Christopher J. Portier

Discussion and summary

The presentations at this workshop demonstrate that, while there have been tremendous strides in the ability to cure childhood leukaemias, the understanding of the genetic and environmental causes of this class of diseases is limited. The epidemiological and the molecular evidence all suggest that childhood leukaemia derives from a multistage process where the initial event starting the process is either inherited or the result of a DNA damaging event during gestation(1). From that initial event, the progression to disease has to occur fairly rapidly because the peak incidence occurs very early in life at around 1–3 y of age depending upon the type of leukaemia(2,3). At later ages, the incidence drops off quite dramatically with a >90% smaller incidence beyond age 15. As discussed at the workshop and in the articles in this special edition, a large number of studies have been conducted to identify risk factors associated with and possibly causal of childhood leukaemia. It is clear from the reviews in this special edition that the only environmental factor that is clearly linked to childhood leukaemia and most likely causes a fraction of the cases is ionizing radiation(4). Extremely low-frequency electric and magnetic fields (EMFs) have also been associated with childhood leukaemia, but most reviews of this literature have found these studies to provide only limited evidence of an association, which is insufficient for a causal association(5). Other environmental factors such as radiofrequency EMF, air pollution, smoking, pesticides, herbicides, persistent organic pollutants, radon gas and solvents, have all yielded either very weak associations or no association at all(4). Indirect indicators of environmental exposures, such as proximity to nuclear power plants, have not provided any additional evidence of a causal exposure. A number of socio-economic factors have also been associated with childhood leukaemias, with mixed results depending upon the metric used for assessing socio-economic status. Although many studies of other cancers have shown an increase in risk with decreasing socio-economic status, childhood leukaemia has mostly shown the opposite with risk increasing as socio-economic status increases. Also, studies evaluating kindergarten attendance and the frequency of interactions during early age have suggested that a lack of interactions with others in early childhood could contribute to increasing the risk. Somewhat consistent with these findings is an association between childhood leukaemias and increased birth weight in infants. Several theories have been brought forward to account for this. The first is that, in combination, these studies of socio-economic factors, birth weight and kindergarten attendance suggest the increased risk of childhood leukaemia occurs if the child lives in a high hygiene environment and has little chance to be exposed to a wide variety of germs and viruses(6). The second suggests that in these more affluent environments, there is an increased exposure to growth factors during gestation and early life that exerts a proliferative stress on haematopoiesis that could clonally expand existing mutations that are either inherited or obtained in very early gestation(7). Given the current knowledge and the relative risks seen in the numerous epidemiology studies of childhood leukaemia, my best estimate is that the attributable fraction has been accounted for by <10%. This means that, for more than 90% of the cases, there is no known or even suggested cause. There has been an increase in childhood leukaemias worldwide that cannot be explained by improvements in diagnosis and tracking alone. This increase strongly suggests an environmental factor (used in the broadest sense to include food, drugs, etc.) is influencing the rate of onset of the disease. But with all of these studies done to date, why have the major risk factors not been identified? Although there have been a number of genetic markers that have been associated with childhood leukaemia, they are not very specific and are rather high penetrance in disease-free children. It is likely that the environmental risk factors for childhood leukaemia combine with genetic risk factors to increase the overall risk for the disease. The number of possible interactions between genes and the environment is virtually infinite and one would need to be extremely strategic if one is to move forward and actually find what is driving this increase in incidence. In that sense, it is necessary to look at what is already done and add to it based upon the knowledge gained from previous studies. The high-hygiene theory of childhood leukaemia is one that should receive some serious follow-up at this point(8). Some may argue that, if it were a virus or bacterial infection that was the largest risk factor for childhood leukaemia, it would have been seen by now. There is no doubt that considerable resources have been spent on looking into viruses and childhood leukaemia and nothing found. However, in addition to the theory outlined earlier, two additional issues suggest that a careful evaluation of viruses and childhood leukaemia would be beneficial at this time. First, many of the studies addressing viruses and childhood leukaemia are fairly old and did not have the benefit of modern technologies that may do a better job of identifying an underlying viral source. Second, the author has tried to find a virus that might cause the leukaemias when it may be the lack of an early lifetime exposure to a virus that is increasing the risk. For example, the countries with the greatest increase in childhood leukaemias over the last two decades are also the countries with the lowest prevalence of many common human viruses such as human papilloma viruses. In the hunt for the causes of childhood leukaemias, size matters. As noted by many speakers, because of the low incidence of these diseases, it will take a worldwide consortium to gain sufficient sample size and power to identify risk factors that have small relative risks. The starting point in these consortiums should be the use of genome-wide association studies to provide a broader number of possible target genes for further exploration linked to complete evaluations of the environments of these children. Using these tools, one should be able to identify a host of new gene and environment links to childhood leukaemias that could then be explored in a broader array of research in smaller studies, laboratories and elsewhere. Epigenetics is emerging as one of the major scientific advances of this decade. The understanding of the heritable control of gene transcription is opening new doors into the understanding of many cellular processes. Because the biochemical processes that are involved in epigenetics are subject to manipulation through environmental factors, this may be one of the most significant means by which the environment can alter disease rates including leukaemias. The degree to which some of the factors weakly linked to childhood leukaemias alter epigenetic control of gene transcription is an area that warrants additional resources in the search for the aetiology of childhood leukaemias. Finally, the revolution in creating induced pluripotent stem cells also holds hope for understanding the root causes of childhood leukaemia. It is likely that this disease results from some form of heritable genetic change, either changes in base-pair sequence or changes in gene transcription, at a very early stage in the development of haematopoiesis. Having access to induced pluripotent stem cells opens a wide array of laboratory studies linked to existing or historical case–control epidemiology studies. In virtually every society on this planet, children are valued and given high priority for protection. Using common scientific resources to address one of the major deadly diseases of childhood is reasonable as a means to prevent the disease. Also, because childhood leukaemias develop so rapidly from the first division after fertilization (<4 y), it is possible that a thorough understanding of childhood leukaemias could have important implications for the understanding and prevention of other cancers. Regardless of why this research is pursued, it is clear that a renewed effort in this area, and a much broader consortium of scientists pursuing the causes of childhood leukaemias is needed.

Compound Cytotoxicity Profiling Using Quantitative High-Throughput Screening

Background The propensity of compounds to produce adverse health effects in humans is generally evaluated using animal-based test methods. Such methods can be relatively expensive, low-throughput, and associated with pain suffered by the treated animals. In addition, differences in species biology may confound extrapolation to human health effects. Objective The National Toxicology Program and the National Institutes of Health Chemical Genomics Center are collaborating to identify a battery of cell-based screens to prioritize compounds for further toxicologic evaluation. Methods A collection of 1,408 compounds previously tested in one or more traditional toxicologic assays were profiled for cytotoxicity using quantitative high-throughput screening (qHTS) in 13 human and rodent cell types derived from six common targets of xenobiotic toxicity (liver, blood, kidney, nerve, lung, skin). Selected cytotoxicants were further tested to define response kinetics. Results qHTS of these compounds produced robust and reproducible results, which allowed cross-compound, cross-cell type, and cross-species comparisons. Some compounds were cytotoxic to all cell types at similar concentrations, whereas others exhibited species- or cell type–specific cytotoxicity. Closely related cell types and analogous cell types in human and rodent frequently showed different patterns of cytotoxicity. Some compounds inducing similar levels of cytotoxicity showed distinct time dependence in kinetic studies, consistent with known mechanisms of toxicity. Conclusions The generation of high-quality cytotoxicity data on this large library of known compounds using qHTS demonstrates the potential of this methodology to profile a much broader array of assays and compounds, which, in aggregate, may be valuable for prioritizing compounds for further toxicologic evaluation, identifying compounds with particular mechanisms of action, and potentially predicting in vivo biological response.

Dose-Additive Carcinogenicity of a Defined Mixture of “Dioxin-like Compounds”

Use of the dioxin toxic equivalency factor (TEF) approach in human risk assessments assumes that the combined effects of dioxin-like compounds in a mixture can be predicted based on a potency-adjusted dose-additive combination of constituents of the mixture. In this study, we evaluated the TEF approach in experimental 2-year rodent cancer bioassays with female Harlan Sprague-Dawley rats receiving 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), 3,3′,4,4′,5-pentachlorobiphenyl (PCB-126), 2,3,4,7,8-pentachlorodibenzofuran (PeCDF), or a mixture of the three compounds. Statistically based dose–response modeling indicated that the shape of the dose–response curves for hepatic, lung, and oral mucosal neoplasms was the same in studies of the three individual chemicals and the mixture. In addition, the dose response for the mixture could be predicted from a combination of the potency-adjusted doses of the individual compounds. Finally, we showed that use of the current World Health Organization dioxin TEF values adequately predicted the increased incidence of liver tumors (hepatocellular adenoma and cholangiocarcinoma) induced by exposure to the mixture. These data support the use of the TEF approach for dioxin cancer risk assessments.

Characterization of the proneural gene regulatory network during mouse telencephalon development

BackgroundThe proneural proteins Mash1 and Ngn2 are key cell autonomous regulators of neurogenesis in the mammalian central nervous system, yet little is known about the molecular pathways regulated by these transcription factors.ResultsHere we identify the downstream effectors of proneural genes in the telencephalon using a genomic approach to analyze the transcriptome of mice that are either lacking or overexpressing proneural genes. Novel targets of Ngn2 and/or Mash1 were identified, such as members of the Notch and Wnt pathways, and proteins involved in adhesion and signal transduction. Next, we searched the non-coding sequence surrounding the predicted proneural downstream effector genes for evolutionarily conserved transcription factor binding sites associated with newly defined consensus binding sites for Ngn2 and Mash1. This allowed us to identify potential novel co-factors and co-regulators for proneural proteins, including Creb, Tcf/Lef, Pou-domain containing transcription factors, Sox9, and Mef2a. Finally, a gene regulatory network was delineated using a novel Bayesian-based algorithm that can incorporate information from diverse datasets.ConclusionTogether, these data shed light on the molecular pathways regulated by proneural genes and demonstrate that the integration of experimentation with bioinformatics can guide both hypothesis testing and hypothesis generation.

Key Characteristics of Carcinogens as a Basis for Organizing Data on Mechanisms of Carcinogenesis.

Background: A recent review by the International Agency for Research on Cancer (IARC) updated the assessments of the > 100 agents classified as Group 1, carcinogenic to humans (IARC Monographs Volume 100, parts A–F). This exercise was complicated by the absence of a broadly accepted, systematic method for evaluating mechanistic data to support conclusions regarding human hazard from exposure to carcinogens. Objectives and Methods: IARC therefore convened two workshops in which an international Working Group of experts identified 10 key characteristics, one or more of which are commonly exhibited by established human carcinogens. Discussion: These characteristics provide the basis for an objective approach to identifying and organizing results from pertinent mechanistic studies. The 10 characteristics are the abilities of an agent to 1) act as an electrophile either directly or after metabolic activation; 2) be genotoxic; 3) alter DNA repair or cause genomic instability; 4) induce epigenetic alterations; 5) induce oxidative stress; 6) induce chronic inflammation; 7) be immunosuppressive; 8) modulate receptor-mediated effects; 9) cause immortalization; and 10) alter cell proliferation, cell death, or nutrient supply. Conclusion: We describe the use of the 10 key characteristics to conduct a systematic literature search focused on relevant end points and construct a graphical representation of the identified mechanistic information. Next, we use benzene and polychlorinated biphenyls as examples to illustrate how this approach may work in practice. The approach described is similar in many respects to those currently being implemented by the U.S. EPA’s Integrated Risk Information System Program and the U.S. National Toxicology Program. Citation: Smith MT, Guyton KZ, Gibbons CF, Fritz JM, Portier CJ, Rusyn I, DeMarini DM, Caldwell JC, Kavlock RJ, Lambert P, Hecht SS, Bucher JR, Stewart BW, Baan R, Cogliano VJ, Straif K. 2016. Key characteristics of carcinogens as a basis for organizing data on mechanisms of carcinogenesis. Environ Health Perspect 124:713–721; http://dx.doi.org/10.1289/ehp.1509912

Variability of safe dose estimates when using complicated models of the carcinogenic process: A case study: Methylene chloride

Advances in understanding carcinogenesis have led to the development of mathematical models that have biologically interpretable parameters. These models utilize more of the available scientific data than the empirical models routinely employed for quantifying carcinogenic risk. They also require consideration of sources of uncertainty in risk estimates that were previously ignored, such as animal-to-animal variability of physiological and pharmacological constants. A numerical technique is proposed for studying the consequences of incorporating the intrapopulation variability of biologically interpretable parameters into the risk assessment process. To demonstrate the technique, the variability of safe dose estimates for exposure to methylene chloride is considered. The results suggest that intrapopulation variability of the model parameters can increase the variability of safe dose estimates an appreciable amount.

Evaluation of Biomonitoring Data from the CDC National Exposure Report in a Risk Assessment Context: Perspectives across Chemicals

Background: Biomonitoring data reported in the National Report on Human Exposure to Environmental Chemicals [NER; Centers for Disease Control and Prevention (2012)] provide information on the presence and concentrations of > 400 chemicals in human blood and urine. Biomonitoring Equivalents (BEs) and other risk assessment–based values now allow interpretation of these biomonitoring data in a public health risk context. Objectives: We compared the measured biomarker concentrations in the NER with BEs and similar risk assessment values to provide an across-chemical risk assessment perspective on the measured levels for approximately 130 analytes in the NER. Methods: We identified available risk assessment–based biomarker screening values, including BEs and Human Biomonitoring-I (HBM-I) values from the German Human Biomonitoring Commission. Geometric mean and 95th percentile population biomarker concentrations from the NER were compared to the available screening values to generate chemical-specific hazard quotients (HQs) or cancer risk estimates. Conclusions: Most analytes in the NER show HQ values of < 1; however, some (including acrylamide, dioxin-like chemicals, benzene, xylene, several metals, di-2(ethylhexyl)phthalate, and some legacy organochlorine pesticides) approach or exceed HQ values of 1 or cancer risks of > 1 × 10–4 at the geometric mean or 95th percentile, suggesting exposure levels may exceed published human health benchmarks. This analysis provides for the first time a means for examining population biomonitoring data for multiple environmental chemicals in the context of the risk assessments for those chemicals. The results of these comparisons can be used to focus more detailed chemical-specific examination of the data and inform priorities for chemical risk management and research.

Global Gene Expression Profiling of a Population Exposed to a Range of Benzene Levels

Background Benzene, an established cause of acute myeloid leukemia (AML), may also cause one or more lymphoid malignancies in humans. Previously, we identified genes and pathways associated with exposure to high (> 10 ppm) levels of benzene through transcriptomic analyses of blood cells from a small number of occupationally exposed workers. Objectives The goals of this study were to identify potential biomarkers of benzene exposure and/or early effects and to elucidate mechanisms relevant to risk of hematotoxicity, leukemia, and lymphoid malignancy in occupationally exposed individuals, many of whom were exposed to benzene levels < 1 ppm, the current U.S. occupational standard. Methods We analyzed global gene expression in the peripheral blood mononuclear cells of 125 workers exposed to benzene levels ranging from < 1 ppm to > 10 ppm. Study design and analysis with a mixed-effects model minimized potential confounding and experimental variability. Results We observed highly significant widespread perturbation of gene expression at all exposure levels. The AML pathway was among the pathways most significantly associated with benzene exposure. Immune response pathways were associated with most exposure levels, potentially providing biological plausibility for an association between lymphoma and benzene exposure. We identified a 16-gene expression signature associated with all levels of benzene exposure. Conclusions Our findings suggest that chronic benzene exposure, even at levels below the current U.S. occupational standard, perturbs many genes, biological processes, and pathways. These findings expand our understanding of the mechanisms by which benzene may induce hematotoxicity, leukemia, and lymphoma and reveal relevant potential biomarkers associated with a range of exposures.

Genetic and environmental pathways to complex diseases

BackgroundPathogenesis of complex diseases involves the integration of genetic and environmental factors over time, making it particularly difficult to tease apart relationships between phenotype, genotype, and environmental factors using traditional experimental approaches.ResultsUsing gene-centered databases, we have developed a network of complex diseases and environmental factors through the identification of key molecular pathways associated with both genetic and environmental contributions. Comparison with known chemical disease relationships and analysis of transcriptional regulation from gene expression datasets for several environmental factors and phenotypes clustered in a metabolic syndrome and neuropsychiatric subnetwork supports our network hypotheses. This analysis identifies natural and synthetic retinoids, antipsychotic medications, Omega 3 fatty acids, and pyrethroid pesticides as potential environmental modulators of metabolic syndrome phenotypes through PPAR and adipocytokine signaling and organophosphate pesticides as potential environmental modulators of neuropsychiatric phenotypes.ConclusionIdentification of key regulatory pathways that integrate genetic and environmental modulators define disease associated targets that will allow for efficient screening of large numbers of environmental factors, screening that could set priorities for further research and guide public health decisions.

A Comparison of Statistical Methods for Low Dose Extrapolation Utilizing Time-To-Tumor Data

The assessment of health risks due to low levels of exposure to potential environmental hazards based on the results of toxicological experiments necessarily involves extrapolation of results obtained at relatively high doses to the low dose region of interest. In this paper, different statistical extrapolation procedures which take into account both time-to-response and the presence of competing risks are compared using a large simulated data base. The study was designed to cover a range of plausible dose response models as well as to assess the effects of competing risks, background response, latency and experimental design on the performance of the different extrapolation procedures. It was found that point estimates of risk in the low dose region may differ from the actual risk by a factor of 1000 or more in certain situations, even when precise information on the time of occurrence of the particular lesion of interest is available. Although linearized upper confidence limits on risk can be highly conservative when the underlying dose response curve is sublinear in the low dose region, they were found not to exceed the actual risk in the low dose region by more than a factor of 10 in those cases where the underlying dose response curve was linear at low doses.