Cancer as an Environmental Disease

This book examines the relationship between environmental influences and the increasing incidence of cancer in humans. There are many changes taking place in the incidence and nature of a number of different diseases that face mankind. Because the environment in which we live is also in a state of flux, it is reasonable to enquire whether such changes might be associated with or indeed responsible for alterations in health patterns. In previous books in this series we have addressed the effects on health of pollution from waste (Nicolopoulou-Stamati et al., 2000) and also of chemical pollutants that disrupt hormones (Nicolopoulou-Stamati et al., 2001). Many of the effects addressed in those books had ‘soft’ endpoints that are awkward to assess, for example subtle changes in human neuro-behavioural patterns. With such endpoints it is notoriously difficult to design studies in human societies that will give emphatic answers, one way or the other. A major obstacle is the collection and maintenance of population-based data, for many of the areas of interest it simply does not exist.

Therefore the study of cancer and its interactions with environmental influences appears, on first inspection, to be a tractable question. After all the incidence of cancer is not a ‘fuzzy’ endpoint, an individual either has it or does not. The basic data is, in fact, binary. In addition most developed countries maintain good cancer registries covering the whole population by region and by age. What is beyond any doubt is that the incidence of cancer in modern industrialised societies is rising rapidly (WHO, 2003). And yet there exists an enormous diversity of opinion between different experts as to the extent of the influence of the environment in the aetiology of cancer. Some members of the scientific establishment state that less than 5 per cent of cancers can be associated with environmental impacts, while others maintain that over 80 per cent are so influenced. Where does the truth lie?

The relative contribution of genetic versus epigenetic factors continues to be a major area of controversy. Which particular environmental or epigenetic effects should be considered? Exposure to known carcinogens, be they chemical, physical or biological, is clearly relevant. What are the effects of the timing of exposure during different periods of life, when vulnerabilities can change? What proportion of the epigenetic influences are concerned with ‘lifestyle’, and therefore under the control of the individual, and what proportion are ‘imposed’ through unavoidable environmental pollution coming in air, water and food?

There remain many more questions than answers. This book does not pretend to be able to provide definitive answers nor address all the areas of concern; the latter would require a much heftier tome. However the contributors to this book do gather much of the available literature together and help to focus on what we do understand and what we do not know. Carcinogenesis is generally accepted as a multi-factorial, multistage process and has been the subject of an enormous amount of research. Many molecular mechanisms and processes have been proposed for cancer development; however the identification of the initiating event leading to malignant transformation remains obscure. In post-modern societies, people are constantly exposed to a variety of known or potential carcinogens. The identification of data gaps is important because it points to ways forward. Ultimately we try to assess the plausibility of the environmental carcinogenesis hypothesis, applying widely accepted criteria on causality. Much remains to be debated, but we consider that this book will be of use to those who want to have a rapid introduction to the range of arguments available in this important area of human health.

The analysis in Section 2 of this chapter presents several lines of evidence that implicate the environment in cancer causation; specifically, findings from wildlife studies, cancer trend reports, immigrant studies, childhood cancer studies and twin studies are reviewed. Having established the general evidentiary basis for the cancer-environment linkage, in Section 3 we turn to a discussion of the current methodological difficulties in incorporating the environmental context in the study of cancer. The particular focus here is on exposure assessment — a key methodological limitation in studying the cancer-environment linkage. It is reasonable to expect that cancer cases arising from point source environmental exposure will tend to cluster geographically. For this reason the role of exposure assessment and other methodological issues in the context of cancer cluster investigations are considered. The case of the Woburn Massachusetts leukaemia cluster is reviewed to illustrate some of the pertinent issues involved. In Section 4, we move to a general discussion of the implications of cancer risk assessment methodologies for cancer policy and intervention. In light of the observational evidence concerning the cancer-environment link (Section 2), as well as the uncertainties involved in assessing the risks associated with environmental carcinogens (Section 3), it is suggested that the precautionary principle be adopted as a guiding principle for cancer policy and intervention. The precautionary principle calls for protective action, even when the evidence of harm remains inconclusive and the adoption of this principle seems warranted under the present technical and policy circumstances. The final section ends the chapter with some recommendations and concluding remarks.

This chapter addresses the evidence available for assessing the incidence of cancer in pre-industrial societies. We conclude that cancer rates must have been very low in pre-industrial societies living traditional lifestyles, and this is considered in the light of the increasing average life expectancy experienced in industrialised societies during recent times. Cancer incidence rates are compared with the more widely reported 5-year mortality rates. Recent epidemiological evidence, from twinning studies, for the connection between human cancer and environmental factors is discussed.

A major section of the chapter is devoted to examining scientific evidence of the feasibility that low dose exposure to environmental pollutants, during critical periods of development, may be able to cause cancer later in life. This includes reviewing a selection of papers involved with hormone disruption and non-genotoxic causes of cancer. We conclude that there is a feasible mechanism whereby current levels of intrauterine exposure to xeno-chemicals could be consistent with carcinogenesis.

Finally, we comment on the public perception of cancer and on the widespread reporting practice of avoiding public discussion in the media of the rising incidence of cancer within the general population. An example of a recent press report is given. We further conclude that there is a need for improved public awareness of the increase of cancer in human societies, before there will be any demand for policies designed to lead to preventative measures.

Environmental pollutants have a strong impact on human health. It has been estimated that up to 90 per cent of all human cancers are due to exposure to chemical carcinogens. These chemicals fall mainly into the category of epigenetic or genotoxic carcinogens. The target of the genotoxic chemicals is clearly DNA. Reactive chemicals or metabolically produced intermediates bind covalently to DNA bases, and several nucleophilic substitution sites in DNA have been noted. The adduct formation in DNA is not a random process, when it is a question of bulky aromatic hydrocarbon like benzo[a]pyrene the process is highly stereoselective. The damage in DNA may be repaired efficiently and with high fidelity depending on the three-dimensional shape and location of adducts. However, if the repair fails for some reason, the DNA modified by the carcinogen may cause a mutation in cell replication. The mutation may bring about cell dysfunction which is the initiation phase of tumourigenesis. The challenge of the environmental exposure-related cancer studies are that the biological outcome is a result of the interaction of the exposure and genome of an individual. This underlines the supposition that gene-environment interactions are at the core of a specific disease outcome.

Large inter-individual variation has been observed in the metabolism of chemical carcinogens, and the genetic basis for some of this variation has been identified. Gene-environment interaction is important in the development of cancer, e.g. genetic polymorphism in enzymes involved in activation and deactivation of environmental carcinogens influences the level of the ultimate carcinogenic compound leading to chromosomal damage, and, subsequently, an altered risk of developing cancer. The effect of the genetic polymorphism depends both on the chemical identity and the level of exposure. In case of exposure to polycyclic aromatic hydrocarbons, smokers with either the CYP1A1*2 or the GSTM1*2/*2 genotypes have an increased cancer risk. However, the risk associated with a genetic polymorphism in a single gene is not very high, as the enzyme only represents one of the many pathways of metabolism. Combining information on genetic polymorphism in many genes involved in competing pathways of biotransformation, i.e., gene-gene interactions, significantly improves the risk-estimate, e.g. smokers possessing both GSTM1*2/*2 and CYP 1A*2 had a significantly higher risk of developing cancer than smokers with only one of the genotypes. Genetic polymorphism in other enzymes involved in xenobiotic metabolism is also associated with an increased risk of different types of cancer, but in each case it is important to focus on both the type and level of exposure, as the genes by themselves do not have any significant effect on the risk of cancer development.

This paper deals with the quantification and the evaluation of cancer risks which are associated with incidental exposures resulting from environmental accidents. To this end, the general risk assessment methodology is applied to the PCB/dioxin exposure incident in Belgium during the period January to June 1999. On that occasion, mineral oil from an electric transformer containing approximately 50 kg of PCBs, 1 g TEQ dioxins and about 2 g TEQ dioxin-like PCBs was added to the food chain.

The hazard identification shows that 2,3,7,8 TCDD is carcinogenic in humans. It is logical to estimate the cancer risk of the other dioxins and PCBs on the basis of their TEQ-value. Non-cancer effects of dioxin/PCB exposure include immunotoxic effects, impacts on brain development and thyroid metabolism, endometriosis and other effects on the reproductive system. The general pattern which emerges from this analysis points to the characteristic effects of endocrine disrupters.

The action mechanism of cancer generation by TCDD does not reveal any argument against a linear-effect relationship without a threshold dose. As a complement to that, maximum limits for dioxins in food have been established. They are based upon the consideration that daily intakes should not exceed 1–4 pg I-TEQ/kg body-weight/day. Also for the PCBs cancer risk, a linear dose without threshold can be accepted.

The exposure assessment shows that, in the past, Belgians faced high background exposure to dioxins and PCBs, resulting in an estimated body burden of 6.88 ng ITEQ/ kg body weight. The incident increased this level on average by 7 per cent. For PCBs, the incident-related increase amounted to 42 per cent.

On the basis of this data, the extra number of cancer deaths resulting from the 1999 Belgian incident is estimated to range between 44 and 8316. Also, changes in the thyroid metabolism and changes in immunological response in babies should be expected as a result of the incident.

These estimations are characterised by uncertainties. The most important concern the insufficiently-detailed exposure assessment, the lack of data on dioxins and PCBs in food during the crisis, the lack of specific risk figures for acute exposure episodes and the lack of knowledge, in particular of the PCB action mechanism.

Overall, the analysis conclusively shows that in a densely-populated, industrialised country such as Belgium, background levels of PCB and dioxin exposure are high. Therefore, any increase in exposure results in a parallel increase, not only in cancer risk, but also in the risk of non-cancer effects. Therefore, and because there is a clear trend of identifying dioxin and PCB-related health effects at decreasing doses, it is of main importance to keep the food chain-related exposure limited. This necessitates lower dioxin values than those the EU put in place on July 1^st, 2002. Also the cost/benefit analysis of the introduction of waste products into the food chain should be reviewed.

Since the nineteen eighties, public health agencies have emphasised the benefits of consuming a diet rich in fresh fruits and vegetables. Vegetables and fruits contain antioxidants, which are known scavengers for reactive oxygen species (ROS). Specific phytochemicals found in cruciferous vegetables are also believed to lead to the induction of phase II enzymes involved in the detoxification of reactive intermediates of chemical carcinogens. Decreases in levels of ROS, and increases in detoxifying enzymes are thought to be the basis for reduction in the risk of developing cancer by dietary means. Animal and epidemiological studies evaluating diets rich in cruciferous fruits and vegetables confirm the cancer preventive effects of such diets. On the basis of the beneficial results realised from these specific diets, one approach to cancer prevention includes dietary supplementation with specific phytochemicals and antioxidants, including beta-carotene (! CT) and various other carotenoids. Unexpectedly, two recent chemoprevention trials, the Alpha-Tocopherol Beta-Carotene Trial and the Beta-Carotene Retinol Efficacy Trial, showed that ! CT, either alone or in combination with vitamins A and E, could actually increase lung cancer risk and mortality in heavy smokers and in asbestos workers. These studies suggest that ! CT possesses co-carcinogenic properties. Similarly, several studies have documented the mutagenic and cancer-promoting activities of anti-oxidants and glucosinolate metabolites derived from cruciferous vegetables. In vitro and in vivo studies demonstrated that some anti-oxidants, while reducing ROS, also could be co-carcinogenic by up-regulating carcinogen metabolising enzymes and by the generation of reactive oxygen species. In contrast to the benefits of a complex diet rich in fruits and vegetables, individual supplementation with anti-oxidants may increase cancer risk by increasing levels of known carcinogenic metabolites.

Cancer is a complex issue that has been the subject of observation and research for many disciplines which have investigated the different aspects of cancer aetiology. The predominant theory for the past 50 years has been that cancer is the result of cumulative mutations that alter specific locations in a cell’s DNA and which alter the proteins encoded by cancer-related genes. Susceptibility to mutation has been mainly researched with respect to people’s genetic make up and their lifestyles, which are largely dictated by behavioural patterns. However, a growing body of scientific evidence strongly implicates the environment in the causation of cancer. Findings from studies of wildlife, cancer trends, human migration, childhood cancer, twinning and industrial accidents suggest that concentrating solely on genetic origin and behavioural pattern as causes of DNA defects, and consequently the main causes of cancer, needs to be re-evaluated. There is now clear scientific evidence implicating exposure to radiation and man-made chemicals as major causes of mutation. In addition there is evidence that non-genotoxic mechanisms causing tissue dysgenesis might be a significant aetiological mechanism during development and this must become an element of any future re-evaluation.

DNA has a tremendous capacity for repair but if this fails for some reason, the DNA modified by carcinogen(s) may cause an abnormality in cell replication. The biological outcome cannot be predicted, as it is a result of the interaction of the exposure of an individual with his or her specific genome. Exposure of the individual to environmental carcinogens may start before birth, during critical periods of development, and this carries implications for the body’s natural defence system.

This paper applies the Bradford-Hill criteria for causation to the question of whether there is sufficient evidence to conclude in favour of the environment as a main cause of cancer in humans. The analysis shows that the current levels of pollution are more likely causally related to the rise of cancer incidence than they are not.

Different methodologies exist now to measure the impact of the environment on cancer. However, the quantification and evaluation of cancer risks in humans can only be performed on populations, e.g., as a consequence of environmental accidents. Furthermore, the idiosyncratic nature of individual human beings, associated with large random variations, makes such phenomena difficult to assess.

This should contribute to a revised, more effective cancer policy in the future, where more emphasis is put on involuntary exposure to carcinogens. Among them, carcinogenic environmental pollutants are of major importance.