J. Blaauboer
Research Institute of Toxicology, Utrecht
University of Utrecht
P.O. Box 80.176
NL-3508 TD Utrecht, The Netherlands
Hugo van Looy (retired)
Organization of Economic Cooperation & Development
Environment Directorate
2 rue Andre Pascal
F75775 Paris, CEDEX 16 France
June Bradlaw
Division of Toxicological Sciences
US Food and Drug Administration
8301 Murikirk Road, HFF162
Laurel, MD 20708
(301) 344-5883.
Judson W. Spalding
National Toxicology Program/NIEHS
Building 101, South Campus
111 Alexander Drive
Research Triangle Park, NC 27709
Leon H. Bruner
(Substitute for John Yam)
The Procter & Gamble Company
P.O. Box 398707
Cincinnati, OH 45239-8707
(513) 627-1430
Horst Spielmann
P.O. Box 330013
D-100 Berlin 33, Germany
Eugene Elmore
National Institute for the Advancement of In Vitro Sciences (NIAIS)
4199 Campus Drive, #550
Irvine, CA 92715
(714) 854-0426
Makoto Umeda
Yokohama City University
Kihara Institute for Biological Research
Nakamura-cho 2-120-3
Yokohama, 232 Japan
John M. Frazier
Environmental Health Sciences
Division of Toxicological Sciences
Johns Hopkins University
615 North Wolfe Street
Baltimore, MD 21205
(410) 955-4689
Neil Wilcox
U.S. Food and Drug Administration/CVM
RM #7-57, HFV-4
5600 Fishers Lane
Rockville, MD 20857
(301) 541-7537
Jack M. Lipman (Substitute for Emil Pfitzer)
Hoffmann-LaRoche Inc.
340 Kingsland Street
Nutley, New Jersey 07110
(201) 235-3028

Workshop Participants -- ConsultedObservers
Michael Balls
Fund for the Replacement of
Animals in Medical Experimentation
Eastgate House
34 Stoney Street
Nottingham, NF1-1NB
London, England
George C. Becking
International Program for Chemical Safety
World Health Organization
P.O. Box 12233, NIEHS
Research Park, NC 27709
(919) 541-7537
Stephen D. Gettings
Cosmetic, Toiletries, and Fragrances Association
1101 17th Street, Suite 300
Washington, D.C. 20036
(202) 331-1770
Patrick Murphy
Commision of the European Community
200 Rue De La Loi
1049 Brussels, Belgium
Emil A. Pfitzer
Hoffmann-Laroche Inc.
340 Kingsland Street
Nutley, NJ 07110
(201) 235-3028
Ian Purchase
Imperial Chemical Industries
Aderley Park, Masslesfield
Cheshire, SK10 4TJ
United Kingdom
John Yam
The Procter & Gamble Company
P.O. Box 398707
Cincinnati, OH 45239
(513) 245-1307
Charles Crepi
GENTEST Corporation
Janis Demitrulius
The Dial Corporation
Barbara Doonan
UST, Inc.
Bjorn Ekwall
Uppsala University, Sweden
Robert Finch
US Army Biomedical, R&D Lab
John Harbell
Microbiological Associates, Inc.
Robert Hay
Cell Culture Department, ATCC
Karen Kohrman
The Procter & Gamble Company
Dennis Laska
Elia Lilly & Co.
Pam Logemann
Advanced Tissue Sciences
Arthur Messier
Naval Sub Medical Research Lab
US Naval Submarine Base
Masaaki Nakamura
Osaka Dental University
Kym O'Brien
Unilever Research
Hiroshi Oshima
Osaka Dental University
Andrea Pfeifer
Toxicological Department
Nestec Research Center
Dennis Triglia
Advanced Tissue Sciences
Kathleen Wallace
Microbiological Associates, Inc.