Document 98Mq8y3xyYzmb6zxOVkYMmYe

Rashmi S. Nair FACTO RS IN TH E EV A LU A TIO N OF 200 NATIONAL CANCER IN STITUTE CARCINOGEN BIOASSAYS Kenneth C. Chu, Cipriano Cueto, Jr., Jerrold M. Ward National Toxicology Program, National Cancer Institute, Bethesda, Maryland In determining the carcinogenicity o f a chemical tested in a National Cancer Institute (NCI) bioassay, the following criteria are considered: (1) the adequacy o f the bioassay data, (2) the presence o f significantly increased incidences o f tumors, (3) the adequacy of the number o f animals at risk o f developing tumors, (4) the adequacy o f the dose of chemical administered, (5) the etiology and pathogenesis o f the lesions, and (6) other factors that may influence an evaluation, such as a shortened latency period for tumor formation in dosed animals or the stability o f the chemical. A decision tree for evaluating these factors is presented. A summary o f the results of 200 NCI carcinogen bioassays is also reported. These procedures are presented in the hope that they may sen-e as discussion points for future developments in the field. INTRODUCTION It has been estimated that many human cancers are caused by chemical and environmental factors (Boyland, 1969; Higginson, 1972; Doll, 1977, 1978). A number of methods have been used to determine the potential health risk of chemicals to humans. These methods include epidemiologic studies, animal experiments (carcinogen bioassays), and in vitro studies. The most direct approach is to determine the effects of chemicals on humans. However, the testing of chemicals in humans is ethically unacceptable. After an accidental or unintentional exposure of humans to chemicals there may be a latency period of 20-30 yr (Doll, 1971, 1978) before the appearance of any carcinogenic `'feet, which precludes an immediate epidemiologic evaluation. Although in vitro tests show promise for the future, the carcinogen bioassay is presently the standard method for determining potential carcinogenic risks to humans. For more than 8 yr the National Cancer Institute (NCI) has been directing such tests. The many discussions with members of the NCI Data Evaluation Group are greatly appreciated. Dr. R. A . Griesemer provided additional insight into the subject matter. Supported in part by Public Health Service contract NOI-CP-43350 to Tracor jitco , Inc., and NOl-CP-43257 to E G & G/Mason, Inc. Requests for reprints should be sent to Kenneth C. Chu, National Toxicology Program, Room 3-A-06, 7910 Woodmont Avenue, Bethesda, Maryland 20205. 251 Journal of Toxicology and Environmental Health, 8:251-280, 1981 Copyright 1981 by Hemisphere Publishing Corporation 0098-4108/81/060251 -30$ 2.25 ,N [ j0 3 7 3 5 252 K . C . C H U E T A U ; During this period, NCI has tested more than 200 chemicals in carcinogen bioassays with two species o f rodents. The vast majority of the results o f these studies are published in the NCI carcinogenesis technical report series. There have been more than 190 reports on 200 chemicals. Because of the importance of these reports in assessing the potential health hazards to humans, we will discuss the general criteria used in evaluating the carcinogenicity o f a chemical tested in an NCI carcinogen bioassay. This paper represents the views of the authors and does not necessarily reflect the views of NCI. M A TERIA LS AND METHODS The NCI Carcinogen Bioassay In a typical NCI chronic bioassay for carcinogenicity (Sontag et al., 1976; Page, 1977), the test chemical is administered continuously for 18-24 mo to groups of 50 male and female (C57BL/6N X C3H/HeN) F l(B 6 C 3 F 1 ) hybrid mice and for 20-24 mo to groups of 50 male and female Fischer 344 rats. On past occasions, Osborne-Mendel and Sprague- Dawley rats have also been used. The dosing period may be followed by a 3-6 mo observation period. Generally, for each species, strain, and sex used there are a control group and two treated groups, one given the maximum tolerated dose (M TD ), termed the high-dose group, and the other given half the M TD, termed the low-dose group. The MTD is taken as " .the,,highest dose that causes no more than a 10% weight decrement, as compared to the appropriate control groups, and does not produce mortality, clinical signs of toxicity, or pathologic lesions (other than those that may be related to a neoplastic response) that would be predicted to shorten the animal's natural life span" (Sontag et al., 1976), as determined from the subchronic study. This dose is used to increase the sensitivity of the test by giving the animals as large a dose as possible, consistent with longevity. During the study, the animals that die are necropsied. At the end of the observation period, all surviving animals are sacrificed and necropsied. Diagnoses o f lesions in any of the 25-30 tissues (Sontag et al., 1976) taken from each animal are made by pathologists by microscopic examination. ** General Evaluation of the Bioassay For the purpose of evaluation, an NCI carcinogen bioassay can be divided into four separate experiments. Each involves the high-dose, low-dose, and control groups of a species, strain, and sex. Thus there is an experiment on the male mice, the female mice, the male rats, and the female rats. Correspondingly, the carcinogenicity of a chemical is assessed for each experiment with a species, strain, and sex. For each experiment, one of four possible conclusions can be drawn EVALUATION OF NATIONAL CANCER INSTITUTE CARCINOGEN BIOASSAYS 253 for a species, strain, and sex: (1) the compound is carcinogenic, (2) the compound is a suspected carcinogen, (3) the compound has not been shown to be carcinogenic, or (4) the data are inadequate for further analysis and do not permit a determination o f the carcinogenicity o f the test chemical (inadequate test). Guidelines for evaluating an NCI ' carcinogen bioassay on a species, strain, and sex are presented in Fig. 1. Evaluating a Bioassay One o f the first steps in evaluating a bioassay is to determine whether the data are adequate for analysis and reporting. Irregularities in the FIG URE 1. Guidelines for evaluating an NCI carcinogen bioassay. 254 K. C. CHU ET.A1 experimental design or in the performance o f the bioassay that bias the. tumor incidences in the controls and/or the dosed groups can make.-' further evaluation meaningless. Tw o major causes of invalid bioassay data, are procedural and operational problems. The first problem occurs when the tumor incidence in the control group is unusually high due to flaws in the execution of the bioassay. For example, in the case of methiodal sodium, an unusually high incidence o f leukemia was found in the control group as well as the dosed groups. These high, incidences were, on further ' investigation, linked to the ip route of administration and/or the injection., vehicle, not the chemical. Consequently, the bioassay was considered, inadequate. The second problem occurs when the dosed animals have-a: high degree of autolysis, preventing histopathologic examination of the tissues. Specifically, in the bioassay of imidazole mustard a large number of animals were autolyzed, so that from each group of 25 animals only 5 to 10 animals could be evaluated. The data from this study were considered inappropriate for further analysis and were not reported in the NCI technical report series. ..--yjaA If the bioassay data are adequate for analysis, the next step is to "-\viL-__.... determine whether there are any chemically induced carcinogenic effects-- -i '- Amin*^ that is, whether there are significantly increased incidences of tumors in .. W iv the dosed groups as compared to the controls. This complex process involves evaluations by pathologists, statisticians, and toxicologists and is discussed later. If there is a significant effect, additional complicating factors may alter the final* conclusion about a chemical. Fo r example, if there is a survival problem with the controls, the tumor incidences in the controls may be artificially low, since they were not at risk of developing tumors for as long a period as the dosed animals. In these cases, it may be appropriate to compare the tumor incidences in the dosed groups with historical data on laboratory controls with similar longevity, if other variables are similar. This may change the nature of the evidence so that it supports, not a carcinogenic effect, but only a suspected carcinogenic effect. Another complicating factor is the stability or reactivity of the test chemical. For example, during the bioassay o f dicofol (N CI-TR-90,. 1978), the chemical changed from a solid to a liquid during storage. If such a change suggests some type of decomposition, it would affect the conclusion' drawn from the bioassay data. In the case of dicofol, although the test substance pro duced hepatocellular carcinomas in 77% of the male mice, the substance was considered a suspect carcinogen, rather than a carcinogen, because its exact identity was not known. These examples show that even-, though there is a significant positive effect, complicating factors, may affect the final conclusion on an experiment. Even when there are no significant positive effects, there may be evidence suggestive of a carcinogenic effect. For example, there may be no EVALUATION OF NATIONAL CANCER INSTITUTE CARCINOGEN BIOASSAYS 255 statistically significant tumors, but a progression of proliferative lesions leading to a neoplasm may be seen in the dosed animals. In another case, a rare tumor may be observed in several dosed animals. In a third case, there may be a positive effect in only the low-dose group while the survival in all groups is similar. A ll these cases may lead to the conclusion that the test chemical is a suspected carcinogen, depending on the particular circumstances. If there are no positive or suggestive effects, then one must examine the survival of the animals. A primary issue is the number of animals at risk of developing a tumor relative to the number initially on test. The basic concern is that a sufficient number o f animals remain alive long enough to be at risk of forming late-developing tumors. To-be considered adequate, an experiment that has not shown a chemical to be carcino genic should have groups of animals with greater than 50% survival at 1 y r (52 wk). Fo r example, in the bioassay of methylchloroform (N CI-TR-3, 1977) no positive results were found. However, the survival in most groups was poor. O f the Osborne-Mendel rats used, 32 of 50 (62%) low-dose males, 36 of 50 (72%) high-dose males, 24 of 50 (48%) low-dose females, and 21 of 50 (42%) high-dose females died within 1 y r after the start of the study. O f the mice, 21 of 50 (42%) low-dose males, 25 o f '50 (50%) high-dose males, 9 of 50 (18%) low-dose females, and 20 o f \40 (40%) high-dose females died within 1 year after the start o f the study' "Because o f this poor survival and the absence o f positive results, the study was considered inadequate for a determination of the carcinogenicity of the chemical. On the other hand, if a study is positive, the survival criteria are not applicable. For instances, in the dibromochloropropane (D BCP) study (NCI-TR-28, 1978) the survival of the dosed mice was less than 50% at 1 y r, but the animals died from stomach tumors induced by DBCP. The bioassay is adequate and the compound is carcinogenic in spite of the low survival rate of the animals. Another major concern with studies that show no significant positive results is the adequacy of the administered dose. In NCI bioassays where .he objective is to determine whether a chemical is carcinogenic at the M TD, the concern is whether the animals actually receive the M TD. In these studies, a dose is considered adequate if there is a detectable weight loss of up to 10% in a dosed group relative to the controls. Fo r example, the doses of chlorpropamide administered to Fischer 344 rats were considered adequate since there was a 10% weight depression in the dosed rats compared to the controls (NC1-TR-45, 1978). The administered dose is also considered an MTD if the dosed animals exhibit clinical signs or severe histopathologic toxic effects attributed to the chemical, so that larger doses of the chemical may be inappropriate or life-threatening. An example is the bioassay of photodieldrin (N CI-TR-17, 1977) in B6C3F1 mice and Osborne-Mendel rats. There was no depression o f mean body weights in the closed animals compared to the controls, and there was no' effect of the chemical on mortality o f the dosed animals. However, there were chemically induced clinical signs--convulsions and hyperexcitability~ in the dosed rats and male mice. Thus the dose administered to these. animals was considered adequate. In addition, doses are considered adequate if the dosed animals show a slightly increased mortality com pared to the controls. For example, in the bioassay of emetine (N G I-TR 43, 1978), the male mice did not exhibit any significant tumor incidences, clinical signs, or weight depression attributed to the chemical. However, there was a decrease in survival in the animals fed 1.6 mg/kg (higher doses led to greater mortality) relative to the controls. The median survival of this dosed group was 72 wk, compared to 80 wk for the controls. This difference in survival time was considered to indicate that a sufficient dose had been administered. Finally, doses of chemicals given as 5% of the diet are generally considered adequate (Sontag et al., 1976) whether or not there are any chemically related effects, since larger doses may interfere with the nutrition and survival o f the dosed animals. The bioassay of diaryl anilide yellow (NCI-TR-30, 1978) is a prime example. There were no chemically related lesions, weight depression, mortality, or other adverse effects, but the doses were considered adequate since the chemical was administered as 5 and 2.5% of the diet. Just as there is concern about administering an insufficient dose, there is also concern that too large a dose may be given. Thus the toxicity of the chemical may cause the dosed animals to die before they can develop tumors. For example, when emetine (NC1-TR-43, 1978) was administered to male mice at 6.4 and 3.2 mg/kg, almost all the animals died before the end of the first year. Therefore a dose of 1.6 mg/kg was administered to a third group of mice. The operational and procedural parts of a bioassay are generally evaluated by the toxicologists. Their evaluations are added to those of the pathologists and statisticians in making the final evaluation of an experiment. Interpretation by the Pathologists A principal role of the pathologist is to decide whether there are any chemically induced lesions. Before the tumor data are interpreted, the histopathologic diagnoses are checked for proper protocols, data recording, and consistency. A ll pathology reports are reviewed by a team of pathologists, and discrepancies are resolved' before acceptance for any further evaluation (Ward et al., 1979). If the data are considered unevaluatable (i.e., incomplete) or uninterpretabie, the experiment is termed inadequate to determine the carcinogenicity of the compound. The pathologists review the microscopic slides, make histopathologic diagnoses, and document the findings. They inspect the tumor and nontumor incidence tables and write up the interpretation of the findings. The pathologist's conclusions are based on (1) a review of the EV A LU A TIO N O F N ATIO N AL CANCER INSTITUTE CARCINOGEN BIOASSAYS 257 incidence o f the lesions, (2) knowledge o f the normal variation in the incidence of the lesions, and (3) an opinion on whether the lesions are induced by the chemical. The latter is determined by answering several questions: 1. Are the lesions morphologically different from naturally occurring lesions? 2. Are the lesions rare in control animals? 3. What are the pathogenesis and histogenesis o f the lesions (e.g., are toxic lesions, hyperplasias, dysplasias, adenomas, and carcinomas of a tissue related)? The histogenesis of a neoplastic lesion is its progression from its nonneoplastic origin, often hyperplasia, to a metastatic cancer that kills the animal. For example, the morphology and histogenesis of rat liver tumors were reviewed by Squire and Levitt (1975). Beginning with foci of cellular alteration (hyperplasia), the lesion progresses to neoplastic nodule, to hepatocellular carcinoma, and eventually to metastatic carcinoma. In certain cases, the morphology of a chemically induced lesion can be differentiated from that of a naturally occurring lesion, e.g., "induced" squamous cell carcinoma of the lung as compared to the usual "spon taneous" alveolar-bronchiolar tumor. From these types of considerations, the pathologist draws a conclusion about the significance of the tumors seen in the experiment. For example, in the bioassay of captan in B6C3F1 mice (NCI-TR-15, 1978), there was no statistically significant increase in duodenal tumors; the incidences were 1 in 43 of the low-dose and 3 in 46 of the high-dose male mice, and 0 in 49 of the low-dose and 3 in 48 of the high-dose female mice. However, the pathologists felt that this rare tumor was related to administration of the chemical. The chemical was called a suspected carcinogen. Another study is in progress to confirm this finding. A major issue in determining the carcinogenicity of a compound is the nature of the significant tumor types (Goodman et al., 1979; Ward et al., 1979). If malignant tumors or a combination .of malignant and benign tumors are produced, then the compound is considered carcinogenic to the animals. If the significant result is only the production of benign tumors, then the compound may pose a potential health hazard and is termed a suspected carcinogen or a carcinogen, depending on the nature of the benign tumor. For example, 2,4-dinitrotoluene (NCI-TR-54, 1978) was considered a suspected carcinogen since it induced only benign tumors (fibromas of the skin and subcutaneous tissue in male Fischer 344 rats and fibroadenomas of the mammary gland in females). Ideally, a distinction should be made between truly benign tumors, which never progress to malignancy, and tumors that are in a benign state according to histo pathologic criteria at . the time of diagnosis. Scientific judgments in this area are limited by inability to predict the biological behavior of a lesion on the basis o f morphological criteria, but it appears that there are few, if any, truly benign tumors in rodents. If this were true, all chemicals that induce benign tumors would be termed carcinogens. , Statistical Analysis A statistical analysis o f tumor incidences is performed to determine whether the proportion o f dosed animals with a tumor is different from the proportion of tumor-bearing animals in the controls. Accordingly, the incidence of neoplastic lesions is expressed as the ratio of th number of animals bearing such lesions at a specific anatomic site to the number of animals in which that site was examined. For example, if 50 animals were necropsied but only 49 liver tissues were examined, the denominator for the proportion with liver tumors, such as hepatocellular carcinomas, is 49. However, when gross examination was required to detect lesions for histological examination (e.g., skin or mammary gland) or when lesions could have appeared at multiple sites (e.g., lymphomas), the denominator is the number of animals necropsied. In almost all NCI bioassays, there are no serial sacrifices (no animals are killed at predetermined intervals before the termination o f the bioassay), so that the reported "tim e to tumor" is the time of death of the animal with the tumor, i.e., the time to the detection of the tumor. If the tumor is life-threatening, then the time of death may be a good approximation for the time of tumor appearance. This is the case for leukemias and lymphomas, gastrointestinal cancers that lead to peritoneal mtastass, and upper respiratory tract tumors that are associated with asphyxiation or pneumonia. In addition, if the tumors are observable and the time o f their appearance is recorded, as for neoplasms of the skin or mammary gland, then the time to tumor appearance can be obtained. On the other hand, tumors that are not immediately life-threatening are generally detected at terminal sacrifice or at the death of the animal if the chemical is toxic. The tumor incidences and time of death of each animal are used in analyzing the significance of tumor incidences. The basic statistics associ ated with the analysis of these ratios are presented here; a full discussion of the statistical methodology for analyzing bioassay data has been presented elsewhere (Gart et al., 1979). Tum or incidences can be analyzed by the one-tailed Fisher exact test (Cox, 1970), which is used to compare the tumor incidence of a control group to that of a group of dosed animals at a given dose level. When the control is compared to k dosed groups pairwise, a correction to ensure an overall significance level of 0.05 is made. The Bonferroni inequality (Miller, 1966) requires that the/? value for any comparison be less than or equal to 0.05/6 to be significant at an overall level of 0.05, where k = 2 in most cases. Tum or ratios can also be analyzed by the Cochran-Armitage test for a EVALUATION OF NATIONAL CANCER INSTITUTE CARCINOGEN BIOASSAYS \ 259 t J f~ linear trend in proportions with continuity correction (Thomas et al., 1977). Under the assumption o f a linear trend, this test determines whether the slope of the dose-response curve is different from zero at the one-tailed 0.05 level o f significance. This method also provides a two-tailed test of departure from linear trend. When time-to-tumor data are available, life-table methods are also appropriate (Gart et al., 1979; Thomas et a!., 1977). A scheme for the initial evaluation o f these statistical tests is given in Table 1. These evaluations are based' on the assumption that there was adequate and approximately equal survival in each group. The scheme is based on observing how statistical results are finally resolved in the 200 NCI reports. It is presented as an empirical formulation and does not necessarily reflect present or future evaluations. It is stressed that the scheme may be useful as a preliminary evaluation o f the statistical tests. Discussions by the pathologists, toxicologists, and statisticans are necessary for a final evaluation. Although these initial evaluations are, in general, independent o f the spontaneous tumor rate, there is a major exception. This is when the low-dose comparison to controls is significant,, while the high-dose compar ison is not significant. If the tumor type is rare, there may be statistical evidence for a chemically related effect. If the low-dose comparison is highly significant, there may be limited evidence for an effect.' Fo r example, in the bioassay of lindane (N CI-TR-14, 1978), the only significant result was hepatocellular carcinomas and neoplastic nodules in male mice, with 5 in 49 o f the pooled controls, 19 in 49 of the low-dose animals, and 9 in 46 of the high-dose animals. The incidence in the low-dose animals was statistically significant {p -- 0 .0 0 1 ); however, the mean incidence of hepatocellular carcinomas and neoplastic nodules in historical control animals at this laboratory was 23%, with spontaneous incidences as high as 35-40%. Thus it was concluded that the results could not be clearly associated with the chemical, and the substance was called a suspected carcinogen. The heavy reliance on statistical evidence to support a positive effect poses the problem of assessing the error rates associated with these statistics. A recent report on error rates in bioassay results (Fears et al., 1977) reaches a number of informative conclusions. A false positive result can arise because the tumor incidence in the dosed group is greater than that in the controls due to sampling variations in the spontaneous tumor incidence rate and not to the chemical. In effect, a false positive result occurs because the dosed group has a random tumor incidence that is statistically significant. A rare tumor has less chance o f occurring than a common tumor and therefore has less chance of causing a false positive. Accordingly, the chance of a false positive result occurring is highly dependent on the spontaneous tumor incidence rate. In general, tissue sites with spontaneous tumor rate above 5% have considerably higher chances ' % } 260 K. C. CHU ET A l 5 T A B L E 1. Initial Evaluations o f the Statstica) Tests t ~-- 1 - ' ( - 1/ : Stat ist )cal, Cochran-Armtage(1 Fisher exactC2) Fisher exactC2) Conclusioni) < 1 X 1- 5 '/. > 5X Si Si HSi Si HSi P P P P P- P - L - L. P P P P P. P P P P P' -()- - - - . P-1 -L PPP -C4)- _ PP -- P = Kay be positive statistical evidence for carcinogenicity L = Kay be limited statistical evidence for carcinogenicity - * May be in su fficient s t a t i s t i c a l e v i d e n c e for c a r c i n o g e n i c i t y (1) In the linear trend* " in dicates n o n - s i g n i f i c a n t result or significant result with a significant departure statistic. (2) For the Fisher exact test* m e a n s test is not si g n i f i c a n t . (3) <1% = tumor types with <1% s p o n t a n e o u s tumors; 1-5% = tu mo r types uith 1-5% spontaneous tumors* other = tumor types >5% spontaneous tumors. Si= significant result: .001 < P < .025 ; HSi = highly significant result P <.001. C4) May be s i g n if ic an t if the tumor type is rare and the b i n o m i a l p r o b ab il it y is s i gn if ic an t* see text. of generating false positives. The spontaneous primary tumor^ rates for untreated animals used in the NCI testing program, which are given in Table 2, show that there are only one to five sites with an incidence at least this high. For determining the overall false positive rate, Gart et al. (1979) proposed the formula 1 -- (1 -- p )r, where t is the number of target sites with spontaneous rates above 5% and p is the nominal r al. lominal o3^- ~S*+Hnn O3< (n/) 0 -+ 1 w , tru+ --3 O r*on<d t "0 r* I -u T) -o T A B L E 2, Percent Spontaneous Primary Tumors in Untreated Species Used at NCJ for Carcinogen Bioassays SPECIES MOUSE STRAIN! B6C3F1 SEX! MALE FEMALE NUMBER HECROPSIED (1) 3543 3617 ORGAN/TISSUE BRAIN <. IX .IX SKIN 3.1 1.7 MAMMARY GLAND 1.3 CIRCULATORY SYSTEM (2 2.9 2.4 LUNG/BRONCHI/TRACHEA 13.7 5.2 LIVER 24.6 4.7 PANCREAS < 1 <. 1 STOMACH .4 .4 INTESTINES (3) .5 .2 KIDNEY .3 <. 1 URINARY BLADDER <. 1 <. 1 PREPUTIAL GLAND (4 TESTES (5) .4 NA OVARY NA .9 UTERUS NA 1.6 PITUITARY .3 3.6 ADRENAL 1.4 .6 THYROID 1.0 1.7 PANCREATIC ISLETS .4 .2 BODY CAVITIES .4 .3 LEUKEMIA/LYMPHOMA 10.3 20.6 RAT FISCHER 344 MALE FEMALE 2960 2924 RAT 05B -MNDl MALE FEMALE 270 27 0 .SX .6X 7.S 3.2 1.5 20.9 .7 .4 3.0 1.9 2.2 1.9 .2 .3 .2 .6 .3 .5 .2 I .3 2.4 1.3 82.3 NA NA NA 17.0 14.7 34.9 12.4 5.2 8.2 6.3 3.9 .8 . 2.6 .4 19.9 13.4 -- 8.9 3.7 4.1 1.5 -1-.-1 --- 4 3.3 ,4 1.6 .7 NA NA 7.4 10.4 9.6 3.0 1.9 3.3 ___ 5.9 28.5 2.6 .7 1.9 --- .4 .4 2.6 -1.2 NA 1.5 3.7 20.0 10.0 1T. 1 1.9 .4 1.3 RAT RAT SPR, -DAU. CHAR. RIV. CD MALE FEMALE MALE FEMALE 440 205 184 184 .7X .5X 3.2 3.4 -1-.-4' 39--.0 --- --- .2 -- -- .5 --- .2 1.0 --.-2 2.0 NA NA NA 4.4 5. 9 40.0 .7 2.9 .9 2.0 .5 .5 .5 1.9 4.3 0.5 Z.7X 7. 1 .5 2.2 1.6 -.-5 ---- 1.6 .5 3.9 NA NA 33.2 7.6 3.8 2.7 2.2 3.3 1.6X 3.3 45. 1 --1.6 2.2 .5 . .5 --- NA .5 3.8 57.6 4.3 ---- 3.3 Cl) STUDIES TERMINATED AT 21-25 MONTHS FOR MICE AND 23-25 MONTHS FOR RATS. (2) HEMANGIOMA AND HEMANGIOSARCOMA. t (3) DUODENUM, JEJUNM, ILEUM, CEDUM AND COLON. (A) CLITORAL GLAND IN FEMALES. (5) SEMINAL VISICLE AND TESTIS. L 262 K - C C H U E T A L ^ I 'tiT&vsEf significance level. If there are 50 animals in the control and dosed groups and the Fisher exact test yields a p of 0.028 for 5 dosed animals with a tumor compared to zero in the controls, the false positive rate for female F344 rats is less than 0.134 (Gart et al., 1979). A possible false positive is characterized by several features. The bioassay of the four experiments has a statistically significant result in only one species, strain, and sex. More specifically, the p value o f the result is in the range 0.005-0.05 and the tumor type has a spontaneous rate above 5%. On the other hand, a statistically significant result in a species, strain, and sex may be considered an effect of the chemical if the tumor type is rare, the incidence in the dosed group is much higher than the spontaneous incidence, or the significance of the tumor is confirmed in the opposite sex or a different species. As stated earlier, the scheme in Table 1 may be useful for initial evaluations. Discussions by the pathologists, toxicologists, and statisticians will determine the final conclusion. Under certain circumstances, the pathologist has indicated that an incidence of a rare tumor is significant, but this result cannot be verified by the statistical methods discussed due '''.'Tv. to the small sample size. To quantify the assertion made by the pathologist, one can calculate the binomial probability that the observed number or a greater number o f tumors will occur (Fears et al., 1977). If one assumes that the tumor incidence is binomially distributed, the probability of obtaining R tumor in N animals is given by:' N \p ` ( i - P f where p is the laboratory historical tumor incidence. When this probability was less than 0.001, it was used to indicate that a substance may be a suspected carcinogen in the particular species, strain, and sex. In other circumstances, the pathologist may indicate that certain types of tumors for a particular species, strain, and sex may have spontaneous incidences that are so high and variable that the use of these tumors as end points is d ifficu lt In these cases, statistically significant results may be nullified. For an extensive discussion of the variability of common naturally occurring lesions, see Tarone et al. (1981). Two examples in F344 rats are interstitial cell tumors of the testes in males, which have an incidence range of 70-100%, and pituitary chromophobe adenomas in females, which have an incidence range of 30-50%. ' In certain other circumstances, discussions between the scientists may indicate that statistically significant results are not meaningful since the incidences in the dosed groups are not above the normal range of incidences in controls in this laboratory. Results that are statistically significant would then be interpreted as not important. This was the case EVA LU A TIO N O F N ATIO N AL CANCER INSTITUTE CARCINOGEN BIOASSAVS 263 in the bioassay of malathion (N CI-TR-24, 1978) in male mice. There was a significant dose-related trend in the incidence of hepatocellular carcinomas, with 5 in 49, 7 in 48, and 11 in 50 o f the control, low-dose, and high-dose animals, respectively. However, several control groups from the same laboratory had incidences of 35-40% , which are comparable to the incidences in the dosed male mice. The principles described in the sections above were'used to evaluate some 200 chemicals. Table 3 shows chemicals that were positive in both sexes of both species; Table 4, chemicals that were positive in at least one sex of both species; and Table 5, chemicals that were positive in one species. In Table 6 the chemicals with suspected carcinogenic effects are listed: Finally, chemicals not shown to be carcinogenic are listed in Table 7 and chemicals with inadequate bioassays are given in Table 8. D IS C U S S IO N . The purposes of this paper are threefold: (1) to explain some of the factors involved in determining the carcinogenicity of a test chemical; (2) .to present summary results for 200 chemicals recently evaluated by the National Toxicology Program at N C I; (3) to explain the principles underlying the bioassay procedures. The first two objectives have been addressed, and the principles of testing will now be discussed. The appropriateness of carcinogen bioassays is based on several char acteristics o f carcinogenicity. (1) Almost all human chemical carcinogens (except possibly As) are animal carcinogens. (2) It is ethically un acceptable to perform carcinogenicity tests on the human population. (3) A number of human carcinogens were first detected as animal carcinogens, i.e., vinyl chloride, Be, Cd, and diethyIstil bestrol (D ES). (4) Due to the small groups o f animals, usually 50-100, used in carcinogen bioassays (designed to detect 5-10% tumor incidences in the dosed groups as compared to none in the controls), the bioassays are relatively insensitive. From these facts the following premises for carcinogen bioassays have been developed. (1) Long-term effects of toxicity studies in animals are applicable to humans. (2) Due to the relative insensitivity of bioassays, exposure of experimental animals to high doses of toxic agents is a necessary and valid method of assessing potential carcinogenic hazards in humans. In the present analysis, the evaluation o f the carcinogenicity o f a chemical is based on the following premises. (1) A carcinogen is any agent that induces malignant or malignant and benign tumors in a valid carcinogen bioassay. (2) Any agent that induces only benign tumors in a valid carcinogen bioassay may be a carcinogen or a suspected carcinogen, depending on the etiology and pathogenesis of the lesions. (3) The NCI carcinogen bioassays (three groups of 50 animals per sex and species) are relatively insensitive, since they are designed to detect levels of tumor T A B L E 3. NCI Carcinogen Bioassays with Positive .Evidence for Carcinogenicity in Both Sexes of Both Species Tested NCI HO. CHEMICAL NAME CAS NUMBER REPORT STATUS ANIMAL ROUTE ORGAN SITES STRAIN CQ3093 3-Amno-9-Ethylcarbazolo, Ilydrochlori do 132-32-1 C03797 o-Anisidine Hydrochloride 139-29-2 COO 191 C03292 Chlordocone CKcpono) 193-50-0 9-Chi oro-o-Phertyl ortadi ami no 95-83-0 C02982 p-Cresidino 120-71-8 C03258 Cupferron 135-20-6 CO 1989 2 ,9-Diaminoanisolo Sulfate 39 156-91-7 (615-05-9) CO 0500A Dibromochloropropane 96-12-8 C00522A 1,2-Di bromoothono 106-93-9 C005T1 12-Dichloroeth.'ne 107-06-2 . REPORT 93 F399 PUBLISHED B6C3F1 FEED 1; REPORT 89 published F399 B6C3F1 FEED REPORT A PUBLISHED REPORT 63 PUBLISHED OSB-MDL B6C3F1 F399 D6C3F1 FEED FEED REPORT 192 F349 FEED PUBLISHED B6C3F t REPORT 100 F349 . FEED PUBLISHED B6C3F1 REPORT 89 PUBLISHED F399 B6C3F1 FEED REPORT' 28 OSB-MDL CAV PUBLISHED B6C3FI' REPDRT 86 published OSB-MDL B6C3F1 * CAV REPORT 55 OSB-MDL GAV PUBLISHED BCC3F1 LIVER, HEPATOCELLULAR TUMORS SKIN/SUBCUTANEOUS TISSUE ZYMBAL 5 GLAND UTERUS NON-SARCOMA KIDNEY BLADDER, URINARY THYROID FOLLICULAR LIVER, HEPATOCELLULAR TUMORS BLADDER, URINARY FORESTOMACH LIVER, HEPATOCELLULAR TUMORS BLADDER, URINARY NASAL LIVER, HEPATOCELLULAR TUMORS HEMANGIOSARCOMA LIVER, HEPATOCELLULAR TUMORS FORESTOMACH ZYMBAL 5 GLAND HARDERIAN GLAND SKIN/SUBCUTANEOUS TISSUE THYROID FOLLICULAR ZYMBAL S GLAND THYROID C-CELL PREPUTIAL GLAND FORESTOMACH MAMMARY GLAND FORESTOMACH HEMANGIOSARCOMA LIVER, HEPATOCELLULAR TUMORS LUNG FORESTOMACH Hs'E}MAfuNGrIcOhSaARCOMA RAT MOUSE M F 'M F PP P PP P PP P PP P PP PP PP PP PP PP PP PP P PP P PP PP PP P PP P P P P PP PP P P PP PP PP P PP P P PP PP P P p 264 tf M fispc;ia r 'A MAMMARY GLAND - PP LUIIG pp UTERUS SARCOMA p 265 C03689B CO 1923 p-Dioxann 123-91-1 Z-Methyl-1-n itroanthraqui none 1Z9-15-7 C02006 Mlchler's Ketone 90-94-6 CO 166 1 CO 1810S Phenoxybenzamine Hydrochloride 63-92-3 Procarbazi ne 366-70-1 C00453 Sulfallate 95-06-7 CO 1707 4,4'-Thiodianiline 139-65-1 CO 1649 Thlo-TEPA 52-24-4 C02335 o-Toluidtne Hydrochloride 636-21-5 C03270 Trla (2,3-Dibro-nopropyl) Phosphate 126-72-7 REPORT 80 PUBLISHED REPORT 29 PUBLISHED OSB-MDL B6C3F1 F344 B6C3F1 WATER FEED NASAL LIVER, HEPATOCELLULAR TUMORS SO FIBROMA LIVER, HEPATOCELLULAR TUMORS FORESTOMACH BLADDER, URINARY HEMAHGIOSARCOMA PP P PP P S s pp pp REPORT 181 PUBLISHED F344 B63F1 FEED LIVER, HEPATOCELLULAR TUMORS HEMANGIOSARCOMA Pp p p REPORT 72 SPR-DAW IP/ZJ PUBLISHED B6C3F1 PERITONEAL SARCOMA Pp pp REPORT 19 PUBLISHED SPR-DAW B6C3F1 IP/IJ NASAL LYMPHOMA MAMMARY GLAND LUNG UTERUS NDH-SARCOMA Pp Pp p ss pp p REPORT 115 PUBLISHED OSB-MDL B6C3F1 FEED ' FORESTOMACH MAMMARY GLAND LUNG P p p p REPORT 47 PUBLISHED F344 B6C3F1 FEED LIVER. HEPATOCELLULAR TUMORS P COLON P .THYROID FOLLICULAR ;P p ZYMBAL S GLAND Pp UTERUS NON-SARCOMA p pp pp REPORT 58 SPR-DAW PUBLISHED B6C3F1 ` REPORT 153 F344 PUBLISHED B6C3F1 ;i REPORT 76 F344 PUBLISHED B6C3F1 IP/IJ FEED FEED SKIN/SUBCUTANEOUS TISSUE HEMATOPOIETIC SYSTEM NASAL UTERUS NON-SARCOMA SPLENIC SARCOMA MESOTHELIOMA 5Q FIBROMA BLADDER, URINARY MAMMARY GLAND HEMANGIOSARCOMA LIVER, HEPATOCELLULAR TUMORS KIDNEY LUNG FORESTOMACH LIVER, HEPATOCELLULAR TUMORS Pp P Ss s p pp Pp P P p p p \p Pp 's'- p pp pp P F344 = Fischer 344 rat OSB-MDL = Osborne Mendel rat SPR-DAW = Sprague Dawlcy rat GAV = Gavaqe I P / I J ' - Intraporltoneal injection P = Positive evidence for carcinogenicity S = Suqqeative evidence for careinoqenictty N = Not shown to bo carcinogenic I = Inadequate test T A B L E 4. NCI Carcinogen Bioassays with Positive Evidence for Carcinogenicity In at Least One Sex o f Both Species Tested 991 NCI NO. CHEMICAL NAME CAS NUMDER CO 1876 2 "Ami noanthraqui none 117--79-3 comt 1-Ami no-2-Mofchylanlhraquinone 82-28-0 C02686 Chloroform 67-66-3 CD3833 C03305 3-(Chloromothy1)Pyrt di no Hydrochlor ida 6959-68-6 4-Chloro-m-Phenyl'enedi ami ne 5131-60-2 C02302 2,4-Diaminotoluena 95-80-7 CO 1854 Hydrazobonzcna 122-66-7 CO 1627 ICRF-159 21416-67-5 CO 1638 Isophosphami do 3778-73-2 / CO <990 4 ,4'-Meihylcnebln(H*N-Dimethyl)Banzon- amino 101-61-1 C03021 1,5-Haphlhalonodiami no 2243^62-1 REPORT STATUS ANIMAL ROUTE ORGAN SITES STRAIN REPORT 144 PUBLISHED REPORT 111 PUBLISHED REPORT 0 PUBLISHED REPORT 95 PUBLISHED REPORT 85 PUBLISHED REPORT 162 PUBLISHED REPORT 92 PUBLISHED REPORT 78 PUBLISHED REPORT 32 PUBLISHED REPORT 186 PUBLISHED REPORT t43 PUBLISHED F344 B6C3F1 F344 B6C3F1 OSB-MDL D6C3F1 F344 B6C3F1 F344 B6C3F1 F344 B6C3F1 F344 B6C3F1 SPR-DAW B6C3F1 SPR-DAW D6C3F1 F344 B6C3F1 F344 B6C3F1 FEED FEED GAV GAV FEED FEED FEED IP/IJ IP/IJ FEED FEED LIVER, HEPATOCELLULAR TUMORS LYMPHOMA KIDNEY LIVER, HEPATOCELLULAR TUMORS KIDNEY LIVER, HEPATOCELLULAR TUMORS FORESTOMACH ADRENAL MEDULLA LIVER, HEPATOCELLULAR TUMORS LIVER, HEPATOCELLULAR TUMORS SO FIBROMA MAMMARY GLAND LYMPHOMA LIVER, HEPATOCELLULAR TUMORS ZYMDAL S GLAND MAMMARY GLAND UTERUS NON-SARCOMA LYMPHOMA LYMPHOMA UTERUS NOH-SARCOMA MAMMARY GLAND THYROID FOLLICULAR LIVER, HEPATOCELLULAR TUMORS CUTORAL GLAHD UTERU5 SARCOMA THYROID FOLLICULAR LIVER, HEPATOCELLULAR TUMORS LUNG THYROID C-CEIL RAT MOUSE MF MF H P PP P P PP N P N P H P PP P P PP P P H P S P S S P PP PP H P N P S N P H P N P PP SP H P P PP P P \P C030Z1 ii3"(iapnnw*> 2243-62-1 PUBLISHED i\ \Vv, < bbc^ri u t e r u s SARCOMA THYROID FOLLICULAR LIVER, HEPATOCELLULAR LUNG THYROID C-CELL ._ TUMORS r hr Br p. Pr p - iiti/, .f. C0379Z Ntthiazldo 139-94-6 C02766 Nitrilotriacotic Acid (NTA) 139-13-9 CO 1967 5-NI troacanaphthene 602-87-9 CO 1934 5-Kitro-o-Aniaiding 99-59-2 REPORT 146 PUBLISHED REPORT 6 PUBLISHED REPORT 118 PUBLISHED REPORT 127 PUBLISHED F344 B6C3F1 F344 . B6C3F1 F344 B6C3F1 F344 B6C3F1 FEED FEED FEED FEED SKIN/SUBCUTANEOUS TISSUE LIVER, HEPATOCELLULAR TUMORS BLADDER, URINARY KIDNEY ZYMBAL S GLAHD LUNG CLITORAL GLAND MAMMARY GLAND LIVER, HEPATOCELLULAR TUMORS OVARY LIVER, HEPATOCELLULAR TUMORS SKIH/SUBCUTAHEOUS TISSUE ZYMDAL S GLAND CLITORAL GLAND N P PS N P P N PP PP P P N P P PP PP P SP C00420B Nitrofen 1836-75-5 REPORT 26 OSB-MDL FEED PUBLISHED B6C3F1 C 02244 CO 1672 P-N1 trosodI phenylamina 156-10-5 Phenazopyri ding Hydrochloride 136-40-3 CO 1558 Phonesteri n 3546-10-9 C50157 Reaerpine 50-55-5 REPORT 190 PUBLISHED REPORT 99 PUBLISHED F344 B6C3F1 F344 B6C3F1 FEED FEED REPORT 60 SPR-DAU GAV PUBLISHED D6C3F1 REPORT ; IN REVIEW 4 F44 B6C3F1 FEED C02904 2,4,6-Tri chiorophonol 88-06-2 REPORT 1^5 E344 PUBLISHED B6C3F1 FEED C0229 9 2,4.5-TrimethyIoni lino 137-17-7 REPORT 160 PUBLISHED F344 D6C3F1 FEED C03781 Trimethyphosphate 512-56-1 REPORT 81 PUBLISHED F344 B6C3F1 FEED F344 = Fischer 344 rat 05B-MDL = Osborne Mendel rat SPR-DAU = Sprague Dawley rat GAV = Gavaqe, 1P/IJ = Intraperitncal injection P = Positive evidence for carcinogenicity PANCREAS ISLET-CELL LIVER, HEPATOCELLULAR TUMORS HEMAIIGIO SARCOMA I P PP P LIVER, HEPATOCELLULAR TUMORS NN PP COLON LIVER, HEPATOCELLULAR TUMORS PP H P MAMMARY GLAND LUNG LYMPHOMA HEART SARCOMA N P PP PP PP ADRENAL MEDULLA SEMINAL VESICLE MAMMARY GLAND N P S P HEMATOPOIETIC 5YSTEM LIVER, HEPATOCELLULAR TUMORS N P PP LIVER, HEPATOCELLULAR TUMORS LUNG PP P N* X. P SQ FIBROMA UTERUS NON-SARCOMA N P N. P S = Suqqostivo evidence for careinoqcnicity N = Not shown to be:carcinogenic I = Inadequate test ffe '. *'.**;i 4J.*. - T A B L E 5. NCI Carcinogen Bioassays with Positive Evidence for Carcinogenicity in Only One Species NCI NO. CHEMICAL NAME CAS NUMBER REPORT STATUS ANIMAL ROUTE 0RGAH SITES STRAIN RAT MOUSE MF MF CO 1536 Acronyci no 7008-42-6 C00044 Aldr in 309-00-2 CO <887 C03963 C03065 C03736 3-Am ino-4-Ethoxyaeotan 111 do 17026-8 1-2 4-Ami no-2-N itrophonol 119-34-6 2-Ami no-5-Ni trothi azole 121-66-4 Aniline Hydrochloride 142-04-1 CO 1569 C02926 5-Azacyt1di no 320-67-2 Azobenzeno 103-33-3 C00077 Captan 133-06-2 CO0055 . Chlorambon 133-90-4 COO 099 Chlordana 57-74-9 , C004 DBA Chlorobenzilate ^ 510-15-6 COO 102 Chlorothaloni1 1897-45-6 C02051 5-Chloro-o-Tolui dine 95-79-4 REPORT 49 PUBLISHED REPORT 21 PUBLISHED REPORT 112 PUBLISHED REPORT 94 PUBLISHED REPORT 53 PUBLISHED REPORT 130 PUBLISHED REPORT 42 PUBLISHED REPORT 154 PUBLISHED REPORT 15 PUBLISHED REPORT 25 PUBLISHED REPORT B p u b l i s h e d REPORT 73 ` PUBLISHED REPORT 4 1 PUBLISHED REPORT 187 PUBLISHED SPR-DAU B6C3F1 OSD-MDL B6C3F1 F344 D6C3F1 F344 B6C3F1 F344 B6C3F1 F344 B6C3F1 SPR-DAU B6C3F1 F344 B6C3FI OSB-MDL D6C3FI 05B-MDL B6C3F1 DSD-MDL B6C3F1 OSB-MDL B6C3F1 OSB-MDL B6C3F1 F344 B6C3FT IP/IJ FEED FEED FEED FEED FEED IP/IJ FEED FEED FEED FEED FEED FEED FEED OSTEOSARCOMA PERITONEAL SARCOMA MAMMARY GLAND THYROID FOLLICULAR ADRENAL CORTEX LIVER, HEPATOCELLULAR TUMORS THYROID FOLLICULAR BLADDER. URINARY HEMATOPOIETIC SYSTEM HEMANGIOSARCDMA MULTIPLE SITES, SARCOMA SPLENIC SARCOMA HEMATOPOIETIC SYSTEM ABDOMINAL CAVITY SARCOMA MULTIPLE SITES, SARCOMA OSTEOSARCOMA DUODENUM LIVER, HEPATOCELLULAR TUMORS P PP P SS 5 HH PS N P P PP P II PP P S NH NN II N P N P NN NH NK I P NN PS SP FIBROUS HISTIOCYTOMA. THYROID FOLLICULAR LIVER, HEPATOCELLULAR TUMORS S S PP LIVER, HEPATOCELLULAR TUMORS NN PP KIDNEY NN PP H N, HEMANGIOSARCOMA . -P P .* '* * . * V *!* * .. '?,< * .* . .V v`! UWHL-WfBHIWWW m Chlorothalonl1 1897-45-6 5-Chloro-o-Tolui dirt 95-7?-* RPUEPBO^R"T'" ^* 1 REPI' '*7 PUD'-.-..'-3 C02368 C03565 C02993 CO 17 18 4-Chioro-o-Tolu(dt no Hydrochlori da 3165-93-3 C.I. Vat Yellow * 1Z8-66-5 m-Crosi di na 102-50-1 Dapsone 80-08-0 C00555 C03827 p, p '-DDE 7Z-55-9 Diaminozf do 1596-84-5 CO 0486 C03816 COZ 175 D icoiol 1 15-32-Z N ,N'-Diethylthiourea 105-55-5 3 13 *-D imethoxybenzidine4 , 4 '-Di i socyanats 91-93-0 C 0 1865 2i4-D1nltrotoluana 1Z1-14-Z C54557 C54579 C54568 CO 1570 Direct Black 38 1937-37-7 Direct Blue 6 2602-46-2 Direct Brown 95 1607 1-86-6 Estradiol Mustard 22966-79-6 C02857 Ethyl Tellurac 30 145-38- 1 COO 180 Hcptachlor 76-44-8 REPORT 165 PUBLISHED REPORT 134 PUBLISHED REPORT 105 PUBLISHED REPORT 20 PUBLISHED REPORT 131 PUBLISHED REPORT 83 PUBLISHED REPORT 90 PUBLISHED REPORT 149 PUBLISHED REPORT 128 PUBLISHED REPORT 54 PUBLISHED REPORT 108 PUBLISHED REPORT 108 PUBLISHED REPORT 108 PUBLISHED; REPORT 59 PUBLISHED REPORT 152 PUBLISHED REPORT 9 PUBLISHED 0SB-P1DL ` FEED B6C3FI F3AA B6C3F1 FEED KIDNEY HEMANOIQSARCOMA PP NN HN PP -- LIVER HEPATOCELLULAR TUMORS F3A<* B6C3F1 FEED HEMANGIOSARCOMA F34A FEED B6C3F t LYMPHOMA F3AA CAV B6C3FI BLADDER, URINARY F3A4 B6C3F1 FEED SPLENIC SARCOMA PERITONEAL SARCOMA OSB-MDl FEED B6C3F1 ' LIVER, HEPATOCELLULAR TUMORS F3A4 B6C3F1 FEED UTERUS NON-SARCOMA UTERUS SARCOMA LIVER, HEPATOCELLULAR TUMORS OSB-MDl FEED B6C3F1 LIVER, HEPATOCELLULAR TUMORS F344 FEED B6C3F1 THYROID FOLLICULAR F3AA B6C3F1 FEED LYMPHOMA ZYMBAL S GLAND SKIN/SUBCUTANE0U5 TISSUE UTERUS NON-SARCOMA F3^^ B6C3F1 FEED SKIN/5UDCUTANEOU5 TISSUE MAMMARY GLAND ^3',$ B6C3F1 FEED LIVER, HEPATOCELLULAR TUMORS F3A A B6C3F1 FEED LIVER, HEPATOCELLULAR TUMORS B6C3F1 FEED LIVER, HEPATOCELLULAR TUMORS SPR-DAW GAV 6C3FI LUNG HEMAHGIOSARCOMA FORESTOMACH LYMPHOMA F3A A B6C3FI FEED MESOTHELIOMA HARDERIAN GLAND OSB-MDl FEED B6C3F1 LIVER, HEPATOCELLULAR TUMORS NH NN PP H P P NN H P P NN PP PP PP P P P P PP PP N P NN N S NH PP PP N P IN NN PP N S N P NN NH NN NN NN NN PS PP PP PP Ps Pp 270 T A B L E 5. N CL Carcinogen Bloassays with Positive Evidence for Carcinogenicity in Only One Species (continued) i NCI NO. CHEMICAL NAME CAS HUMBER REPORT 51ATU5 ANIMAL ROUTE ORGAN SITES STRAIN CO1601A CO 1517 CO 1178 Hexachioroothane 67-72-1 IPD 3158-22-8 Lasiocarptno 303-31-1 CO 1778 CO 1912 C00120A C02222 C03076 3 ,-Ni tro-p-Acetophaneti da 1777-81-0 6-Ni trobcnztmi dazols 91-52-0 HItrofen 1836-75-5 2-N1tro-p-Phanylenediamino 5307-11-2 3-Nitroproplonto Acid 501-88-1 C02380 N-N ttrosodi phenylami no 86-30-6 CO 1813 5-N1Iro-o-Toloidino 99-55-8 CO 1115A HTA Trlsodium 5aIt.H20 18662-53-8 C02821 C01126 CO 3850 Piperonyl Sulfoxide 120-62-7 Pi valolactono 1955-15-9 p-Quinono Dioxime I05- M - 3 REPORT 68 PUBLISHED REPORT 18 PUDLISHEO REPORT 39 PUBLISHED REPORT 133 PUBLISHED REPORT 117 PUBLISHED REPORT 181 PUBLISHED REPORT 169 PUBLISHED REPORT 52 PUBLISHED REPORT 161 PUBLISHED REPORT 107 PUBLISHED t REPORT 6 PUBLISHED REPORT 121 PUBLISHED REPORT 110 PUBLISHED REPORT 179 PUDUSHED OSB-MDL GAV B6C3F1 LIVER, HEPATOCELLULAR TUMORS SPR-DAU IP/IJ B6C3F1 PERITONEAL SARCOMA F311 FEED LIVER, HEPATOCELLULAR TUMORS LIVER, ANGIOSARCOMA LYMPHOMA F311 B6C3F1 FEED LIVER, HEPATOCELLULAR TUMORS FEED LIVER, HEPATOCELLULAR TUMORS F311 B6C3FT FEED LIVER, HEPATOCELLULAR TUMORS F311 FEED B6C3F1 LIVER, HEPATOCELLULAR TUMORS F311 GAV D6C3FI LIVER, HEPATOCELLULAR TUMORS PAHCREAS ISLET-CELL F311 B6C3F1 FEED BLADDER, URINARY SO FIBROMA F311 B6C3F1 FEED HEMAHGIDSARCOMA LIVER, HEPATOCELLULAR' TUMORS F311 FEED KIDNEY URETER BLADDER, URINARY F311 FEED B6C3FT LIVER. HEPATOCELLULAR TUMORS F311 GAV B6C3F1 FORESTOMACH F311 B6C3F1 FEED BLADDER, URINARY RAT MOUSE MF MF NN PP SS PP PP PP P NH HH HN KH H P S PP S HH N P PP P 'P N P *H H HH PP PP PP P P NN PP N P N P H N. NN r C0385Q p-QuInonc Dioxfme 105- 11-3 REPORT V - F344 FEED N HN PUBLISH B6C3FI BLADDER .URI.HAR--------------------- E-------- 003554 004 520A COO 162 1 1>2 2 - T o t r a c h l o r o e t h a n e 79-34-5 Talrachloroethylcne 127-18-4 TetracMorvlnplias 961-U-5 CO 1581 CO 0259 B-TGDR 789-61-7 Toxapheno 800 1-35-2 C04579 1> 12-Trichloroethano 79-00-5 C04546A CO 0442 Trichloroethylene 79-01-6 Tr 1fluralIn 1582-09-8 C02186 Trimethylthi ourea 2439-77-2 F344 = Fischer 344 rat OSB-MDL = Osborne Mendel rat 5PR-DAW ~ Sprague Dawloy rat GAV = Gavaqe IP/IJ = Intraperitoneal Injection P s Positive evidence for carcinogenicity S = Suqgestive evidenca for carcinogenicity N = Hot shown to be carcinogenic I = Inadequate test REPORT 27 PUBLISHED REPORT U PUBLISHED REPORT 33 PUBLISHED REPORT 57 PUBLISHED REPORT 37 PUBLISHED REPORT 74 PUBLISHED REPORT 2 PUBLISHED REPORT 34 PUBLISHED REPORT 129 PUBLISHED OSB-MDL B6C3F1 OSB-MDL BGC3F1 OSB-MDL B6C3F1 GAV GAV FEED LIVER, HEPATOCELLULAR TUMORS LIVER, HEPATOCELLULAR TUMORS THYROID C-CELL ADRENAL CORTEX LIVER. HEPATOCELLULAR TUMORS SPR-DAU B6C3F1 OSB-MDL B6C3F1 OSB-MDL B6C3F1 OSB-MDL B6C3F1 OSB-MDL B6C3F1 F344 B6C3F1 IP/IJ FEED GAV GAV FEED FEED 2YMBAL S GLAND THYROID FOLLICULAR LIVER, HEPATOCELLULAR TUMORS LIVER. HEPATOCELLULAR TUMORS ADRENAL LIVER, HEPATOCELLULAR TUMORS LIVER, HEPATOCELLULAR TUMORS LUNG FORESTOMACH THYROID FOLLICULAR NN I1 H S s Sp PP PP PS II ss NN HN NN N P PP PP SP PP N P s s NH T A B L E 6. NCI Carcinogen Qloassays with Evidence Suggestive of a Carcinogenic Effect HCI HO. CHEMICAL NAME CA5 NUMBER REPORT STATUS ANIMAL ROUTE ORGAN SITES STRAIN 003247 C046 15 C0375S C02664 CO 006 6 Acctohcxami da . 768-81-0 Allyl Chloride 107-05-1 p-Anisidina Hydrochloride 20265-97-8 A r o c h l o r - 1254 27323-18-8 Asi nphosmothyl 86-50-0 C03521 IH-Benzotriazole 95-14-7 C0359B C0ZD39A Butylatod Hydroxytoluene (8HT) 128-37-0 p-Chloroani1 Ino 106-47-8 003316 CO 043 1 2-Chloro-p-Phcnylenadi araino Sulfato 61702-44-1 C l o n 1t r a 1 1d 1420-04-8 002028 C03667B Dlbutylfcln Dlacotata 1067-33-0 2,7-Dichlorodlbonzo-p-Dloxtn CDCDD) 33857-26-0 C04535 1( I-Dlchloroethano 75-34-3 C00 113A 004524 Dtchlorvos 62-73-7 N.N'-Di cyclohoxylthlourea REPORT 50 PUBLISHED REPORT 73 PUQLTSHED REPORT 116 PU0L1SKED REPORT 38 PUBLISHED REPORT 69 PUBLISHED REPORT 88 PUBLISHED REPORT 150 PUBLISHED REPORT 189 PUBLISHED REPORT 113 PUBLISHED REPORT 91 PUBLISHED REPORT 183 PUBLISHED REPORT 123 PUBLISHED REPORT 66 PUBLISHED REPORT 10 PUBLISHED REPORT 56 F344 D6C3M OSB-tIDL B6C3F1 F344 B6C3F1 F344 FEED GAV FEED FEED OSB-nDL FEED B6C3F1 F344 FEED B6C3F1 F344 B6C3F1 F344 B6C3F1 FEED FEED F344 B6C3F1 OSB-MDL B6C3F1 FEED FEED F344 B6C3F1 OSB-MDL B6C3F1 FEED FEED OSB-MDL GAV B6C3F1 OSB-MDL B6C3F1 F344 FEED FEED LEUKEMIA FORESTOMACH PREPUTIAL GLAND LIVER, HEPATOCELLULAR TUMORS GASTROINTESTINAL TRACT THYROID FOLLICULAR PANCREAS ISLET-CELL BRAIN LUNG LUNG SPLENIC SARCOMA HEMANGIOSARCOMA BLADDER, URINARY MAMMARY GLAND FORESTOMACH UTERUS NON-SARCOMA LYMPHOMA LIVER, HEPATOCELLULAR TUMORS HEMANGIOSARCOMA MAMMARY GLAND HEMANGIOSARCOMA UTERUS NON-SARCOMA ESOPHAGUS RAT MOUSE MF MF N S NN H S SS 5S H S S sS HN H S sS H S 5 K S NN N S 5 HN H NN SS HH NK N S N S SS II H rN HN N S Ss N S sH NN cool 13A Dichlorvos 62-73-f C04524 HiH*-Dicyclohunylthlcured 10 iKED REPORT 56 Styrene 100-42-5 72-54-8 F344 - Fischer 344 rat OSB-fIDL = Osborno Mendol rat SPR-DAW = Sprague Dowley rat GAV = GavaqB IP/IJ = Intraporitoneal injection P = Positive evidence for carcciinnoocgen)clty S = Sugqestive evidence for carcinogenicity N = Hot shown to be carcinogenic 2 = Inadequate test PUBLISHED REPORT 21 PUBLISHED REPORT 132 PUBLISHED REPORT 156 PUBLISHED REPORT 103 PUBLISHED REPORT 70 PUBLISHED REPORT 16 PUBLISHED REPORT 23 PUBLISHED REPORT 5 PUBLISHED REPORT 185 PUBLISHED REPORT 131 PUBLISHED OSB-MDL BSC3FI F3AA FEED FEED UTERUS'NQN-SARCOMA ESOPHAGUS HH H S H/ S HH ^r-i"v?^ C- "Tt*` ' B6C3Ft OSB-MDL B6C3F1 F3AA B6C3F1 F3AA B6C3F1 F3AA B6C3FI OSB-MDL B6C3FI OSB-MDL B6C3F1 FEED FEED FEED FEED FEED FEED OSB-MDL B6C3F1 F3AA B6C3FI FEED FEED F3AA B6C3F1 OSB-MDL B6C3F1 GAV FEED THYROID FOLLICULAR LIVER. HEPATOCELLULAR TUMORS LIVER. HEPATOCELLULAR TUMORS LIVER, HEPATOCELLULAR TUMORS SKIN/SUBCUTANEOUS TISSUE ADRENAL CORTEX hemangxosarcoma THYROID C-CELL LIVER, HEPATOCELLULAR TUMORS GASTROINTESTINAL TRACT LIVER, HEPATOCELLULAR TUMORS LUNG THYROID FOLLICULAR 5 NH NH NN NN S5 S S ss I s N, N N s N s N S N S N S NH NH NN I S H s NH 't .t, . T A B L E 7. NCI Carcinogen Bloassays with No Significant Evidence for Carcinogenicity MCI NO. CHEMICAL NAME CA5 NUMBER C0669Q C08689 CO 1730 C02697 C02937 C03805 C03770 cosm CO 3907 C00533 C0209Q CO 1752 CQ5662 C00969 C03Z69 C08673 C03656A Aldi carb 116-06-3 Anilazine 101-05-3 o-Anthrani1ic Acid 1 18-92-3 Asprin. Phonacatin And Caffeine 8003-03-0 Calcium Cyanamido 156-62-7 Carbromal 77-65-6 9 '- (ChloroacetylJ-Acatanl1 1da 190-99-8 (2-Chloroafchyl) Trimathylammoniuro Chi or do 999-8 1-5 2-(Chloromofchyl)Pyridino Hydrochlorido 6959-97-3 Chloron icr in 76-06-2 3-Chlor-p-Toluidine 95-79-9 Clilorpropnm ido 99-20-2 Coumaphos 56-72-9 DDT 50-29-3' DiaryInni lido Yo IIo h 6358-85-6 Diasi non 333-9 1-5 Dibonso-p-Dioxin 262-12-9 REPORT STATUS ANIMAL ROUTE ORGAN SITES STRAIN REPORT 136 PUBLISHED REPORT 109 PUBLISHED REPORT 36 PUBLISHED REPORT 67 PUBLISHED REPORT 163 PUDLISHED REPORT 173 PUDLISHED REPORT 177 PUDLISHED REPORT 158 PUDLISHED REPORT 178 PUDLISHED REPORT 65 PUDLISHED REPORT 195 PUDLISHED REPORT 95 PUBLISHED REPORT 96 PUDLISHED REPORT 131 PUDLISHED REPORT 30 PUDLISHED REPORT 137 PUDLISHED REPORT 122 PUDLISHED F399 B6C3F1 F399 B6C3F1 F39 9 D6C3F1 F359 D6C3F1 F399 D6C3F1 F399 B6C3F1 F399 B6C3F1 F399 D6C3F1 F399 D6C3F1 OSD-MDL B6C3F1 F399 D6C3F1 F399 06C3F1 F399 B6C3F1 OSD-MDL D6C3F1 F399 D6C3F t F399 B6C3F1 OSD-MDL D6C3F 1 FEED FEED TEED FEED FEED FEED FEED FEED CAV GAV FEED FEED FEED FEED FEED FEED FEED .i -'i` RAT MOUSE MF MF NN HN HH HN HN NN NH NN HN NN NH HN HH NH NN HH HN NN II NN NN HH KN N H HN NH NN NN HN NN NN NN NN NN CO 1673 C03636A Di 3231n3o-4n 1-3 D ibcnso-p-DIaxin 262-12-4 RPUEPBOLRISTH `V REPORT ' PUBLISHED ivf1* 4" PPflliPWP COO 124 D1 eldrl n 60-57-1 REPORT 22 PUBLISHED COO 135 Dimethoate 60-51-5 REPORT 4 PUBLISHED C02255 2.4-Dimethoxyaniline Hydrochloride 54150-69-5 REPORT 171 PUBLISHED C50055 Dimethyl Terephthalato 120-6 1-6 REPORT 121 PUBLISHED CO 0395 Di oxathion 78-34-2 REPORT 125 PUBLISHED CD3974 EDTA Trisodium Salt CTrihydrate) 150-38-9 REPORT 11 PUBLISHED CO 0566 Endosu1fan 1 15-29-7 REPORT 62 PUBLISHED COO 157 Endr in 72-20-8 REPORT 12 PUBLISHED CO 16 94 Ethi onamfda 536-33-4 REPORT 46 PUBLISHED N*L4n> C03010 Formulated Fenaminosulf 140-56-7 C02653 Hexachlorophene 70-30-4 REPORT 101 PUBLISHED REPORT 40 PUBLISHED C04568 Iodoform 75-47-6 REPORT 110 PUDLI51IED CO289 1 Lead D 1mcthyldi thlocarbmata 19010-66-3 REPORT 151 P U D L 1SHED CO 0204 L indana 55-89-9 REPORT 14 PUBLISHED C0386 1 Llthocholic Acid 434-13-9 REPORT 175' PUBLISHED C00215A Malathi on 121-75-5 REPORT 24. PUBLISHED 1 C50000 dl-Menthol 89-78-1 REPORT 98' PUBLISHED . CO 04 97 Mathoxychlor 72-43-5 REPORT 35 PUBLISHED C0297 1 tlathyl Parathlon 298-00-0 REPORT 157 PUBLISHED C00544 Mexacarbate 3 15-18-4 REPORT 147 PUBLISHED F3**i BC3FI OSB-MDL 86C3F1 FEED FEED ' iin F3<iA FEED OSB-MDL B6C3F1 F3A4 B6C3F1 F344 DC3F1 OSB-MDL B6C3F1 F3A4 B6C3F1 OSB-MDL D6C3F1 OSB-MDL B6C3F1 OSB-MDL B6C3F1 FS't* B6C3F1 F3AA FEED FEED FEED FEED FEED FEED FEED FEED FEED FEED OSB-MDL D6C3F1 F3A4 B6C3F1 OSB-MDL B6C3FI F34A B6C3F1 OSB-MDL B6C3FI F3AA B6C3F1 OSD-MDL B6C3F1 F3A B6C3F1 OSB-MDL B6C3FI GAV FEED FEED GAV FEED FEED FEED FEED FEED HN HH NN HH \ 51 '*'r' NN NN NN HN NH NN NH NK HN HM H N IH IN NK HH HH HN HN HN NN NH NH HH NH NH HH HH HH H H HN HH HN HM HH NH NH HH HN 276 T A B L E 7. N CL Carcinogen Bloassays with No Significant Evidence for Carcinogenicity {continued) HCI NO. c h e m i c a l NAME CAS HUMBER C03281 CO 1965 CO 1956 C03961 CO22 T1 co m s s N-( 1-NopMhyl )Ethylenodiamino.2HCl 1A5-2S-A 6 -Ni troanthranilie Acid 619-17-0 1-Ni tronnphthalene 86-57-7 6 - H 1fcro-o-Phcnylenodiamino 99-56-9 Bota-Mitrostyreno and Styrene 102-96-5 100-62-5 NTA Trisodium 5all.H20 18662-53-8 REPORT 5TATUS (. REPORT 168 PUBLISHED REPORT 109 PUBLISHED REPURT 66 PUBLISHED REPORT 180 PUBLISHED REPORT 170 PUBLISHED REPORT 6 PUBLISHED ANIMAL STRAIN F 366 B6C3F1 F366 86C3F1 F366 D6C3F1 F366 0CC3F1 F366 B6C3FI F366 B6C3F1 ROUTE ORGAN SITES FEED FEED FEED FEED GAV FEED C00619A CO 176 I CD3930 C03952 CDZZ33 COZ017 C0059? C036 12 C03601 CO28 13 Pcntacbloroni trobenzena 82-68-8 Phonformi n 1 16-86-3 p-Phonylonediamtno Dihydrochloride 626- 18-0 1-Phcnyl-3~MethyI"5-Pyrazolono 89-25-8 N-Phenyl-p-Phenylenodi ami ne 101-56-2 t-Phonyl-2-Thi ouroa 103-85-5 Photodi cldrIn 13366-73-9 Phthalamido * 88-96-0 Phthalic AnhydridB .-- 85-66-9 Piperonyl Butoxido 51-03-6 REPORT 61 PUBLISHED REPORT 7 PUBLISHED REPORT 176 PUBLISHED REPORT 161 PUBLISHED REPORT 82 PUBLISHED REPORT 168 PUBLISHED REPORT 17 PUBLISHED * REPORT 161 PUBLISHED REPORT 159 PUBLISHED REPORT 120 PUBLISHED OSB-MDL D6C3F1 F366 B6C3F1 F366 B6C3F1 5366 B6C3F1 F366 B6C3F1 F36 6 B6C3F1 OSB-MDL B6C3F1 F366 B6C3F1 F366 B6C3F1 F366 B6C3F1 FEED FEED FEED FEED FEED FEED FEED FEED FEED FEED RAT MOUSE nF MF HH NN HH NH HH HH HN HH HH NN HH NN HH HN NN HH NH HN HH NK NH NN NH NN NN NH ' HH HN N N HN H H HN CQ3601 C028 13 Phthalic Anhydride 65-44-9 Plperonyl Butoxldo 31-03-6 CO 1633 C02835 C5DD22 Pyrazinamida 93-96-4 Pyrimethamine 53- 14-0 Sodium Oi cthyldithiocarbamata 148-18-5 Suif1soxazolo 127-69-5 C04557 C03032 C02959 C04240 C03327 C01763 CO 1832 C04637 C00260 CO 1729 3-Sulfolene 77-79-2 2*35f6"Totrachloro-4-m troanisola 2438-68-2 Tetraethylthiuram Disulflda 97-77-8 Titanium Dioxide 13463-67-7 1309-63-3 Tolazamjdo 1156-19-0 Tolbutamida 64-77-7 2 , 5-Toluenediamina Sulfata 6369-59-1 Trichlorofluoromathane 75- 69-4 Triphenyltln Hydroxide 76- 87-9 1-Tryptophan 73-22-3 F344 s Fischer 344 rat OSB-MDL = Osborne Mendel rat SPR-DAU = Sprague Dawley rat GAV = Gavaae IP/IJ = Intraperitoneal injection H = Not shown to be carcinogenic I - Inadequate test KLru*i i'jt PUBLISHED REPl' PUBI' 0 ) rjii B6C3FI F344 B6C3F1 I ULLI FEED HN HH < REPORT 46 PUBLISHED REPORT 77 PUBLISHED REPORT 17Z PUBLISHED REPORT 138 PUBLISHED F344 B6C3F1 F344 B6C3F1 F344 B6C3F1 F344 B6C3F1 FEED FEED FEED GAV REPORT 102 PUBLISHED REPORT 114 PUBLISHED REPORT 166 PUBLISHED REPORT 97 PUBLISHED REPORT 51 PUBLISHED REPORT 31 PUBLISHED REPORT 126 PUBLISHED REPORT 106 PUBLISHED REPORT 139 PUBLISHED REPORT 71 PUBLISHED OSB-MDL B6C3F1 F344 B6C3F1 F344 B6C3F1 F344 B6C3F1 F344 B6C3F1 F344 B6C3F1 F344 6C3F1 OSB-MDL D6C3F1 F344 D6C3F1 F344 B6C3F1 GAV FEED FEED FEED FEED FEED FEED GAV FEED FEED HH NI HH IH HN N H HN H H NN HH HH HN HH NH HH NH NH NH NN HH NH HH II NH HH HH , NH HN \ T A B L E 8. NCI Carcinogen Bioassays That Are Inadequate NCI NO. CHEMICAL NAME CAS NUMBER REPORT STATUS ANIMAL ROUTE STRAIN 278 COZ 108 C02095 C0 151$ C02120 COZ 13 1 CQ$5 9 1A C0 1821 CO 1605 C50259 C02 1$2 C03 190 00208$ COZ 153 C0$626A C 0 2 1 19 Acotamido 60-35-5 NO Tf;CH RPT INCONCLUSIVE Adi pami de 628-9$-$ NO TECH RPT: INCONCLUSIVE A dr1amyc1n 23Z1$-92-8 NO TECH RPT; INCONCLUSIVE l-Arqlnine Glutamate $320-30-3 NO TECH RPT; INCONCLUSIVE N-Butyluroa 592-31-$ NO TECH RPT; INCONCLUSIVE Carbon Disulfide 75-15-0 NO TECH RPT; INCONCLUSIVE N.N-Dimolhyl-p-Ni trosoani1 ino 138-89-6 NO TECH RPT; INCONCLUSIVE Emeti ne $83-18-1 REPORT $3 PUBLISHED Nexame thylroclamino 6$5-05-6 NO TECH RPT; INCONCLUSIVE Hexanarnide 628-02-$ MO TECH RPT; INCONCLUSIVE Mcthansulfon-m-AnIs ide. $*-(9-Acridinvl- NO TECH RPT; ami no) - 1 monohydrochlor ido UK1L057 9 INCONCLUSIVE Nitrous Acid. Sodium Salt 763Z-00-0 NO TECH RPT; INCONCLUSIVE p-Tolylurca 622-51-5 NO TECH RPT; INCONCLUSIVE 1. 1. 1-1r ichioroothone 71-55-6 REPORT 3 PUBLISHED Urea 57-13-6 ' NO TECH RPT; INCONCLUSIVE F3$$ B6C3F1 F3$ $ B6C3F1 F3$$ B6C3F1 F 3$ $ D6C3F1 F3$$ D6C3F1 OSD-MDL B6C3F1 F3$$ D6C3F1 SPR-DAW B6C3F1 F3$$ D6C3F1 F3$$ D6C3F1 F3$$ D6C3F1 B6C3F1 F3$$ D6C3F1 0SB-NDI B6C3FI F3$$ 06C3F1 FEED FEED IP/IJ FEED FEED GAV FEED IP/U IP/IJ FEED IP/IJ FEED FEED GAV FEED F3$$ = Fischer 34$ rat OSD-MDL ~ Osborne Mondai rat SPR-DAW s Sprague Pauley rat GAV = Gavage IP/IJ = Intraperitoneal Injoction nl u* --co MO >0 nL. c tcrt2>i co t3r 4 o COa --i* L. C P L un, oOotOoia. nn p *Q>-v. >&p +c> un*on < i p i- II i sT I I 1tuo.mo em. <M0. EVALUATION OF NATIONAL CANCER INSTITUTE CARCINOGEN BIOASSAYS 279 incidences higher than 10% when there are no tumors in the control group; as a consequence, negative results do not necessarily mean that the test chemical is not a carcinogen. (4) Any substance that is deemed to be a carcinogen or a suspected carcinogen is a potential hazard to humans unless proved otherwise. The extent o f the health risk to humans may be a matter for further research. These premises are scientific issues. Increasing knowledge about chemical carcinogenesis may validate, alter, or eliminate any o f them. As a consequence, the design, conduct, and evaluation of the carcinogen bioassay will undergo an evolutionary process. We hope that the procedures described in this paper will serve as discussion points for further developments of rodent bioassays. E*rC', R E F E R E N C E S Boyland, E. 1969. The correlation of experimental carcinogenesis and cancer in man. Prog. E xp . Tumor Res. 2:222-234. Cox, D. R. 1970. Analysis o f Binary Data, pp. 48-52. London: Methuen. Doll, R. 1971. The age distribution of cancer: Implications for models of carcinogenesis. /. R . Stat. Soc. A 34:133-166. Doll, R. 1977. Strategy for detection of cancer hazards to man. Nature (L a n d .) 265:589-596. Doll, R. 1978. An epidemiological perspective of the biology of cancer. Cancer Res. 38:3573-3583. Fears, T. R., Tarone, R. E., and Chu, K. C. 1977. False-positive and false-negative rates for carcinogenicity screens. Cancer Res. 37:1941-1945. Gart, J. ]., Chu, K. C., and Tarone, R, E. 1979. Statistical issues in the interpretation o f chronic bioassay tests for carcinogenicity. /. Natl. Cancer inst. 62:957-974. Goodman, D. G., Ward, ]. M., Squire, R. A., Chu, K. C., and Linhart, M. 5. 1979. Neoplastic and nonneoplastic lesions in aging F344 rats. Toxicol, Appf. Pharmacol. 48:237-248. Higginson, J. 1972. The role of geographical pathology of environmental carcinogenesis. In Proceedings o f the 1971 Symposium on Fundamental Cancer Research on the Environm ent and Cancer, ed. R. L. Clark, pp. 69-92. Baltimore: Williams & Wilkins. Miller, R. G. 1966. Simultaneous Statistical Inference, pp. 6-10. New York: McGraw-Hill. NCI-TR-3. 1977. Bioassay of 1,1,1-trichloroethane for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 3, D HEW Pubi. 78-828. NCI-TR-14, 1978. Bioassay of lindane for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 14, DHEW Publ. 78-828. NCI-TR-15. 1978. Bioassay of captan for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 15, DHEW Publ. 77-81S. NCI-TR-17. 1977. Bioassay o f photodieldrin for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 17, D HEW Publ. 78-828, NCI-TR-24. 1978. Bioassay o f malathion for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 24, DHEW Publ. 78-828. NCI-TR-25. 1978. Bioassay o f chloramben for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 25, DHEW Publ. 78-828. NCI-TR-28. 1978. Bioassay of dibromochloropropane for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 28, D HEW Publ. 78-828. NCI-TR-30. 1978. Bioassay of diarylanilide yellow for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 30, DHEW Publ. 78-828. NCI-TR-43. 1978. Bioassay of emetine for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 43, DHEW Publ. 78-828. NCI-TR-45. 1978. Bioassay of chlorpropamide for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 45, DHEW Pubi. 78-845. 280 K.C . CHU E T A L NCI-TR-54. 1978. Bioassay of 2,4-dinitrotoluene for possible carcinogenicity. Carcinogenesis Tech. . Rep. Ser. 54, D H EW Pub/. 78-828. NC1-TR-90. 1978, Bioassay o f dicofol for possible carcinogenicity. Carcinogenesis Tech. Rep. Ser. 90, DHEW Pubt. 78-1340. Page, N. P. 1977. Concepts o f a bioassay program in environmental carcinogenesis. In Advances in Modern Toxicology: Environmental Cancer, eds. H. Kraybill and M. Mehlman, vol. 3, pp. 87-171. New York: Wiley. Sontag, J. A., Page, N. P., and Saffiottl, U. 1976. Guidelines for carcinogen bioassay in small rodents. Carcinogenesis Tech. Rep. Ser. 1, D HEW PubI (N!H)76-801. Squire, R. A. and Levitt, M. 1975. Report of a workshop on classification of specific hepatocellular lesions in rats. Cancer Res. 35:3214-3215. Tarone, R. E., Chu, K. C., and Ward, J. M. 1981. Variability in the rates of some common naturally occurring tumors in F344 rats and B6C3F1 mice. } . Natl. Cancer Inst. 66:1175-1181. Thomas, D. G., Breslow, N., and Gart, J. J. 1977. Trend and homogeneity analyses of proportions and life table data. Comput. Biomed. Res. 10:373-381. Ward, J. M., Goodman, D. G., Griesemer, R. A., Hardisty, J. F., Schueler, J. D., Squire, R. A., and Strandberg, J. 0. 1978. Quality assurance for pathology in rodent carcinogenesis tests. J. Environ. Pathol. Toxicol. 2:371-378. Ward, |. M., Goodman, D. G., Squire, R. A., Chu, K. C., and Linhart, M. S. 1979. Neoplastic and nonneoplastic lesions in aging (C57BL/6N X C3H/HeN)F1 (B6C3F1) mice. /. Natl. Cancer Inst. 63:849-854. Received January 29, 1980 Accepted December 15, 1980 0- H