Human carcinogenic risk evaluation, part II: contributions of the EUROTOX specialty section for carcinogenesis.

There is growing recognition that carcinogenic risk extrapolation to low doses (and standard setting) should consider the mode of action of a given chemical. So far, there is agreement on distinguishing between genotoxic and nongenotoxic chemicals; yet, further differentiations seem appropriate. For genotoxic carcinogens, case studies of chemicals point to many possibilities for assessing carcinogenic risk. There are numerous, apparently genotoxic carcinogens where practical thresholds are a matter of discussion. For instance, positive data of chromosomal effects only, in the absence of mutagenicity, may support the characterization of a compound that produces carcinogenic effects only at high, toxic doses. There is a wide consensus that for non-DNA-reactive genotoxicants, such as aneugens, thresholds should be defined. Specific mechanisms of clastogenicity have been repeatedly addressed as also having thresholds, such as topoisomerase II poisons, or mechanisms based on reactive oxygen. These and other arguments together lead to the distinction of four groups of carcinogens, which have been introduced (C. Streffer et al., 2004, Springer-Verlag). There are nonthreshold genotoxic carcinogens (for low-dose risk assessment, the linear nonthreshold [LNT] model appears appropriate); genotoxic carcinogens, for which the existence of a threshold cannot be sufficiently supported (in these cases the LNT model is used as a default assumption, based on the scientific uncertainty and backed by the precautionary principle); genotoxic carcinogens for which a practical threshold is supported; and nongenotoxic carcinogens and non-DNA-reactive carcinogens (for these compounds a true [perfect] threshold is associated with a clearly founded no-observed-adverse-effect level).     

Risk-Assessment Implications of Mechanistic Model��S Prediction of Low-Dose Nonlinearity of Liver Tumor Risk for Mice Fed Fumonisin B1

A two-stage, clonal-expansion model of liver tumor risk in mice was developed by Kodell et al. (Food Addit Contam 18:237–253, 2001) based on the hypothesis that fumonisin B₁, a naturally occurring mycotoxin in corn, is not genotoxic, but rather causes cancer through the disruption of sphingolipid metabolism. This disruption is assumed to cause an increase in apoptosis, in response to which cells proliferate to compensate for reduced tissue mass. The resulting differential increase in the number of pre-neoplastic cells at risk of mutation during cell division is assumed to lead to an increase in the incidence of tumors. Two-year liver tumor incidences predicted by the model using data on organ weight, cell proliferation, and sphingolipid metabolism provided a reasonable match to the actual 2-year observed incidences in a study conducted at the National Center for Toxicological Research. The predictions indicated no risk at low doses (even a possible hormetic effect) and high risk at high doses in females, as well as a complete absence of a dose response (or perhaps, a hormetic effect) in males. This paper provides a commentary on the risk-assessment implications of the modeling results, pointing out that the model’s low-dose predictions provide scientific support and justification for the U.S. Food and Drug Administration’s low-ppm guidance levels in corn products. These guidance levels are significantly higher than would be obtained using linear extrapolation, the method most often used for genotoxic carcinogens and other carcinogens for which low-dose linearity cannot be ruled out.     

Possible Contribution of Indomethacin to the Carcinogenicity of Nongenotoxic Bladder Carcinogens That Cause Bladder Calculi

Nongenotoxic bladder carcinogens that form bladder calculi have been concluded to be of low carcinogenic risk to humans because bladder stones would be expelled or surgically removed before they had a chance to exert their carcinogenic effect. It is the aim of this report to examine the possible contribution of indomethacin to the carcinogenic risk posed by nongenotoxic bladder carcinogens that cause bladder stones. Indomethacin may act as a tumor promoter in the bladder by interfering with the synthesis of prostaglandins. Prostaglandins have a cytoprotective function in the gastric mucosa and possibly also in the urinary bladder. Diminished cytoprotection may be implicated in bladder carcinogenesis as β-naphthylamine, a human bladder carcinogen, also inhibits prostaglandin synthesis in vitro. The presence of other tumor promoters in the bladder may further ensure that tumors would be formed even if bladder stones were expelled. People who are exposed to nongenotoxic bladder carcinogens that are present in the environment and that form bladder stones, therefore, may be at an increased risk for developing bladder cancer if they are also exposed to tumor promoters, such as indomethacin.     

Biomarkers of carcinogenicity and their roles in drug discovery and development

The screening of drug candidates to assess their carcinogenic potential has long been a challenge for drug development. While genotoxic compounds can be readily detected with a battery of standard tests, including short-term in vitro and in vivo assays, predicting nongenotoxic carcinogenicity remains a major challenge. The 2-year rodent bioassay has been held as the gold standard for the assessment of carcinogenic risk to humans. However, due primarily to the continuing doubt over their relevance to human risk assessment, there has been an increased demand for more efficient and accurate approaches to predict and understand human relevant risk of carcinogenicity. Novel biomarkers have helped to shed light on our understanding of the factors that lead to and are characteristic of the carcinogenic phenotypes. Tissue biomarkers of carcinogenicity identified to be concordant with drug exposures resulting in tumor outcome may assist the drug development process by resolving ambiguities, shortening timelines and enabling earlier decisions on compounds. This information could vastly improve the efficiency with which nongenotoxic carcinogens are identified and ensure earlier insight into the relevance for humans.     

The Comet Assay with Multiple Mouse Organs: Comparison of Comet Assay Results and Carcinogenicity with 208 Chemicals Selected from the IARC Monographs and U.S. NTP Carcinogenicity Database

ABSTRACT

The comet assay is a microgel electrophoresis technique for detecting DNA damage at the level of the single cell. When this technique is applied to detect genotoxicity in experimental animals, the most important advantage is that DNA lesions can be measured in any organ, regardless of the extent of mitotic activity. The purpose of this article is to summarize the in vivo genotoxicity in eight organs of the mouse of 208 chemicals selected from International Agency for Research on Cancer (IARC) Groups 1, 2A, 2B, 3, and 4, and from the U.S. National Toxicology Program (NTP) Carcinogenicity Database, and to discuss the utility of the comet assay in genetic toxicology.

Alkylating agents, amides, aromatic amines, azo compounds, cyclic nitro compounds, hydrazines, halides having reactive halogens, and polycyclic aromatic hydrocarbons were chemicals showing high positive effects in this assay. The responses detected reflected the ability of this assay to detect the fragmentation of DNA molecules produced by DNA single strand breaks induced chemically and those derived from alkali-labile sites developed from alkylated bases and bulky base adducts. The mouse or rat organs exhibiting increased levels of DNA damage were not necessarily the target organs for carcinogenicity. It was rare, in contrast, for the target organs not to show DNA damage. Therefore, organspecific genotoxicity was necessary but not sufficient for the prediction of organ-specific carcinogenicity. It would be expected that DNA crosslinkers would be difficult to detect by this assay, because of the resulting inhibition of DNA unwinding. The proportion of 10 DNA crosslinkers that was positive, however, was high in the gastrointestinal mucosa, stomach, and colon, but less than 50% in the liver and lung. It was interesting that the genotoxicity of DNA crosslinkers could be detected in the gastrointestinal organs even though the agents were administered intraperitoneally.

Chemical carcinogens can be classified as genotoxic (Ames test-positive) and putative nongenotoxic (Ames test-negative) carcinogens. The Ames test is generally used as a first screening method to assess chemical genotoxicity and has provided extensive information on DNA reactivity. Out of 208 chemicals studied, 117 are Ames test-positive rodent carcinogens, 43 are Ames test-negative rodent carcinogens, and 30 are rodent noncarcinogens (which include both Ames test-positive and negative noncarcinogens). High positive response ratio (110/117) for rodent genotoxic carcinogens and a high negative response ratio (6/30) for rodent noncarcinogens were shown in the comet assay. For Ames test-negative rodent carcinogens, less than 50% were positive in the comet assay, suggesting that the assay, which detects DNA lesions, is not suitable for identifying nongenotoxic carcinogens. In the safety evaluation of chemicals, it is important to demonstrate that Ames test-positive agents are not genotoxic in vivo. This assay had a high positive response ratio for rodent genotoxic carcinogens and a high negative response ratio for rodent genotoxic noncarcinogens, suggesting that the comet assay can be used to evaluate the in vivo genotoxicity of in vitro genotoxic chemicals. For chemicals whose in vivo genotoxicity has been tested in multiple organs by the comet assay, published data are summarized with unpublished data and compared with relevant genotoxicity and carcinogenicity data.

Because it is clear that no single test is capable of detecting all relevant genotoxic agents, the usual approach should be to carry out a battery of in vitro and in vivo tests for genotoxicity. The conventional micronucleus test in the hematopoietic system is a simple method to assess in vivo clastogenicity of chemicals. Its performance is related to whether a chemical reaches the hematopoietic system. Among 208 chemicals studied (including 165 rodent carcinogens), 54 rodents carcinogens do not induce micronuclei in mouse hematopoietic system despite the positive finding with one or two in vitro tests. Forty-nine of 54 rodent carcinogens that do not induce micronuclei were positive in the comet assay, suggesting that the comet assay can be used as a further in vivo test apart from the cytogenetic assays in hematopoietic cells. In this review, we provide one recommendation for the in vivo comet assay protocol based on our own data.     

Epigenetic mechanisms of chemical carcinogenesis

Chemically induced cancer is a multi-step process involving damage to the genome initially followed by clonal expansion of the DNA damaged cell eventually leading to a neoplasm. Chemical carcinogens have been shown to impact at all of the stages of the tumorigenesis process. It has become apparent that chemical and physical agents that induce cancer may do so through several different cellular and molecular mechanisms. Epigenetic (nongenotoxic) chemical carcinogens are those agents that function to induce tumor formation by mechanisms exclusive of direct modification or damage to DNA. These agents appear to modulate cell growth and cell death and exhibit dose response relationships between exposure and tumor formation. The exact and/or exclusive mechanisms by which these agents function have not been established, however, changes in cell growth regulation and gene expression are important to tumor formation. This review focuses on several potential mechanisms and cellular processes that may be involved in nongenotoxic chemical carcinogenesis.     

Possible contribution of indomethacin to the carcinogenicity of nongenotoxic bladder carcinogens that cause bladder calculi

Nongenotoxic bladder carcinogens that form bladder calculi have been concluded to be of low carcinogenic risk to humans because bladder stones would be expelled or surgically removed before they had a chance to exert their carcinogenic effect. It is the aim of this report to examine the possible contribution of indomethacin to the carcinogenic risk posed by nongenotoxic bladder carcinogens that cause bladder stones. Indomethacin may act as a tumor promoter in the bladder by interfering with the synthesis of prostaglandins. Prostaglandins have a cytoprotective function in the gastric mucosa and possibly also in the urinary bladder. Diminished cytoprotection may be implicated in bladder carcinogenesis as beta-naphthylamine, a human bladder carcinogen, also inhibits prostaglandin synthesis in vitro. The presence of other tumor promoters in the bladder may further ensure that tumors would be formed even if bladder stones were expelled. People who are exposed to nongenotoxic bladder carcinogens that are present in the environment and that form bladder stones, therefore, may be at an increased risk for developing bladder cancer if they are also exposed to tumor promoters, such as indomethacin.     

Toxicogenomics and cancer risk assessment: A framework for key event analysis and dose–response assessment for nongenotoxic carcinogens

In order to determine a threshold for nongenotoxic carcinogens, the traditional risk assessment approach has been to identify a mode of action (MOA) with a nonlinear dose–response. The dose–response for one or more key event(s) linked to the MOA for carcinogenicity allows a point of departure (POD) to be selected from the most sensitive effect dose or no-effect dose. However, this can be challenging because multiple MOAs and key events may exist for carcinogenicity and oftentimes extensive research is required to elucidate the MOA. In the present study, a microarray analysis was conducted to determine if a POD could be identified following short-term oral rat exposure with two nongenotoxic rodent carcinogens, fenofibrate and methapyrilene, using a benchmark dose analysis of genes aggregated in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and Gene Ontology (GO) biological processes, which likely encompass key event(s) for carcinogenicity. The gene expression response for fenofibrate given to rats for 2 days was consistent with its MOA and known key events linked to PPARα activation. The temporal response from daily dosing with methapyrilene demonstrated biological complexity with waves of pathways/biological processes occurring over 1, 3, and 7 days; nonetheless, the benchmark dose values were consistent over time. When comparing the dose–response of toxicogenomic data to tumorigenesis or precursor events, the toxicogenomics POD was slightly below any effect level. Our results suggest that toxicogenomic analysis using short-term studies can be used to identify a threshold for nongenotoxic carcinogens based on evaluation of potential key event(s) which then can be used within a risk assessment framework.     

Risk assessment of dietary exposures to compounds that are genotoxic and carcinogenic--an overview

The process of risk assessment of dietary exposures to genotoxic carcinogens is summarised. Exposures to six genotoxic carcinogens in food (acrylamide, aflatoxin B(1), benzo(a)pyrene, dimethylnitrosamine, ethyl carbamate, PhIP) have been used to illustrate the process. The margin of exposure (MOE) approach is seen as a useful method to be used in the risk characterisation step of assessing exposures to genotoxic carcinogens. This approach combines information on animal potency and human exposure, and can be used to indicate levels of concern and also the ranking between various exposures to such agents. Both the T25 and the BMDL10 methods may be used as a reference point. Should a specific MOE value be developed as a cut-off between levels of concern and levels of low concern, the value using T25 data is proposed to be 2.5-times higher than using BMDL10 data. Linear low-dose extrapolation using either T25 or BMDL10, may also be applied. However, it should be understood that this approach should not be interpreted as giving a precise estimate of human risk. For exposures to mutagens in food lacking carcinogenicity data, it is proposed to apply the MOE approach to the lowest effective dose (LED) for in vivo genotoxicity.     

In silico approaches to predicting cancer potency for risk assessment of genotoxic impurities in drug substances

The current risk assessment approach for addressing the safety of very small concentrations of genotoxic impurities (GTIs) in drug substances is the threshold of toxicological concern (TTC). The TTC is based on several conservative assumptions because of the uncertainty associated with deriving an excess cancer risk when no carcinogenicity data are available for the impurity. It is a default approach derived from a distribution of carcinogens and does not take into account the properties of a specific chemical. The purpose of the study was to use in silico tools to predict the cancer potency (TD₅₀) of a compound based on its structure. Structure activity relationship (SAR) models (classification/regression) were developed from the carcinogenicity potency database using MultiCASE and VISDOM. The MultiCASE classification models allowed the prediction of carcinogenic potency class, while the VISDOM regression models predicted a numerical TD₅₀. A step-wise approach is proposed to calculate predicted numerical TD₅₀ values for compounds categorized as not potent. This approach for non-potent compounds can be used to establish safe levels greater than the TTC for GTIs in a drug substance.     

Induction effect of coadministration of soybean isoflavones and sodium nitrite on DNA damage in mouse stomach

We have already found that nitrite-treated isoflavones exhibit genotoxic activities toward Salmonella typhimurium TA 100 and 98 strains (submitted: nitrite-treated genistein). However, we have not demonstrated genotoxic activity induced by simultaneous treatment with isoflavones and NaNO₂in vivo. In the present study, we examined whether coadministration of isoflavones (such as daidzein and genistein) and NaNO₂ induces DNA damage in the stomach of ICR male mice. Mice were coadministered with isoflavones (1 mg/kg body weight) and NaNO₂ (10 mg/kg body weight), and dissected to collect tissues at 1, 3, and 6 h after administration. We used comet assay combined with repair enzyme formamidopyrimidine-N-glycosylase (FPG) to detect FPG-sensitive sites. An HPLC–ECD system was employed to determine 8-oxo-2′-deoxyguanosine (8-oxodG) in the stomach. In addition, we observed leukocyte infiltration by histopathological investigation, and measured total superoxide dismutase (SOD) in the stomach. We confirmed that oxidative DNA damage in the stomach was significantly increased by coadministration. Total SOD activities were also significantly stimulated by coadministration. However, the induction of inflammation in the stomach was not found. These data suggest that coadministration of isoflavones and NaNO₂ can cause DNA damage in the stomach because of the formation of radicals.     

Promotion of hepatocarcinogenesis in humans and animal models

Risk assessment based on rodent carcinogenicity data depends on the assumption of similarity between rodents and humans. While this assumption is conceivable in the case of genotoxic initiating carcinogens, considerable species differences have been observed with nongenotoxic tumor promoters. This heterogeneous group of agents increases the probability of cancer by stimulating selection and clonal expansion of cells transformed during tumor initiation. Since tumor promoters differentially affect normal tissue and preneoplastic cell clones, their action cannot be discussed without knowledge of persistent genomic and epigenetic alterations occurring during initiation and formation of preneoplastic cells. Chemical carcinogenesis, and in particular, tumor promotion, is known to be tissue specific. We focus on hepatocarcinogenesis in humans and in animal models and emphasize two different modes of action: (1) chronic cytotoxicity leading to promotion of liver carcinogenesis in both humans and animal models; (2) sustained activation of orphan receptors such as CAR, PPARalpha and Ah receptor leading to promotion of rodent but probably not human hepatocarcinogenesis. Further studies on the different modes of action may help to avoid overestimation of the risk of liver tumor promotion.     

Re-examination of the ED01 Study: Adjusting for Time on Study

Gaylor (1980) describes conclusions drawn from the initial analysisof the ED₀₁ study. These conclusions failed to take adequateaccount of differences in time on study for animals at differentdoses. Several consequences of this are discussed. Two conclusionsof Gaylor's are identified that are open to question. The ED₀₁liver and bladder tumor data, adjusted for time-on-study, areused to firmly reject several mathematical models that havebeen proposed for low-dose extrapolation. The complexities thatarise from differences in time-on-study are shown to cast seriousdoubt on the general concept of low-dose extrapolation, whenit is attempted independent of time to tumor. It is pointedout that the problems of statistical analysis that arise whendosed animals live longer than controls (as occurred in theED₀₁ study) raise questions about the design of lifetime feedingstudies in general.     

Pulmonary Mechanical Responses to Intravenously Administered Capsaicin in Guinea-pigs: Effect of Peptidase Inhibitors

Summary: In order to examine the role of peptidases in modulating bronchoconstrictor responses to iv administered capsaicin, a potent C-fiber stimulant, we measured changes in pulmonary conductance (G_L) and dynamic compliance (C_dyn) in anesthetized mechanically ventilated guinea-pigs. Control guinea-pigs, and guinea-pigs treated with the neutral endopeptidase (NEP) inhibitors thiorphan (1.7 mg/kg) or SCH32615 (1 mg/kg), the angiotensin converting enzyme (ACE) inhibitor captopril (5.7 mg/kg), or combinations of NEP and ACE inhibitors, were given increasing doses of capsaicin by rapid iv injection. The doses of capsaicin required to cause a 50% decrease in G_L and C_dyn (ED₅₀G_L and ED₅₀C_dyn respectively) were computed for each animal. None of the peptidase inhibitors, when given alone, had any effect on the changes in pulmonary mechanics induced by capsaicin. However, combined administration of thiorphan and captopril, or SCH32615 and captopril, caused a decrease in ED₅₀C_dyn for capsaicin, and prolonged the time during which the peak changes in G_L induced by capsaicin persisted. These data support the hypothesis that substances whose degradation is inhibited by combined NEP and ACE inhibitors contribute to the bronchoconstriction induced by iv administered capsaicin. This profile of enzymatic degradation is consistent with the tachykinin, substance P.     

Evaluation of nonthreshold leukemogenic response to methyl nitrosourea in p53-deficient C3H/He mice

The classic controversy of whether genotoxic chemicals induce cancers with or without a certain low-dose limit, i.e., the threshold, is revisited because of a number of current publications available addressing the plausibility of "practical" thresholds even for genotoxic carcinogens, the mechanism of which may be hypothesized to be due, in part, to a repair system composed of ordinarily available various defense mechanisms under the steady-state DNA damage. The question of whether an absolute nonthreshold or a relative nonthreshold, i.e., a "practical" threshold specifically in the low-dose level, is present may not be answered even with the use of a prohibitively large number of wild-type mice. Could the excessive incidence of tumorigenesis in p53-deficient mice contribute to our understanding of the threshold vs nonthreshold issue in genotoxic carcinogenesis? This is considered because an exaggeration of tumorigenesis in p53-deficient mice is hypothesized to reduce or eliminate the range of threshold due to the p53-deficiency-mediated reduction of DNA repair and apoptosis. The present study of chemical leukemogenesis in p53-deficient mice by transplantation assay was designed to answer this question. Briefly, 218 C3H/He mice were lethally irradiated and repopulated with bone marrow cells from wild-type, heterozygous p53-deficient, and homozygous p53-deficient C3H/He mice. This was followed by treatment with a single and graded dose of methyl nitrosourea at 6.6, 14.8, 33.3, 50.0, and 75.0 mg/kg body wt, with the vehicle-treated control groups treated with zero dose for each genotype. Whereas mice repopulated with p53-deficient bone marrow cells showed a marked reduction of the threshold for leukemogenicity, mice repopulated with wild-type bone marrow cells did not exhibit leukemia at a dose of 33.3 mg/kg body wt and showed a curve with a high probability for the linear regression model with a positive dose intercept, predicting a threshold by the likelihood ratio test. Thus, the failure of wild-type mice to show an increase in incidence of leukemogenesis at low doses of genotoxic carcinogens may be due not to a statistical rarity, but to various p53-related pharmacophysiological functions, possibly including DNA repair and apoptosis that may account for a threshold.     

Supplemental role of the Ames mutation assay and gap junction intercellular communication in studies of possible carcinogenic compounds from diesel exhaust particles

This study presents a new strategy for the carcinogenic evaluation of complex chemical mixtures based on genotoxic and nongenotoxic assays. We studied the ability of organic extracts of diesel exhaust particles (DEP) to induce point mutations in five different Salmonella typhimurium strains (Ames test) and to inhibit gap junction intercellular communication (GJIC) in rat liver epithelial cell lines. A crude extract of DEP was prepared by extraction with dichloromethane (DCM), and fractionated according to polarity into five fractions: aliphatic hydrocarbons, polycyclic aromatic hydrocarbons (PAH), nitro-PAH, dinitro-PAH, and polar compounds. Statistical experimental design, multivariate data analysis, and modeling were used to quantify the mutagenicity of individual and combined DEP fractions in the Ames assay. Quantitative determination of GJIC was carried out using a recently described combination of scrape loading and digital image analysis. Both assays responded to the DEP extract, but the responses were due to different fractions. The nitro-PAH fraction showed the strongest mutagenic potential, followed by the dinitro-PAH fraction. The effect on GJIC was due to the fraction containing the polar components, followed by the dinitro-PAH fraction. The extract was found to induce both basepair substitutions and frameshift mutations, through activation by bacterial nitroreductases. Hyperphosphorylation of connexin43, the major connexin in the epithelial cell lines, was less evident for DEP extract than for other communication inhibitors such as phorbol esters and growth factors, and consequently inhibitors of the protein kinase C (PKC) and mitogen-activated protein (MAP) kinase pathway were unable to counteract the inhibition by DEP extract. Since the Ames test is a well accepted method to screen for substances with genotoxic activity while inhibition of GJIC is associated with effect of tumor promoters and nongenotoxic carcinogens, it is not surprising but encouraging and interesting that the present data indicate that the two endpoints supplement each other as screening tests and in the evaluation of hazardous compounds in complex mixtures.     

An organizational approach for the assessment of DNA adduct data in risk assessment: case studies for aflatoxin B1, tamoxifen and vinyl chloride

The framework analysis previously presented for using DNA adduct information in the risk assessment of chemical carcinogens was applied in a series of case studies which place the adduct information into context with the key events in carcinogenesis to determine whether they could be used to support a mutagenic mode of action (MOA) for the examined chemicals. Three data-rich chemicals, aflatoxin B₁ (AFB₁), tamoxifen (Tam) and vinyl chloride (VCl) were selected for this exercise. These chemicals were selected because they are known human carcinogens and have different characteristics: AFB₁ forms a unique adduct and human exposure is through contaminated foods; Tam is a pharmaceutical given to women so that the dose and duration of exposure are known, forms unique adducts in rodents, and has both estrogenic and genotoxic properties; and VCl, to which there is industrial exposure, forms a number of adducts that are identical to endogenous adducts found in unexposed people. All three chemicals produce liver tumors in rats. AFB₁ and VCl also produce liver tumors in humans, but Tam induces human uterine tumors, only. To support a mutagenic MOA, the chemical-induced adducts must be characterized, shown to be pro-mutagenic, be present in the tumor target tissue, and produce mutations of the class found in the tumor. The adducts formed by AFB₁ and VCl support a mutagenic MOA for their carcinogenicity. However, the data available for Tam shows a mutagenic MOA for liver tumors in rats, but its carcinogenicity in humans is most likely via a different MOA.     

A comprehensive approach for integration of toxicity and cancer risk assessments

Experimental observations and theoretical considerations indicate a dose threshold for most chemically induced noncancer toxic effects below which the increased risk of toxicity is zero. Thus, the historical approach for minimizing risk from toxic chemicals has been to experimentally determine a no-observed-adverse-effect-level (NOAEL) and then to apply safety or uncertainty factors to estimate a dose not expected to produce that toxic effect in humans. In contrast, for radiation and chemically induced cancer, it has been assumed that all agents operate by a genotoxic mode of action and that some risk can be assigned to even vanishingly small doses. Accordingly, risk assessments for carcinogens have commonly been based on the assumption that the tumor dose-response curve at low doses is linear and passes through the origin. Mode of action is defined as a fundamental obligatory step in the induction of toxicity or cancer. It is now clear that tumor induction can arise in a variety of ways including not only a DNA-reactive genotoxic mode of action, but also non-DNA-reactive nongenotoxic-cytotoxic and nongenotoxic-mitogenic modes of action. Initial risk assessment approaches that recognized this distinction identified a chemical carcinogen as either genotoxic or nongenotoxic, with no middle ground. The realization that there is a continuum whereby different chemicals can act by a combination of modes of action and the recent explosion of research into molecular mechanisms of carcinogenesis indicate that all relevant information should be integrated into the risk assessment process on a case by case basis. A comprehensive approach to risk assessment demands that default assumptions be replaced with an integrated understanding of the rate-limiting steps in the induction of toxicity or cancer along with quantitative measures of the shapes of those dose-response curves. The examples of more contemporary risk assessments are presented for chloroform and vinyl acetate.