Toxicologic Pathology

The ILSI/HESI Workshopon Alternatives to Carcinogenicity Testing aims to develop and apply new methods for assessment of potential carcinogenic risk to humans from various chemicals. The Workshoprepresents a major cooperative scientific effort. The long-term goals should be to greatly enhance the effi ciency and reliability of such testing and to supplant, not just supplement, lifetime rodent bioassays. There are now well-established frameworks for risk assessment and risk management, putting risks into public health context and engaging stakeholders. The Lave-Omenn value-of-information model provides a useful way to assess the social costs and benefits of different strategies for testing large numbers of chemicals.

The willingness of the agencies involved in the regulation of pharmaceuticals to accept data from newly proposed models for carcinogenicity testing (eg, transgenic animals, neonatal rodent models, initiation-promotionmodels) has stimulated international interest in gaining experience and a greater understanding of the strengths and limitations of the specifi c models. Over a 4-year period, the International Life Sciences Institute (ILSI) Health and Environmental Science Institute (HESI) has coordinated a large-scale collaborative research program to help to better characterize the responsiveness of several models proposedfor use in carcinogenicity assessment. The overall objective of this partnership among industry, government, and academic scientists was to evaluate the ability of these new models to provide useful information for human cancer risk assessment. This research program reflected a commitment of nearly US$35 million by over 50 industrial, government, and academic laboratories from the United States, Europe, and Japan. Evaluation of the models required the development of standardized protocols to allow reproducibility and comparability of data obtained across multiple laboratories. Test compounds were selected on the basis of mechanistically meaningful carcinogenic activity or noncarcinogenicity in the rodent bioassay as well as humans. Criteria were established for dose selection, pathology review, quality control, and for evaluation of study outcome. The database from these studies represents an important contribution to the future application of new models for human cancer risk assessment. Beyond the data, the collaborative process by which the models were evaluated may also represent a prototype for assessing new methods in the future.

The generation, evaluation, and presentation of data from the ILSI Alternatives to Carcinogenicity Testing (ACT) program was standardized to ensure that the results of studies performed in multiple laboratories could be reliably compared. To this end, standardized experimental protocols, tissue collection procedures, histopathology nomenclature, diagnoses, and terminology were employed by study participants. In the experimental phase, this approach provided important cross-model consistency. To ensure comparability in the data evaluation phase of the project, interpretive criteria were defi ned to allow the characterization of study outcome as positive, negative, or equivocal in regards to carcinogenic response. These criteria helped to provide consistency across models because separate Assay Working Groups were established to evaluate the results of each model. To organize and compile the data from the ILSI ACT program, a database has been developed and data entered in standardized format to facilitate cross- and intramodel comparisons. In summary, the early development of standardized test protocols, evaluation procedures, and interpretive criteria has resulted in a data set in which users can have a high level of assurance that results in the database refl ect consistently applied experimental and interpretive guidelines.

The heterozygous Trp 53 null allele C57BL/6 (N5) mouse is susceptible to the rapid development of neoplasia by mutagenic carcinogens relative to control strains. This mouse model of chemical carcinogenesis demonstrates 1) dose-related rapid induction of tumors (26 wks), 2) multiple sites of carcinogen-specifi c tissue susceptibility, and 3) carcinogen-induced loss of heterozygosity involving the Trp53 wild-type allele or a p53 haploinsufficiency permitting mutation of other critical protooncogenes and/or inactivation of tumor suppressor genes driving tumorigenesis. Demonstration of mutation or loss ofheterozygosity involving the Trp53 locus is consistent with a common finding in human cancers and supports extrapolation between rodents and humans. Using diverse experimental protocols, almost all mutagenic rodent carcinogens (including all mutagens that are carcinogenic to humans), but not nonmutagenic rodent carcinogens, induce tumors within 26 weeks of continuous exposure. These characteristics and results indicate that the mouse heterozygous for the Trp53 null allele may be of significant use for the prospective identification of mutagenic carcinogens of potential risk to human health.

The performance of the p53 +/- transgenic (knockout) mouse model was evaluated through review of the data from 31 short-term carcinogenicity studies with 21 compounds tested as part of the International Life Sciences Institute's (ILSI) Alternatives to Carcinogenicity Testing (ACT) project, together with data from other studies which used comparable protocols. As expected based on the hypothesis for the model, a significant number (12/16 or 75%) of the genotoxic human and/or rodent carcinogens tested were positive and the positive control, p-cresidine, gave reproducible responses across laboratories (18/19 studies positive in bladder). An immunosuppressive human carcinogen, cyclosporin A, was positive for lymphomas but produced a similar response in wild type mice. Two hormones that are human tumorigens, diethylstilbestrol and 17β-estradiol, gave positive and equivocal results, respectively, in the pituitary with p53-defi cient mice showing a greater incidence of proliferative lesions than wild type. None of the 22 nongenotoxic rodent carcinogens that have been tested produced a positive response but 2 compounds in this category, chloroform and diethylhexylphthalate, were judged equivocal based on effects in liver and kidney respectively. Four genotoxic noncarcinogens and 6 nongenotoxic, noncarcinogens were also negative. In total (excluding compounds with equivocal results), 42 of 48 compounds or 88% gave results that were concordant with expectations. The technical lessons learned from the ILSI ACT-sponsored testing in the p53+/- model are discussed.

The Tg.AC (v-Ha-ras) transgenic mouse model provides a reporter phenotype of skin papillomas in response to either genotoxic or nongenotoxic carcinogens. In common with the conventional bioassay, the Tg.AC model responds to known human carcinogens and does not respond to noncarcinogens. It also does not respond to most chemicals that are positive in conventional bioassays principally at sites of high spontaneous tumor incidence. The mechanism of response of the Tg.AC model is related to the structure and genomic position of the transgene and the induction of transgene expression through specific mediated interactions between the chemicals and target cells in the skin.

In a Government/Industry/Academic partnership to evaluate alternative approaches to carcinogenicity testing, 21 pharmaceutical agents representing a variety of chemical and pharmacological classes and possessing known human and or rodent carcinogenic potential were selected for study in several rodent models. The studies from this partnership project, coordinated by the International Life Sciences Institute, provide additional data to better understand the models' limitations and sensitivity in identifying carcinogens. The results of these alternative model studies were reviewed by members of Assay Working Groups (AWG) composed of scientists fromgovernmentand industry with expertise in toxicology, genetics, statistics, and pathology. The Tg.AC genetically manipulated mouse was one of the models selected for this project based on previous studies indicating that Tg.AC mice seem to respond to topical application of either mutagenic or nonmutagenic carcinogens with papilloma formation at the site of application. This communication describes the results and AWG interpretations of studies conducted on 14 chemicals administered by the topical and oral (gavage and/or diet) routes to Tg.AC genetically manipulated mice. Cyclosporin A, an immunosuppresant human carcinogen, ethinyl estradiol and diethylstilbestrol (human hormone carcinogens) and clofi brate, an hepatocarcinogenicperoxisome proliferator in rodents, were considered clearly positive in the topical studies. In the oral studies, ethinyl estradiol and diethylstilbestrol were negative, cyclosporin was considered equivocal, and results were not available for the clofibrate study. Of the 3 genotoxic human carcinogens (phenacetin, melphalan, and cyclophosphamide), phenacetin was negative by both the topical and oral routes. Melphalan and cyclophosphamide are, respectively, direct and indirect DNA alkylating agents and topical administration of both caused equivocal responses. With the exception of clofi brate, Tg.AC mice did not exhibit tumor responses to the rodent carcinogens that were putative human noncarcinogens, (di(2-ethylhexyl )phthalate, methapyraline HCl, phenobarbital Na, reserpine, sulfamethoxazole or WY-14643, or the nongenotoxic, noncarcinogen, sulfisoxazole) regardless of route of administration. Based on the observed responses in these studies, it was concluded by the AWG that the Tg.AC model was not overly sensitive and possesses utility as an adjunct to the battery of toxicity studies used to establish human carcinogenic risk.

The rasH2 mouse is a hemizygous transgenic mouse carrying the c-Ha-ras oncogene and that gene's promotor/enhancer within the genetic background of a BALB/cByJ C57BL/6J F1 mouse. Approximately 3 copies of the transgene are integrated in a tandem array into chromosome number 15. The transgene is transmitted stably without point mutation in hot spots and is expressed in all tissues over 20 backcross generations. The homozygous c-Ha-ras genotype is lethal. Hemizygotes are selected by polymerase chain reaction (PCR) analysis of tail tips after birth. Spontaneous tumors in hemizygous transgenic mice are rare until 6 months of age. The observed rasH2 tumor spectrum, including lung adenoma/adenocarcinoma, forestomach and skin papillomas, Harderian gland adenoma, liver proliferative lesions, splenic hemangioma/sarcoma, and lymphoma is consistent with the BALB/c and C57BL/6 background. In the rasH2 mouse, point mutations of the transgene induced by genotoxins are reported frequently but not in all tumors. Elevated levels of transgene expression were detected in all genotoxin-induced tumors in the rasH2. Increased transgene expression was independent of the mutation rate in transgenic and endogenous ras genes. These observations suggest that the overexpression of transgenic c-Ha-ras is responsible for accelerated tumor development.

This article presents data from short-term carcinogenicity studies of compounds tested in the CB6F1-rasH2 transgenic mouse as part of the International Life Sciences Institutes' (ILSI) Health and Environmental Sciences' (HESI) Alternative to Carcinogenicity Testing (ACT) project. Additionally, data from other studies that were not conducted as part of the ILSI program, but used comparable or slightly modifi ed protocols, are included here. A signifi cant number (3 of 4) of the genotoxic carcinogens tested were positive in the rasH2 mouse; the other compound was equivocally positive. The positive control, N-Methyl-N-nitrosurea (MNU), gave reproducible responses across all participating laboratories with tumors noted at multiple sites in the animal. The immunosuppressive human carcinogen, Cyclosporin A, was equivocal. Two hormones that are human tumorigens, Diethylstilbestrol and 17β-Estradiol, gave positive and negative results, respectively. Of the twelve additional compounds tested that are classifi ed as non-genotoxic rodent carcinogens and putative human non-carcinogens, only the two peroxisome proliferators (clofi brate and diethylhexylphthalate(DEHP)) produced a positive response (liver effects). The three non-genotoxic non-carcinogens that were tested also gave negative responses in the rasH2 model. This result provides confi dence that the model is likely to have a low false-positive rate.

Xeroderma pigmentosum (XP) is a rare autosomal recessive disease in which repair of ultraviolet (UV)-induced DNA damage is impaired or is totally absent due to mutations in genes controlling the DNA repair pathway known as nucleotide excision repair (NER). XP is characterized, in part, by extreme sensitivity of the skin to sunlight, and XP patients have a more than 1000-fold increased risk of developing cancer at sun-exposed areas of the skin. To study the role of NER in chemical-induced tumorigenesis in more detail, the authors developed Xpa-/- homozygous knockout mice with a complete defect in NER (designated as Xpa mice or XPA model). Xpa mice develop skin tumors at high frequency when exposed to UV light, and as such, they mimic the phenotype of human XP. Moreover, the Xpa mice also appear to be susceptible to genotoxic carcinogens given orally. Based on these phenotypic characteristics, the Xpa mice were considered to be an attractive candidate mouse model for use in identifying human carcinogens. In an attempt to further increase both the sensitivity and specifi city of the XPA model in carcinogenicity testing, the authors crossed Xpa mice with mice having a heterozygous defect in the tumor suppressor gene p53. Xpa/p53+/- double knockout mice develop tumors earlier and with higher incidences upon exposure to carcinogens as compared to their single knockout counterparts. Here the authors describe the development and features of the Xpa mouse and present some examples of the Xpa and Xpa/p53+/- mouse models' sensitivity towards genotoxic carcinogens. It appeared that the Xpa/p53 +/- double knockout mouse model is favorable over both the Xpa and p53+/- single knockout models in short-term carcinogenicity testing. In addition to the fact that the double knockout mice respond more robustly to carcinogens, they also appear to respond in a very discriminative way. All compounds identified thus far are true (human) carcinogens, and, therefore, the authors believe that the Xpa/p53+/- mouse model is an excellent candidate for a future replacement of the chronic mouse bioassay, at least for certain classes of chemicals.

DNA repair defi cient Xpa-/- and Xpa-/- /p53 +/- knock-out mice in a C57BL/6 genetic background, referred to as respectively the XPA and XPA/p53 model, were investigatedinthe international collaborative research program coordinated by InternationalLife Sciences Institute(ILSI)/Health and Environmental Science Institute. From the selected list of 21 ILSI compounds, 13 were tested in the XPA model, and 10 in the XPA/p53 model. With one exception, all studies had a duration of 9 months (39 weeks). The observed spontaneous tumor incidence for the XPA model after 9 months was comparable to that of wild-type mice (total 6%). For the XPA/p53 model, this was somewhat higher (9%/13% for males/females). The 3 positive control compounds used, B[a]P, p-cresidine, and 2-AAF, gave positive and consistent tumor responses in both the XPA and XPA/p53 model, but no or lower responses in wild-type mice. From the 13 ILSI compounds tested, the single genotoxic carcinogen (phenacetin) was negative in both the XPA and XPA/p53 model. Positive tumor responses were observed for 4 compounds, the immunosuppressant cyclosporin A, the hormone carcinogens DES and estradiol, and the peroxisome proliferator WY-14,643. Negative results were obtained with 5 other nongenotoxic rodent carcinogens, and 2 noncarcinogens tested. As expected, both DNA repair defi cient models respond to genotoxiccarcinogens. Combined with previous results, 6 out of 7 (86%) of the genotoxic human and/or rodent carcinogens tested are positive in the XPA model. The positive results obtained with the 4 mentioned nongenotoxicILSI compounds may point to other carcinogenic mechanisms involved, or may raise some doubts about their true nongenotoxicnature. In general, the XPA/p53 model appears to be more sensitive to carcinogens than the XPA model.

The neonatal mouse model, in various forms, has been used experimentally since 1959 and a large number of chemicals have been tested. The neonatal model is known to be very sensitive for the detection of carcinogens that operate via a genotoxic mode of action. In contrast, it is known not to respond to chemicals that act via epigenetic mechanisms, commonly observed in the two-year carcinogenicity studies. As such, the model has a high sensitivity and specifi city in its response. Dose selection for the neonatal model is based on the maximum tolerated or feasible dose. Traditionally, compounds have been tested via the IP route of administration in this model. In some cases, this has limited the amount of material that can be administered because of the low dosing volumes (10 to 20 μL) that can be administered IP. For the ILSI project, the neonatal model was adapted for oral administration, which has the advantages of being the same route for which most pharmaceuticals are administered. In addition, a 10-fold increase in the volume of administration (100 to 200 μL) and the ability to dose drugs in suspension, permits much higher doses to be used as compared to the IP route of administration. The spontaneous tumors in the neonatal model occurred mainly in the liver of male mice and lung of male and female mice with a few tumors observed in the Harderian gland. The positive control, DEN produced a robust, uniform, and reproducible tumor response with the target organs essentially limited to liver and lung. A total of 13 compounds out of the 21 ILSI ACT compounds were evaluated in the neonatal model involving 18 studies with duplicate studies for some compounds. The genotoxic carcinogens including those used as positive controls were clearly positive (cyclophosphamide, diethylnitrosamine, 6-nitrochrysene). The non-genotoxicrodent carcinogens were clearly negative (chlorpromazine, sulfi soxazole, sulfamethoxazole, clofi brate, DEHP, haloperidol, metaproteranol, and phenobarbital). The non-genotoxic human carcinogen (cyclosporin) was clearly negative. The two other human carcinogens phenacetin and DES were negative and interestingly estradiol was negative in one of the two oral studies, but was clearly positive in the other. Considering the mode of action for three of the human carcinogens (DES, cyclosporin and phenacetin), which were negative in this model, the mode of action in humans is likely to be epigenetic. Overall, for the 3 clearly genotoxic chemicals, all were positive. For the 9 clearly non-genotoxic chemicals, all 9 were negative. The two human carcinogens for which genotoxicity may or may not play a role (DES and phenacetin) were negative and estradiol was positive in 1 of the two oral studies. Overall, the extensive database for compounds tested in the neonatal mouse model would support its use as an alternative model for the assessment of the carcinogenic potential of a chemical. The model responds to chemicals that act via a genotoxic mode of action that represent a greater concern for human cancer risk.

The Syrian hamster embryo (SHE) cell-transformation assay represents a short-term in vitro assay capable of predicting rodent carcinogenicity of chemicals with a high degree of concordance (LeBoeuf et al [1996]. Mutat Res 356: 85—127). The SHE assay models the earliest identifiable stage in carcinogenicity, morphological cell transformation. In contrast to other short-term in vitro assays, both genotoxic and epigenetic carcinogens are detected. The SHE assay, originally developed by Berwald and Sachs (J Natl Cancer Inst 35: 641—661) and modifi ed as described by LeBoeuf and Kerckaert (Carcinogenesis 7: 1431—1440), was included in the International Life Sciences Institute, Health and Environmental Sciences Institute (ILSI/HESI). Alternative Carcinogenicity Testing (ACT) collaboration to provide additional information on the use of short-term in vitro tests in predicting carcinogenic potential. A total of 19 ILSI compounds have been tested in the SHE assay: 15 were tested for this project, whereas clofi brate, methapyrilene, reserpine, and Di(2-ethylhexyl ) phalate (DEHP) were tested previously. Of the 3 noncarcinogenic compounds tested, 2 were negative in the SHE assay, whereas ampicillin was tested positive. The remaining 16 compounds tested were either known rodent carcinogens and/or human carcinogens. From this group, 15 tested positive in the SHE assay whereas phenacetin, a genotoxic carcinogen, was tested negative. Therefore, overall concordance between the SHE assay and rodent bioassay was 89% (17/19), whereas concordance with known or predicted human carcinogens was 37% (7/19). Based on these data, it is concluded that the SHE cell-transformation assay has utility for predicting the results of the rodent carcinogenesis bioassay but lacks the selectivity to distinguish between rodent and human carcinogens.

The p53 tumor suppressor gene has been shown to be critical in preventing cancer in humans and mice. We have generated and extensively characterized p53-deficient mice lacking one (p53+/-) or both (p53-/-) p53 alleles. The p53-defi cient mice are much more susceptible to an array of different tumor types than their wild-type (p53+/+) littermates. The enhanced tumor susceptibility of the p53+/- mice has made them one of several transgenic mouse models that are being considered as substitutes for standard 2-year rodent carcinogenicity assays. In order to fully exploit this model, it will be important to understand some of the basic biological and molecular mechanisms that underlie its enhanced tumor susceptibility. With this in mind, we have explored the fate of the remaining wild-type p53 allele in spontaneously arising p53+/-tumors and have shown that over half of these tumors retain an intact, functional wild-type p53 allele. This suggests that p53 is haploinsufficient for tumor suppression and that mere reduction in p53 dosage is suffi cient to promote cancer formation. To support the idea that p53 is indeed a haploinsufficient tumor suppressor, we show here that normal p53+/-cells exhibit reduced parameters of growth control and stress response compared to their p53+/+ counterparts. We hypothesize that the reduced p53 dosage in the p53+/- cells provides an environment more conducive to the development of further oncogenic lesions and the initiation of a tumor. Finally, we have assessed p53 loss of heterozygosity (LOH) in carcinogen-induced p53+/- tumors and have found that some agents induce tumors that almost invariably exhibitp53 LOH, whereas other agents induce tumors that often retain the wild-type p53 allele. Our preliminary data suggest that LOH is dependent on both the mechanism of genotoxicity of the agent utilized and the tissue type targeted.

The 2-year rodentbioassay has long had a central role in determining whether a compound is carcinogenic. It has recently been suggested that the use of 6-month studies in transgenic mice could reduce costs and animal numbers, without impairing the validity of cancer risk assessment. The p53 +/- hemizygous knockout mouse model is phenotypically stable and develops tumors during the 6-month study period only in response to chemical and physical stimuli, showing a high concordance with genotoxic rodent carcinogens. We treated p53+/- mice and wild-type parent strain (C57BL/6J) animals with diethylstilbestrol (DES), 500 μmol/kg i.p. for 4 days. Following sacrifice, DNA was extracted from various tissues and adducts measured by a modifi ed monophosphate version of the 32P-postlabelling assay. Major DES adducts were detected in the liver DNA of DES-treated wild-type mice at a level of 118.7 17.0 (mean SD relative adduct level [RAL]/10 10 nucleotides) compared with 207.7 36.4 in p53 +/-mice. No such adducts were detected in vehicle-treated animals. Total adduct levels, including endogenous I-compound adducts, in wild-type mice were 192.4 17.5 and 311.5 58.6 in p53 +/-animals. These data support the hypothesis that defi cient p53-dependent global genomic repair of DES adducts in p53+/-mice may result in the persistence of exogenous and endogenous DNA adducts that could contribute to earlier carcinogenicity in this model. We also prepared hepatic microsomes from male and female p53+/-and wild-type mice exposed to DES or vehicle. Western blot analysis demonstrated modestly higher basal levels of various cytochrome P450 (CYP) enzymes in the untreated p53 +/-mice compared to the wild-type mice. Furthermore, P450 levels were higher in female DES-treated p53 +/-mice compared to treated wild-type mice. For the p53 +/- knockout mice to be used with confi dence in drug safety studies, a further understanding of the metabolic capacity of these animals is needed.

Drug-metabolizing enzymes are involved in the metabolic activation or detoxification of carcinogens. To evaluate animals developed as models for alternative carcinogenicity testing, we investigated whether or not a gene manipulation including the transgene of ras and the knocking out of a tumor suppressor gene such as p53 orXPA could alter the expression of representative drug-metabolizing enzymes directly or indirectly. Expression of several isoforms of cytochrome P450 (CYP) in the liver of rasH2, p53 ( +/- ), Tg.AC, and XPA (-/-) mice with or without treatment of prototype inducer, phenobarbital or 3-methylcholanthrene, was analyzed by Western immunoblotting in comparison with their parental strains of mice. In addition, the activities of 3 major phase II enzymes, UDP-glucronosyltransferase, sulfotransferase, and glutathione S-transferase, were compared between the gene-manipulated and the corresponding parental strains of mice. Results demonstrate that XPA gene knockout appeared to increase constitutive expression of CYP2B and CYP3A isoforms. Overexpression of human c-Ha-ras gene or p53 gene knockout appeared to increase constitutive UGT activity toward 4-nitrophenol. The content or activities of almost all other enzymes examined in the present study do not appear to be affected by the gene manipulation.

This perspective is based upon the data presented at the International Life Sciences Institute (ILSI), Health and Environmental Sciences Institute Workshop on the Evaluation of Alternative Methods for Carcinogenicity Testing (ILSI Workshop). It is important to understand that all models discussed at the Workshop have limitations and that they are not designed to be employed as stand-alone assays. Although they may have other, appropriate applications, I do not recommend use of the SHE cell assay and the Tg.AC model for the regulatory purposes of a safety assessment. In my view, the neonatal mouse, p53 +/- , XPA -/- , XPA-/- and p53 +/-, and the rasH2 models can, as a component of an overall assessment, provide information on potential carcinogenicity of a chemical that is appropriate for consideration in a regulatory context. Generally, these models exhibit the ability to detect genotoxic compounds. In most cases these compounds would be detected in a standard battery of genotoxicity tests and, therefore, quite often the use of an alternative is not necessary. Actually, I believe that a bioassay in rats will suffi ce most of the time, that is, in my view, a routine bioassay in mice is not necessary. Specifi c circumstances where data obtained from one of the “recommended” alternative models might be helpful are discussed. With regard to lessons for the future, there is a particular need for models that are responsive to chemicals that exhibit a nongenotoxic mode of action. Additionally, new models will continue to be developed and their half-life will likely be substantially shorter than the time required for traditional validation. The development of enhanced paradigms for validation should be a priority so that improved safety assessment decisions can be made more quickly. However, while evaluating and validating such models, it is important to consider the fundamental issues, for example, rational dose selection, evaluation of mode of action in the context of dose-response relationships including the existence of thresholds and secondary mechanisms, and species-to-species extrapolation. The alternatives to carcinogenicity testing project was a very major undertaking. In addition to the valuable information provided, it serves to illustrate the value of cooperation between academia, government, and industry. Furthermore, the involvement of the International Life Sciences Institute as the overall organizing, facilitating umbrella was crucial for the success of the project.

The results of the present study have advanced dramatically the database on transgenic mouse abbreviated carcinogenicity bioassay models. As such, it will provide a secure foundationfor future evaluations of these assays and for their eventual validation as models for the prediction of possible human carcinogens. Based upon the results derived from the present study, it is suggested that 5 areas require discussion as a prelude to the further evaluation of existing models and the future evaluation of new models. First, there is the need to agree a standard list of calibration chemicals to be studied and to derive agreement on optimal bioassay group sizes, statistical methods, and exposure periods. Second, general agreement must be reached regarding the classes/types of known rodent carcinogens so that it is acceptable for the new models to fi nd negative, by implication, those rodent carcinogens considered not to pose a carcinogenic hazard to humans. Third, current understanding of mechanisms of carcinogenesis should be integrated into the evaluation of new bioassay models. Fourth, any changes made to the standard rodent carcinogenicity bioassay protocol will require compromises being made, and these should be commonly owned between interested parties in order to reduce the number of regional/agencyspecifi c carcinogenicity testing schemes. Fifth, a mechanism needs to be developed by which assays can be adopted or rejected for use in the routine bioassay of chemicals. In the absence of such initiatives the increasing number of new bioassay models will come to exist along side of the standard 2-species bioassay, and this may potentially lead to confusion regarding the true future role of these assays.

Twenty-one chemicals were evaluated by standardized protocols in 6 mouse models that have been sugggested as alternatives to the 2-year mouse bioassay. Included were genotoxic and nongenotoxic chemicals, carcinogens and noncarcinogens, immunosuppressive and estrogenic agents, peroxisome proliferators, and chemicals producing cancer in rodents by other mechanisms. Mice were sacrificed at the end of 6 to 12 months, depending on the model. Standardized histopathology, biostatistical analyses, and criteria for overall evaluation of the results were employed. The TgAC transgenic (dermal and oral administration), the Tg-rasH2 transgenic, the heterozygous p53 gene knockout, the homozygous XPA and homozygous XPA—heterozygous p53 gene knockout, and the neonatal mouse models were evaluated. The chemicals were also evaluated in the in vitro SHE assay. Comparison of the results between the various in vivo models suggest that they might have usefulness as screening bioassays for hazard identification for potential human carcinogens. They have the benefits of being quicker, less expensive, and involve fewer animals than the traditional 2-year mouse bioassay. They do not appear to be overly sensitive. However, they do not definitively distinguish between genotoxic and nongenotoxic carcinogens, and they do not have 100% specifi city for identifying human carcinogens. Like the 2-year bioassay, the results from these models need to be evaluated in conjunction with other information on a chemical in an overall weight-of-evidence, integrated analytical approach to assess risk for human exposures.

The International Life Sciences Institute Alternative Carcinogenicity Testing (ILSI ACT) Workshopconcluded with a panel discussion that addressed the framework issue of the appropriate application of alternative models to human cancer risk assessment. This discussion encompassed both technical issues relating to the level of understanding of these models and their output as well as issues relating to the regulatory acceptance of these data. Although there were many different perspectives represented by the panelists, there was also significant consensus on many broad issues. This article focuses on several key areas of emphasis that were addressed by panelists and are considered critical issues for future discussions and evaluations.