|
28 ELR 10254 | Environmental Law Reporter | copyright © 1998 | All rights reserved A Single Exposure to Many Carcinogens Can Cause CancerEdward J. Calabrese, Ph.D., and Robyn BlainEditors' Summary: This Dialogue addresses whether a single exposure to a carcinogenic agent can cause cancer. Using a database that contains nearly 5,000 studies that have assessed the ability of a single dose of a chemical to cause cancer, the authors discover that approximately 400 chemicals representing more than 30 chemical classes have been found to cause tumor development. These chemicals have been shown to induce tumors in numerous animal models, including more than 400 mouse and 100 rat strains, as well as in less frequently used models such as hamsters, guinea pigs, gerbils, rabbits, primates, and several fish species. Similarly, this phenomenon has been found in numerous tumor types, through multiple routes of exposure, in both genders, and during various stages of development and maturation. The authors argue that a single dose need not be exceedingly high nor approach acute toxicity in order to cause cancer. The implications of these findings for carcinogen testing and risk assessment procedures, particularly those concerning less than lifetime exposures, are also addressed in light of potential mechanistic interpretations by which single dose carcinogens can act. The authors conclude that episodic exposures to short-term high exposures may be an important yet generally overlooked area of concern. Dr. Calabrese is a board-certified toxicologist who is a professor of toxicology at the University of Massachusetts, School of Public Health, Amherst. He has researched and written extensively in the area of host factors affecting susceptibility to pollutants. He is currently chair of the Biological Effects of Low Level Exposures (BELLE) Advisory Committee, and is a former member of the U.S. National Academy of Sciences, NATO Countries Safe Drinking Water committees, and the Board of Scientific Counselors for the Agency for Toxic Substances and Disease Registry (ATSDR). Robyn Blain received her B.A. at Mount Holyoke College in 1990 and her Master's degree from the University of Massachusetts in 1995. She is in the process of completing her Ph.D. at the University of Massachusetts and is expected to finish by fall 1998. Is it possible that a single exposure to a carcinogenic agent can cause cancer? This has been an often debated question, but one that traditional testing protocols have not been designed to answer. The typical rodent cancer bioassay1 involves daily exposure to the maximum tolerated dose and fractions thereof for about two years. Nevertheless, continuing concerns remain over whether a single exposure would be able to cause cancer, especially in light of the episodic nature of many human exposures to environmental, occupational, and medicinal carcinogens. Over the past few years much interest has focused on evaluating the underlying principal assumptions used in the process of cancer risk assessment. The overwhelming impression that has emerged is that there are multiple, highly conservative, and independent assumptions involved that, when taken collectively, are likely to interact multiplicatively and result in a cascading of risks that ultimately grossly exaggerate the actual cancer risks.2 Despite the impression that the principal assumptions employed in the risk assessment process are highly conservative, however, a number of factors inherent in the process of carcinogenesis testing and risk assessment estimation are also likely to contribute to an underestimation of risk. These factors may include post-weaning exposures that miss a developmental period of possible enhanced susceptibility to a variety of agents;3 the use of homogeneous environmental conditions with respect to diet, temperature, and lighting; the lack of disease; and the testing of single agents despite normal exposure to multiple agents. Tumor formation resulting from a single exposure to a carcinogen is an important public health issue because exposure patterns within the human population are highly variable. Transitory or episodic exposures to elevated pollutant levels may occur, for example, because of one's occupation or hobbies, hazardous spills and their cleanup, certain medicines and medicinal treatments, and numerous other experiences in which one is inadvertently exposed to a relatively large dose of a carcinogenic agent for a brief period of time. This Dialogue documents numerous experimental studies in the published literature where animal models have [28 ELR 10255] been assessed for tumor development following the administration of a single exposure4 to a carcinogenic agent without the assistance of an exogenous promotional agent or a promotional procedure.5 Because the search for single-exposure carcinogen studies having these criteria does not lend itself to easy identification in a key word computer-assisted search of cancer literature, this proved to be a difficult endeavor. Nevertheless, a surprisingly large number of studies were discovered in which a single exposure to a carcinogen was administered without the additional treatment of a promotional agent or process. Before providing the results of the assessment, it is necessary to know that the following evaluation is based on information contained in a developing relational retrieval database. At the time of this writing, the database contained the results of more than 5,000 studies that used single-exposure carcinogen experimental protocols discovered in articles published in the open-literature, principally in widely recognized toxicological and cancer-oriented peer-reviewed journals. From each study satisfying the single-exposure carcinogen bioassay criteria, information was abstracted from numerous fields, including citation information, the subject's age at exposure, gender, chemical name, the type of statistical analysis employed, the route of exposure, the animal model, the type and quality of control groups, the type of pathology performed, dose response relationships, the capacity for time-to-tumor evaluations, the capacity to assess a single versus a fractionated dose, the relationship of a single dose to the dose at which 50 percent of the subjects die within a specified time (LD50), the number of doses per experiment, the number of animals per dose, and the type of tumor produced. Additionally, information has been included on the chemical class as well as on numerous physical and chemical characteristics of the chemical tested. With such a multitude of subject areas, it is possible to conduct an extremely wide array of queries depending on the interest of the investigator. This query system was used to provide the descriptive information provided in the Dialogue. It is difficult to make generalizations about the entire database because each study was designed for the testing of a specific hypothesis unique to the investigators. In general, the principal issue addressed in this Dialogue is whether the collective database allows one to judge properly whether a single exposure to a chemical carcinogen causes cancer in an animal model. While one may legitimately question the adequacy of selected individual studies because of sample size, lack of concurrent controls, lack of an adequate number of doses to establish a dose response, and other limitations, these studies, when taken collectively, represent a strongly supportive and extremely robust database for assessing the general hypothesis posed. Agents Causing Cancer With a Single Dose This assessment revealed that nearly 400 chemical agents cause tumorigenesis in animal models, including mammals and fish, with but a single exposure. With respect to mammalian studies, these agents represent several dozen chemical classes. Table 1 lists the chemical classes in which single-exposure carcinogens have been reported and the number of chemicals per class that were positive. Most of these agents are believed to be genotoxic6 carcinogens, which require bioactivation7 to an ultimate carcinogenic form, while others are thought to be direct acting, which do not need bioactivation. In addition to the wide range of chemical classes, a substantial number of these agents are generally recognized as having community and occupational exposure relevance, such as Benzo(a)pyrene, a by-product of burning fuel, and are not simply relegated to highly restricted industrial or experimental processes. Role of Species and Strain Susceptibility One of the most striking aspects of this review is that single exposures to numerous carcinogens have been shown to cause cancer in a wide range of species or strains.8 More specifically, positive studies have been reported in nearly a dozen species, including mice, rats, hamsters, gerbils, rabbits, guinea pigs, opossums, and fish, as depicted in Table 2. Furthermore, positive findings have been reported in 402 strains and/or substrains of mice and 133 strains and/or substrains of rats. Such similarities in cross-species and cross-strain responses support the premise that susceptibility to a single carcinogen exposure may be a broadly generalizable phenomenon, and suggest that humans are likely to respond in a qualitatively similar fashion. In fact, positive findings are reported in species of generally widely varying susceptibility to genotoxic and carcinogenic agents. Consequently, there is no basis to support the assumption that only very sensitive models respond positively in single-exposure carcinogen protocols. Another way to evaluate the role of species or strain susceptibility by using the information contained in the database is to determine how many different species or strains have been involved in the evaluation of a particular agent. Table 3 lists 10 different single-exposure carcinogens and the animal models in which the chemicals tested positively. Of particular interest is the remarkable number of different species and strains in which each of these carcinogenic agents tested positively in single-exposure carcinogen bioassays. The consistent response across such a large number of animal models represents a powerful argument for the broad interspecies generality of these findings. Single-Exposure Studies in Fish Models Although this assessment primarily focuses on mammalian responses to single-carcinogen exposures, a sizeable [28 ELR 10256] number of positive studies in the database involve fish models. For example, when rainbow trout were exposed in water to a number of carcinogens—aflatoxin B[1] (AFB[1]), diethylnitrosamine (DEN), dimethylnitrosamine (DMN), and methylazoxymethanol acetate (MAM)—on a single occasion for durations lasting between 30 minutes and 24 hours, marked increases in a variety of tumor types were observed. Exposing trout embryos to a solution containing 0.5 parts per million (ppm) of AFB[1] for only 15 minutes produced a 62 percent tumor incidence 12 months later.9 In addition, a single intraperitoneal injection of labeled 10 micrograms AFB[1] in rainbow trout caused a persistence of liver deoxyribonucleic acid (DNA) adducts10 three weeks after treatment. The rainbow trout's heightened sensitivity is related both to its enhanced capacity to develop initial macromolecular damage and its limited capacity for DNA repair.11 Other methodological approaches involving the rainbow trout revealed that single exposures can cause cancer. Using a micro injection of nanogram quantities of test carcinogens into embryonated fish eggs, the carcinogens AFB, dimethylbenzanthracene (DMBA), and 2-anthramine-induced liver tumors after one year,12 and positive findings were obtained using the same technique in rainbow trout with polycyclic aromatic hydrocarbon and nitrosamine carcinogens after nine months.13 In addition to the research involving rainbow trout, a single two-hour exposure of MAM in doses between 50-100 milligrams per liter in specimens aged 6-10 days' old induced cancer in seven fish species: Fathead Minnow, Gulf Killfish, Guppy, Inland Silverside, Matico, Minnow, Rivulus, and Sheepshead.14 These findings surrounding fish species are consistent with the mammalian database findings—a single exposure to a carcinogen can cause a significant increase in cancer incidence. Influence of Dose and Dose Response Relevant to this discussion is the claim that it is not surprising that a single exposure of a carcinogen could cause cancer if the dose were sufficiently high. The implicit assumption in such a statement is that the single dose used must be exceedingly high and approach acute toxicity by itself. In order to assess the hypothesis that only high doses can cause cancer with a single dose, each study indicating that a single exposure caused cancer in a mammalian model was analyzed in order to determine the relationship of the doses employed with respect to the LD50. Although many studies do not report or provide reference to LD50 values for the agent and animal model under the conditions of their study, such comparisons were available for a number of agents. As expected, it was common to see investigators use doses that approached the LD50 value—doses ranging between 0.1 of the LD50 to the LD50 value itself. A very high tumorincidence was usually noted in these cases. However, a number of studies exist in which the dose tested was below 1/50 of the LD50.15 This illustrates that single doses of these agents, some approaching levels as low as 1/125 of the LD50, have caused cancer in the respective animal models. One study demonstrated that the ratio of the chronic oral no observed adverse effect level (NOAEL) to the oral LD50 for a large number of agents approximate 1/50 to 1/75 of the LD50 using the geometric mean as the measure of central tendency.16 Even in experiments where the LD50 values are not provided, the dose used typically caused no lethality in the initial months following treatment. Such observations indicate that doses that cause tumors with a single administration typically did not cause even a measurable elevated risk of acute mortality. These studies reveal that single nonlife-threatening doses can induce tumors. Despite this conclusion, however, a single exposure far below the LD50 value with no measurable risk for causing mortality from acute exposure may still be well above ambient exposures. That is, such an apparent "low" exposure—one having a very low fraction of the LD50-may be higher than normal community or work environment exposures. For this reason, it is critical that the entire dose response continuum be assessed. While it has been assumed by U.S. regulatory agencies that cancer risk is linear at low doses, there are sufficient examples in the cancer bioassay literature to challenge the generality of this pervasive belief. In fact, dose response relationships for the various stages of cancer—initiation, promotion, and tumor formation or progression17—have been "U" or "J"-shaped in numerous experimental studies.18 In such cases, the lowdose [28 ELR 10257] groups display a noticeable reduction in response relative to the controls, while at the higher end of the dose spectrum, the enhancement of tumor-related endpoints occurs. Further, low-dose treatment groups display less damage than both the unexposed and high-dose treatment groups. Such U-J shaped responses occur in both single-and multiexposure experimental protocols. This type of response is commonly referred to as hormesis, a widely recognized phenomenon that occurs independent of the chemical, animal model, and endpoint, including cancer. The hormesis hypothesis states that most, if not all, chemical and physical agents, such as radiation, have the capacity to stimulate biological effects at doses below the toxicity threshold, while causing toxicity at doses above the threshold.19 In the case of hormesis and cancer, it is necessary that the control group have a high background incidence in order to assess the hormesis hypothesis. However, there is no biological basis to believe that hormesis would only occur when the background tumor incidence is high. An adaptive response induced by low-level chemical or radiation stress would be expected to be independent of the incidence background. Based on the above information, the authors believe that there is no contradiction between the concepts of hormesis20 and single-exposure carcinogenesis. A single dose causing cancer may be low in comparison to an LD50 value, but would be greater than a dose that would have a hormetic response for the same agent in the same experimental system. However, the two concepts principally, although not necessarily exclusively, emphasize different points on the dose response spectrum. Single-Exposure Carcinogenesis Does Not Equal Single-Hit Theory of Carcinogenesis Even though the single-exposure carcinogen database has several thousand positive studies, it does not automatically mean that the single-hit theory of carcinogenesis21 is now overwhelmingly supported. In fact, one study has clearly differentiated the two concepts.22 This study sought to determine whether the process of carcinogenesis was more consistent with the single-hit or multistage theory of carcinogenesis. They assessed the capacity of the carcinogen DMN to induce kidney tumors in a rat model that never spontaneously developed kidney tumors, meaning that there was a 0 percent incidence of kidney tumor development in unexposed controls. After the rats were given a single exposure to DMN, the rats were sacrificed at various times corresponding to the different stages of the kidney carcinogenesis process. Under the single-hit theory of carcinogenesis, DNA-adduct formation would occur during the early stage, foci formation would occur at the middle stage, and tumor formation would occur during the late stage. The authors of the study supported a linear dose response relationship for both adduct and foci formation—observations that were consistent with the single-hit theory of carcinogenesis.23 However, the kidney tumor yield during the late stage was strikingly nonlinear. This observation is more consistent with the traditional sigmoidal nature of the dose response curve. The lack of linearity for the tumor dose response led to the conclusion that the process of carcinogenesis is clearly multistage rather than single hit in nature. However, demonstrating that a single-exposure experimental protocol24 follows a multistage theory of carcinogenesis [28 ELR 10258] separates the concept of single-exposure carcinogenesis from the single-hit theory. A wide variety of routes of administration have been employed in studies assessing single-exposure carcinogens. These include oral exposures principally by gavage; dermal exposures via skin application; injections by any of a variety of routes, including under the skin, into the peritoneal cavity, into the muscle, or into the vein; respiratory routes via inhalation studies or via direct injections into the respiratory tract; and selective types of implantations. While implantation and injection exposures have limited quantitative relevance for environmental exposures, oral, inhalation, and dermal exposures are more directly related to typical human exposures. Age has been reported to be an important factor affecting susceptibility to chemically induced cancer.25 The assessment of mammalian studies under the database revealed numerous examples of fetal (transplacental), neonatal, and adult single-exposure studies. However, as shown in Table 4, the number of agents tested and the number of species or strains used were clearly greater for adults followed by neonates and then transplacental exposure. While observations that a single exposure causes cancer have not been specifically addressed in the risk assessment process, predicting lifetime cancer risks from short-term exposures to cancer-causing agents has been the object of some debate. In the absence of an appropriate database, a number of groups involved in cancer risk assessments have simply partitioned the total dose received over a short period of exposure, such as a single dose, for an equal but considerably lower exposure on a daily basis for an entire lifetime. This procedure is performed because the cancer bioassay is typically based on near lifetime exposures.26 Experiments in the field of radiation carcinogenesis have revealed that fractionation of dose for low linear energy transfer radiation results in a lower incidence of cancer than a single, massive exposure.27 While the radiation experiments may question the validity of risk assessment practices that fractionate a single dose over a lifetime, the database may permit a direct comparison of this issue for chemical carcinogens. Several dozen studies that have been published concerning singleexposure carcinogenesis and dose fractionation are included in the database. Such studies may permit an evaluation of hypotheses related to dose fractionation, especially in relation to potency factors. Given the controversial nature of this topic, it is important to address the quality of the data included in the single-exposure carcinogen database. The widespread publication of positive single-exposure studies reveals that the phenomenon repeatedly passed the peer-review process of multiple editorial boards and individual reviewers. Moreover, the majority of these findings have been published within the past 25 years. In addition to such general considerations that speak to database quality issues, it is of value to consider specific characteristics of the studies in the database. Evaluation of the single-exposure carcinogen database reveals a wide range of experimental protocols that depend on the intent of the investigators. Analysis of the nearly 5,000 studies concerning single-exposure carcinogens included an examination of the sample sizes, the nature of control groups, the number and range of dose levels, and the type of statistical analysis of the data. These analyses revealed that 38 percent—more than 1,400—of the positive studies had over 30 animals per treatment group, and nearly 18 percent—700—had more than 50 animals per treatment group (Table 4). Fifty-two percent of the positive studies had concurrent controls, while 42 percent did not report a control group. Twenty-one percent of the positive studies used at least two treatment groups, while 12 percent of the positive studies had three or more treatment groups. The type of statistical analysis varied markedly amongst the studies, with 45 percent utilizing hypothesis testing. A separate analysis of studies with negative results in the single-exposure protocol provided comparable percentages of the evaluated criteria. Despite the fact that approximately 400 agents representing approximately 30 chemical classes have been shown to cause cancer in a wide variety of animal species and strains with but a single exposure without the additional application of an exogenous promotional stimulus, it is a little recognized and appreciated massive series of observations. In fact, no review paper was found that addressed the phenomenon that a single exposure to a carcinogen can cause cancer in anything more than an extremely limited manner. Most research has focused on the realization that the process of carcinogenesis is a multistage phenomenon that includes initiation, promotion, and progression. It has generally been recognized that the process of carcinogenesis involves an initiation stage, including "fixation" of the genetic alteration, followed by a rather prolonged period of promotional stimulation. That initiation could in fact result from a single "subcarcinogenic" dose as long as the promotional stimulation was adequate. Optimal promotional exposures to agents, such as the enhancement of skin tumor development, have involved administration of such agents usually twice per week for at least four months. Given the historical context that the two-stage model of carcinogenesis28 is derived from studies of skin cancer in various mouse strains, it is not unexpected that the overwhelmingly dominant evidence [28 ELR 10259] of tumor promotion is within the skin model, although strong efforts have been directed to other systems, especially the liver and colon. Experimental systems have been designed to separate initiation from promotion so that these events could be clearly differentiated. In addition, emphasis has been placed on the temporal nature of promotional exposures. Within this experimental context, much has been learned about the process of carcinogenesis—including the sequential steps of promotion. A consistent observation is that promotional events are generally reversible. Another consistent observation is that for tumorigenesis to proceed, a consistent exposure to a promoter is required. Eventually, the initiated and promoted mass of cells becomes transformed into a malignant stage via a process called progression. The capacity to affect progression has been found to vary significantly amongst promotional agents. Despite the strong emphasis on understanding the multistage process of tumorigenesis involving exposure to initiating, promoting, and progressing agents, little attention has addressed how specific carcinogens cause benign and malignant tumors to develop with only a single dose. The recognition that numerous agents can affect this process is testimony to its occurrence and generality. However, the mechanism by which single exposures cause cancer has been inadequately studied on multiple levels, including genetic, biochemical, and histopathological domains. The most direct way to account for the occurrence of single-exposure-induced cancer is to assume a genetic lesion is accompanied by massive tissue necrosis followed by extensive reparative synthesis.29 This would be comparable to the two-stage system of the rat liver. While this initiation-promotion explanation is likely to account for some studies in the single-exposure carcinogen database, it is unlikely to be an adequate explanation for a large proportion of the studies. Possible mechanisms by which a single exposure without an exogenous promotional stimulus can cause carcinogenesis include cell-cycle alteration and oncogene activation as seen in epidermal cells, receptor mediated promotion by an initiator, endogenous promotional stimuli, and activation of obligatory biochemical events in promotion. While a detailed explanation is beyond the scope of this Dialogue, these four additional mechanisms do not necessarily require a damage-or injury-based promotional mechanism and are consistent with substantial data, including those that show apparently nontoxic doses of some single-exposure carcinogens can cause cancer in animal models. An element missing from the current assessment is that of epidemiologic evidence to either support or challenge the concept that single exposures to chemical carcinogens can cause cancer in humans. Consequently, this section will briefly examine evidence relevant to the hypothesis that humans may develop cancer following a single exposure or "quasi"-single exposure to a chemical carcinogen. Data on this point are quite limited because chronic disease epidemiological studies, such as those involving cancer, are traditionally concerned with prolonged exposures. Because cancer generally develops many years after initial exposure, it may be difficult to relate the cancer with a short-term exposure. It is also possible that persons having very brief exposures of less than one year may tend to forget such short-term exposures during the reporting period in comparison to the more prolonged exposures. Nonetheless, attempts were made to identify reports in the occupational epidemiologic literature where a brief exposure was associated with the development of cancer. The term "brief" is obviously a relative term, and as used here describes durations lasting less than one year. For purposes of this Dialogue, exposure periods of the shortest duration were analyzed where there was some link between the exposure and the development of cancer. A limited number of agents were identified from evidence in which a brief occupational exposure lasting less than one year was implicated as the causal agent in the development of human cancer. These agents are benzene; beryllium; vinyl chloride; and the aromatic amines of benzidine, aniline, and 4-aminobiphenyl. While these suggestive studies are inadequate to sufficiently resolve causal issues, diethylstilbestrol (DES) data are more powerful. Substantial evidence exists indicating that the administration of DES during pregnancy may produce clear-cell adenocarcinoma of the vagina in female offspring.30 Clearcell adenocarcinomas of the vagina and cervix in young females, which had been considered a rare form of cancer, started to be observed with increasing frequency during the late 1960s and early 1970s in association with intrauterine exposure to stilbestrol and chemically related nonsteroidal estrogens. Because of the rare nature of the cancer, a DES registry was created to centralize data and to facilitate the acquisition of information on this disease. In 1974, a comprehensive analysis of 170 DES registry cases was published.31 Consideration of the 170 cases revealed a 50-fold variation in the total dose—300 milligrams was the lowest dose. The duration of therapy ranged from as short as seven days to nearly the entire pregnancy. These data demonstrate that very few exposures with some very low doses were capable of causing tumors in the offspring of treated mothers. Because DES causes tumors at a specific time period at low doses, it address two points of the single-exposure carcinogen database. The first being that the age of treatment is important, and the second being that humans are not exempt from the phenomenon that a limited exposure can cause cancer. Future Directions and Implications Some important biomedical implications of the single-exposure carcinogen findings and directions for future research were identified through the assessment. For example, the role of episodic exposures to carcinogens may be more important than previously recognized. Particular attention needs to be directed to exposure patterns and susceptibility [28 ELR 10260] to chemically induced cancer. Because a number of agents can cause cancer in experimental animal models at nonacutely toxic levels based on a single exposure, these data suggest the need to prevent episodic exposures during accidental spills at work, and other similar circumstances. New approaches for predicting risk from single-exposure conditions also need to be addressed. The regulatory agencies' widely employed practice in risk assessment of fractionating the single day's dose over 70 years of life needs to be reevaluated. This is particularly true in the environmental arena, where single-exposure carcinogen data can be useful to those who are concerned with unexpected releases and spills. Developmental and age susceptibility for carcinogenic agents is likely to play an important role in enhancing an understanding of how some single exposures may cause cancer. These stages provide the potential for a high promotional stimuli (cell division) and, therefore, can facilitate in the process of carcinogenesis. If a single exposure to some agents causes cancer because of a period of special susceptibility, such as breast cancer in Sprague-Dawley female rats, then the summation of doses also applied in periods of very reduced sensitivity as is typically done may lead to an underestimation of actual risk. Such a summation of a total lifetime dose improperly reduces potency values. This clearly illustrates the importance of recognizing developmental periods of susceptibility in both study design and interpreting the cancer bioassay data. Last, if cancer can be induced bya single exposure, is it possible to alter traditional lifetime exposure protocols? This would have the advantage of reducing the cost of chronic bioassays as well as reducing the handling of the animals during the experiment. However, the data are insufficient to support such a recommendation. It is possible that potency and relative tissue sensitivity to a specific agent may significantly change as a result of different dosage patterns. Information on such parameters would be required in order to begin assessing the implications of single or limited numbers of doses as part of a standard protocol. In addition, the single-exposure scheme would appear to be relevant only for complete carcinogens, that is, those capable of both initiation and promotion. It is also possible that a single exposure to a chemical agent may not elicit the range of tumors that multiple exposures do. For example, it has been reported that while a single exposure of DMBA to female Sprague-Dawley rats caused mammary cancer, multiple doses were required for leukemia induction.32 This lack of multiple organ and multiple system cancer involvements—as in the case of a single dose of DMBA causing only mammary cancer—can influence the U.S. Environmental Protection Agency's weight-of-evidence judgment for carcinogen ranking. A single exposure to more than 400 carcinogens representing 30 chemical classes can cause cancer in mammalian models. The level of exposure varied according to the study, but sufficient examples exist where exposure was less than 1/100 of the LD50 and approach that of a chronic NOAEL. These data suggest that episodic exposures to short-term high exposures may be an important yet generally overlooked area of concern. It is recommended that a new direction in cancer studies occur in which the notion of episodic exposures are incorporated into study designs for animal testing protocol. In addition, environmental/occupational epidemiological studies should take greater pains to document episodic exposure patterns because these data suggest that they may play a greater than previously recognized role in chemically induced cancer. Table 1 Listing of chemical classes with the number of chemicals per class that were demonstrated to be positive in at least one study in the Single-Exposure Carcinogen Database
List of species with the number of strains and/or substrains that had positive results in the Single-Exposure Carcinogen Database
Table 3 List of selected chemicals with the strains of animals where the chemical was found to be positive in the Single-Exposure Carcinogen Database
Description of the Single-Exposure Carcinogen Database with number and percent of studies in various categories where a single dose was used with a positive outcome
2. A.L. Nichols & R.J. Zeckhauser, The Dangers of Caution: Conservation in Assessment and the Mismanagement of Risk, 4 ADVANCES IN APPLIED MICROECONOMICS 55 (1986); A.L. Nichols & R.J. Zeckhauser, The Perils of Prudence: How Conservative Risk Assessments Distort Regulation, 8 REGULATORY TOXICOLOGY AND PHARMACOLOGY 61 (1988). 3. EDWARD J. CALABRESE, AGE AND SUSCEPTIBILITY TO TOXIC SUBSTANCES (1986). 4. A single exposure can be conducted through a variety of methods, including single gavage, injection, dermal application, and placing the subject in the inhalation chamber for a limited time (e.g., 24 hours or less). 5. An example of an exogenous, or external, promotional agent includes croton oil and its derivatives, as in the case of the two-stage model of carcinogenesis. Another example of a promotional procedure would be partial hepatectomy (typically removal of up to two-thirds of the liver). 6. Genotoxic means that the compound damages the deoxyribonucleic acid (DNA). 7. Bioactivation is the process in which an organism transforms a compound into a more biologically active compound. Therefore, the transformation may allow the compound to cause more damage. 8. A species is a type of animal model (i.e., a rat or mouse). A strain is a specific animal type, such as Strain A mice. And a substrain is typically from a specific laboratory that may have slight differences due to long periods of breeding within the lab (i.e., A/J mice are Strain A mice that are from Jackson (J) Laboratory and have been bred there for multiple generations. Therefore, they are typically referred to as A/J mice and not simply Strain A mice). 9. Jerry D. Hendricks et al., Rainbow Trout Embryos: Advantages and Limitations for Carcinogenesis Research, 61 NAT'L CANCER INST. MONOGRAPHS 129 (1984). 10. A DNA adduct is when a chemical mutagen binds to one of four base pairs in the DNA. An example would be a benzo(a)pyrene metabolite forming an adduct with guanine. In replicating the DNA during cell division, this adduct may cause a misreading of the DNA, leading to mutations within the DNA. 11. George S. Bailey et al., The Sensitivity of Rainbow Trout and Other Fish to Carcinogens, 15 DRUG METABOLISM REV. 725 (1984). 12. C.D. Metcalfe & R.A. Sonstegard, Microinjection of Carcinogens Into Rainbow Trout Embryos: An In Vivo Carcinogenesis Assay, 73 J. NAT'L CANCER INST. 1125 (1984). 13. John J. Black et al., A Reliable, Efficient, Microinjection Apparatus and Methodology for the In Vivo Exposure of Rainbow Trout and Salmon Embryos to Chemical Carcinogens, 75 J. NAT'L CANCER INST. 1123 (1985). 14. William E. Hawkins et al., Carcinogenicity Tests With Small Fish Species, 11 AQUATIC TOXICOLOGY 113 (1988). 15. E.g., H. Druckrey et al., Organotropic Carcinogenic Effects of 65 Various N-nitroso-compounds on BD Rats, 69 Z. KREBSFORSCH 103 (1967); James A. Swenberg et al., Quantitative Aspects of Transplacental Tumor Induction With Ethylnitrosourea in Rats, 32 CANCER RES. 2656 (1972); Ulrich Mohr & Jorn Hilfrich, Effect of a Single Dose of N-diethylnitrosamine on the Rat Kidney, 42 J. NAT'L CANCER INST. 1729 (1972). 16. David W. Layton et al., Deriving Allowable Daily Intakes for Systemic Toxicants Lacking Chronic Toxicity Data, 7 REG. TOXICOLOGY AND PHARMACOLOGY 112 (1987). 17. Initiation is an irreversible alteration of the genome. Promotion is a reversible process that further develops the process to a tumor. And progression is an irreversible process that typically describes the process that leads to the malignancy of a tumor. 18. For examples of experimental studies on initiation, see L. Camurri et al., Sister Chromatid Exchange in Workers Exposed to Low Doses of Styene, in SISTER CHROMATID EXCHANGES, PART B 957-63 (Raymond R. Tice & Alexander Hollaender eds., 1983); Kirk T. Kitchin & Janice L. Brown, Dose-Response Relationship for Rat Liver DNA Damage Caused by 49 Rodent Carcinogens, 88 TOXICOLOGY 31 (1994); Kirk T. Kitchin & Janice L. Brown, Response to: Dose-Response Studies of Genotoxic Rodent Carcinogens: Thresholds, Hockey Sticks, Hormesis or Straight Lines? BELLE NEWSL., Feb. 1995, at 16-19 (Biological Effects of Low Level Exposures Advisory Committee, Northeast Regional Environmental Public Health Center, University of Massachusetts, Amherst, Mass.); Kirk T. Kitchin & Janice L. Brown, Dose-Response Relationship for Rat Liver DNA Damage Caused by 1,2-dimethythydrazine. 114 TOXICOLOGY 113 (1996); H.E. Kleczkowska & F.R. Althaus. Response of Human Kerinocytes to Extremely Low Concentrations of N-methyl-N'-nitro-N-nitrosoguanidine, 367 MUTATION RES. 151 (1996); Y. Liu et al., O<6">-methylguanine-DNA Methyltransferase Activity in Human Buccal Mucosal Tissue and Cell Cultures. Complex Mixtures Related to Habitual Use of Tobacco and Betel Quid Inhibit the Activity In Vitro, 18 CARCINOGENESIS 1889 (1997). For examples of experimental studies on promotion, see Rory B. Conolly & Melvin E. Andersen, Hepatic Foci in Rats After Diethylnitrosamine Initiation and 2,3,7,8-tetrachlorodibenzo-p-dioxin Promotion: Evaluation of a Quantitative Two-Cell Model and of CYP 1A1/1A2 as a Dosimeter, 146 TOXICOLOGY AND APPLIED PHARMACOLOGY 281 (1997); Thomas D. Downs & R.F. Frankowski, Influence of Repair Process on Dose-Response Models, 13 DRUG METABOLISM REV. 839 (1982); Thomas Goldsworthy et al., The Natural History and Dose-Response Characteristics of Enzyme-Altered Foci in Rat Liver Following Phenobarbital and Diethylnitrosamine Administration, 5 CARCINOGENESIS 67 (1984); Werner K. Lutz et al., Dose Response for the Stimulation of Cell Division by Caffeic Acid in Forestomach and Kidney of the Male F344 Rat, 39 FUNDAMENTAL AND APPLIED TOXICOLOGY 131 (1997); OFFICE OF TECHNOLOGY ASSESSMENT, CANCER TESTING TECHNOLOGY AND SACCHARIN (1977); Henry C. Pitot et al., A Method to Quantitate the Relative Initiating and Promoting Potencies of Hepatocarcinogenic Agents in Their Dose-Response Relationships to Altered Hepatic Foci, 8 CARCINOGENESIS 1491 (1987). For examples of experimental studies on tumor formation, see J.J. Broerse et al., Mammary Carcinogenesis in Rats After X-and Neutron Irradiation and Hormone Administration, in 2 LATE BIOLOGICAL EFFECTS IONIZING RADIATION 13 (Int'l Atomic Energy Agency, Vienna, 1978); J.J. Broerse et al., Mammary Carcinogenesis in Different Rat Strains After Single and Fractionated Irradiations, COMMISSION OF THE EUROPEAN COMMUNITIES REPORT. LUXEMBOURG, COMMISSION OF EUROPEAN COMMUNITIES, DIRECTORATE GENERAL INFORMATION MARKET AND INNOVATION 155-68 (1982); J.J. Broerse et al., Mammary Carcinogenesis in Different Rat Strains After Irradiation and Hormone Administration, 51 INTERNATIONAL J. RADIATION BIOLOGY AND RELATED STUDIES IN PHYSICS, CHEMISTRY, AND MEDICINE 1091 (1987); Ralph R. Cook, Responses in Humans to Low Level Exposures, in BIOLOGICAL EFFECTS OF LOW LEVEL EXPOSURES: DOSE-RESPONSE RELATIONSHIPS 99-109 (Edward J. Calabrese ed. 1994); Richard J. Kociba et al., Results of a Two-Year Chronic Toxicity and Oncogenicity Study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Rats, 46 TOXICOLOGY ANDAPPLIED PHARMACOLOGY 279 (1978); S. Nesnow et al., Cyclopenta[cd]pyrene-induced Tumorigenicity, Ki-ras Codon 12 Mutations and DNA Adducts in Strain A/J Mouse Lung, 15 CARCINOGENESIS 601 (1994); Roger W. O'Gara et al., Induction of Tumors in Mice Given a Minute Single Dose of Dibenz[a,h]anthraccne or 3-methylcholanthrene as Newborns. A Dose-Response Study, 35 J. NAT'L CANCER INST. 1027 (1965); A.K. Prahalad et al., Dibenzo[a,l]pyrene-induced DNA Adduction, Tumorigenicity, and Ki-ras Oncogene Mutations in Strain A/J Mouse Lung, 18 CARCINOGENESIS 1955 (1997); Michael P. Waalkes et al., Cadmium Carcinogenesis in Male Wistar [Crl:(WI)BR] Rats: Dose-Response Analysis of Tumor Induction in the Prostate and Testes and at the Injection Site, 48 CANCER RES. 4656 (1988). 19. Edward J. Calabrese, Hormesis Revisited: New Insights Concerning the Biological Effects of Low-Dose Exposures to Toxins, 27 ELR 10526 (Oct. 1997). 20. Id. 21. The single-hit theory of carcinogenesis assumes that a single biological event takes place that alters the cell leading to tumor formation. Therefore, the tumor response is directly proportional to the dose (linear). The multistage theory of carcinogenesis assumes that several independent, progressive, stages are required. 22. H. Elizabeth Driver et al., Dose-Response Relationships in Chemical Carcinogenesis: Renal Mesenchymal Tumors Induced in the Rat by Single Dose Dimethylnitrosamine, 68 BRITISH J. EXPERIMENTAL PATHOLOGY 133 (1987). 23. The single-hit theory assumes a one-way irreversible progression to the development of tumors. Therefore, if there is a linear dose response in adduct and foci formation, the single-hit theory would have a linear dose response in tumor development. The multistage theory assumes that there is the same progression through DNA adducts, foci, and tumor development, but that the process can be reversed or halted at any stage during this process. Thus, the formation of foci does not necessarily lead to the formation of tumors. 24. Driver, supra note 22. 25. CALABRESE, supra note 3. 26. Dose fractionation does not have to occur over the lifetime. Dose fractionation simply takes a large dose and breaks it up, thereby giving several doses that equal the single dose over a period of time. The period of time may be daily for several days or weekly for several weeks or spread over months. 27. Arthur C. Upton. Influence of Dose Rate in Mammalian Radiation Biology Quality Effects, in DOSE RATE IN MAMMALIAN BIOLOGY, Conf.—680410, U.S. ATOMIC ENERGY COMMISSION DIVISION OF TECHNICAL INFORMATION, OAK RIDGE (D.G. Brown et al. eds., 1968). 28. The two-stage model of carcinogenesis is one type of multistage carcinogenesis. It typically means that the initiation was administered as a single dose followed by a series of repeat exposures to a promotional stimuli. 29. Necrosis is the death of cells or tissues through injury or disease, and reparative refers to the act of repairing damaged tissue through the replacement of cells. 30. Peter Greenwald et al., Vaginal Cancer After Maternal Treatment With Synthetic Estrogens, 285 NEW ENG. J. MED. 390 (1971); A.L. Herbst et al., Adenocarcinoma of the Vagina. Association of Maternal Stilbestrol Therapy With Tumor Appearance in Young Women, 284 NEW ENG. J. MED. 878 (1971); A.L. Herbst et al., Clear Cell Adenocarcinoma of the Vagina and Cervix in Girls: Analysis of 170 Registry Cases, 119 AM. J. OBSTETRICS & GYNECOLOGY 713 (1974). 31. Herbst et al., Registry Cases, supra note 30. 32. K.C. Silinskas & A.B. Okey, Protection by 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane (DDT) Against Mammary Tumors and Leukemia During Prolonged Feeding of 7,12-dimethylbenz(a)anthracene to Female Rats, 55 J. NAT'L CANCER INST. 653 (1975). 28 ELR 10254 | Environmental Law Reporter | copyright © 1998 | All rights reserved |