Interventional probability of causation (IPoC) with epidemiological and partial mechanistic evidence: benzene vs. formaldehyde and acute myeloid leukemia (AML).

INTRODUCTION: Causal epidemiology for regulatory risk analysis seeks to evaluate how removing or reducing exposures would change disease occurrence rates. We define interventional probability of causation (IPoC) as the change in probability of a disease (or other harm) occurring over a lifetime or other specified time interval that would be caused by a specified change in exposure, as predicted by a fully specified causal model. We define the closely related concept of causal assigned share (CAS) as the predicted fraction of disease risk that would be removed or prevented by a specified reduction in exposure, holding other variables fixed. Traditional approaches used to evaluate the preventable risk implications of epidemiological associations, including population attributable fraction (PAF) and the Bradford Hill considerations, cannot reveal whether removing a risk factor would reduce disease incidence. We argue that modern formal causal models coupled with causal artificial intelligence (CAI) and realistically partial and imperfect knowledge of underlying disease mechanisms, show great promise for determining and quantifying IPoC and CAS for exposures and diseases of practical interest. METHODS: We briefly review key CAI concepts and terms and then apply them to define IPoC and CAS. We present steps to quantify IPoC using a fully specified causal Bayesian network (BN) model. Useful bounds for quantitative IPoC and CAS calculations are derived for a two-stage clonal expansion (TSCE) model for carcinogenesis and illustrated by applying them to benzene and formaldehyde based on available epidemiological and partial mechanistic evidence. RESULTS: Causal BN models for benzene and risk of acute myeloid leukemia (AML) incorporating mechanistic, toxicological and epidemiological findings show that prolonged high-intensity exposure to benzene can increase risk of AML (IPoC of up to 7e-5, CAS of up to 54%). By contrast, no causal pathway leading from formaldehyde exposure to increased risk of AML was identified, consistent with much previous mechanistic, toxicological and epidemiological evidence; therefore, the IPoC and CAS for formaldehyde-induced AML are likely to be zero. CONCLUSION: We conclude that the IPoC approach can differentiate between likely and unlikely causal factors and can provide useful upper bounds for IPoC and CAS for some exposures and diseases of practical importance. For causal factors, IPoC can help to estimate the quantitative impacts on health risks of reducing exposures, even in situations where mechanistic evidence is realistically incomplete and individual-level exposure-response parameters are uncertain. This illustrates the strength that can be gained for causal inference by using causal models to generate testable hypotheses and then obtaining toxicological data to test the hypotheses implied by the models-and, where necessary, refine the models. This virtuous cycle provides additional insight into causal determinations that may not be available from weight-of-evidence considerations alone.

Underground salt and potash workers exposed to nitrogen oxides and diesel exhaust: assessment of specific effect biomarkers.

PURPOSE: Occupational exposure limits (OEL) for nitrogen oxides (NO, NO2) and diesel exhaust (EC-DPM) were reassessed by the German authorities in 2016/2017. We performed a clinical cross-sectional study among salt and potash underground workers exposed to these substances at relatively high levels to examine possible indicators of acute effects on workers' health. METHODS: We measured post- versus pre-shift differences in cardiovascular, inflammatory, immune, and respiratory effect biomarkers and assessed their associations with personal exposures measured during the same shift. We also compared post- versus pre-shift differences in biomarker levels between exposure groups defined based on work site and job type. RESULTS: None of the above-ground workers exceeded the OEL for NO2 and only 5% exceeded the OEL for EC-DPM exposure. Among underground workers, 33% of miners and 7% underground maintenance workers exceeded the OEL for NO2; the OEL for EC-DPM was exceeded by 56% of miners and 17% of maintenance workers. Some effect biomarkers (thrombocytes, neutrophils, MPO, TNF-α, IgE, FeNO) showed statistically significant differences between pre- versus post-shift measurements; however, there were no consistent associations between pre- and post-shift differences and exposure group or personal exposure measurements during the shift. CONCLUSIONS: We did not find evidence of associations between workplace exposure to NO, NO2 or EC-DPM and clinically relevant indicators of acute cardiovascular, inflammatory and immune, or respiratory effects among salt and potash underground workers in Germany.

Integration of Evidence on Community Cancer Risks from Elongate Mineral Particles in Silver Bay, Minnesota.

The potential for cancer-related risks to community members from ambient exposure to elongate mineral particles (EMPs) in taconite processing has not been formally evaluated. We evaluated 926 ambient air samples including 12,928 EMPs (particle structures with length-to-width ratio ≥3:1) collected over 26 years near a taconite processing facility in Silver Bay, Minnesota. Eighty-two percent of EMPs were ≤3 µm in length and 97% of EMPs had an average aspect ratio <20:1. A total of 935 (7.3%) EMPs had length >5 µm and AR ≥3:1. Average ambient concentration of NIOSH countable amphibole EMPs over all years was 0.000387 EMPs per cubic centimeter (EMP/cm3 ). Of 12,765 nonchrysotile EMPs, the number of amphiboles with length and width dimensions that correlate best with asbestos-related carcinogenicity ranged from four (0.03%) to 13 (0.1%) and the associated ambient amphibole air concentrations ranged from 0.000003 to 0.000007 EMP/cm3 . After 65 years of taconite processing in Silver Bay, evidence of an increased risk of mesothelioma and lung cancer in community members who did not work in the taconite industry is lacking. The absence of an increased risk of asbestos-related cancer in the Silver Bay community is coherent with supporting evidence from epidemiological and toxicological studies, as well as ambient exposure data and lake sediment data collected in Minnesota Iron Range communities. Collectively, the data provide consistent evidence that nonasbestiform amphibole minerals lack the carcinogenic potential exhibited by amphibole asbestos.

Extended Analysis and Evidence Integration of Chloroprene as a Human Carcinogen.

ß-Chloroprene is used in the production of polychloroprene, a synthetic rubber. In 2010, Environmental Protection Agency (EPA) published the Integrated Risk Information System "Toxicological Review of Chloroprene," concluding that chloroprene was "likely to be carcinogenic to humans." This was based on findings from a 1998 National Toxicology Program (NTP) study showing multiple tumors within and across animal species; results from occupational epidemiological studies; a proposed mutagenic mode of action; and structural similarities with 1,3-butadiene and vinyl chloride. Using mouse data from the NTP study and assuming a mutagenic mode of action, EPA calculated an inhalation unit risk (IUR) for chloroprene of 5 × 10-4 per µg/m3 . This is among the highest IURs for chemicals classified by IARC or EPA as known or probable human carcinogens and orders of magnitude higher than the IURs for carcinogens such as vinyl chloride, benzene, and 1,3-butadiene. Due to differences in pharmacokinetics, mice appear to be uniquely responsive to chloroprene exposure compared to other animals, including humans, which is consistent with the lack of evidence of carcinogenicity in robust occupational epidemiological studies. We evaluated and integrated all lines of evidence for chloroprene carcinogenicity to assess whether the 2010 EPA IUR could be scientifically substantiated. Due to clear interspecies differences in carcinogenic response to chloroprene, we applied a physiologically based pharmacokinetic model for chloroprene to calculate a species-specific internal dose (amount metabolized/gram of lung tissue) and derived an IUR that is over 100-fold lower than the 2010 EPA IUR. Therefore, we recommend that EPA's IUR be updated.

Occupational exposure to formaldehyde and risk of non hodgkin lymphoma: a meta-analysis.

BACKGROUND: Formaldehyde, a widely used chemical, is considered a human carcinogen. We report the results of a meta-analyses of studies on the relationship between occupational exposure to formaldehyde and risk of non-Hodgkin lymphoma (NHL). METHODS: We performed a systematic review and meta-analysis according to international guidelines and we identified 12 reports of occupational populations exposed to formaldehyde. We evaluated inter-study heterogeneity and we applied a random effects model. We conducted a cumulative meta-analysis and a meta-analysis according to estimated average exposure of each study population. RESULTS: The meta-analysis resulted in a summary relative risk (RR) for NHL of 0.93 (95% confidence interval 0.83-1.04). The cumulative meta-analysis suggests that higher RRs were detected in studies published before 1986, while studies available after 1986 did not show an association. No differences were found between different levels of occupational exposure. Conclusions Notwithstanding some limitations, the results of this meta-analysis do not support the hypothesis of an association between occupational exposure to formaldehyde and risk of NHL.

Considerations for refining the risk assessment process for formaldehyde: Results from an interdisciplinary workshop.

Anticipating the need to evaluate and integrate scientific evidence to inform new risk assessments or to update existing risk assessments, the Formaldehyde Panel of the American Chemistry Council (ACC), in collaboration with the University of North Carolina, convened a workshop: "Understanding Potential Human Health Cancer Risk - From Data Integration to Risk Evaluation" in October 2017. Twenty-four (24) invited-experts participated with expertise in epidemiology, toxicology, science integration and risk evaluation. Including members of the organizing committee, there were 29 participants. The meeting included eleven presentations encompassing an introduction and three sessions: (1) "integrating the formaldehyde science on nasal/nasopharyngeal carcinogenicity and potential for causality"; (2) "integrating the formaldehyde science on lymphohematopoietic cancer and potential for causality; and, (3) "formaldehyde research-data suitable for risk assessment". Here we describe key points from the presentations on epidemiology, toxicology and mechanistic studies that should inform decisions about the potential carcinogenicity of formaldehyde in humans and the discussions about approaches for structuring an integrated, comprehensive risk assessment for formaldehyde. We also note challenges expected when attempting to reconcile divergent results observed from research conducted within and across different scientific disciplines - especially toxicology and epidemiology - and in integrating diverse, multi-disciplinary mechanistic evidence.

Peak Exposures in Epidemiologic Studies and Cancer Risks: Considerations for Regulatory Risk Assessment.

We review approaches for characterizing "peak" exposures in epidemiologic studies and methods for incorporating peak exposure metrics in dose-response assessments that contribute to risk assessment. The focus was on potential etiologic relations between environmental chemical exposures and cancer risks. We searched the epidemiologic literature on environmental chemicals classified as carcinogens in which cancer risks were described in relation to "peak" exposures. These articles were evaluated to identify some of the challenges associated with defining and describing cancer risks in relation to peak exposures. We found that definitions of peak exposure varied considerably across studies. Of nine chemical agents included in our review of peak exposure, six had epidemiologic data used by the U.S. Environmental Protection Agency (US EPA) in dose-response assessments to derive inhalation unit risk values. These were benzene, formaldehyde, styrene, trichloroethylene, acrylonitrile, and ethylene oxide. All derived unit risks relied on cumulative exposure for dose-response estimation and none, to our knowledge, considered peak exposure metrics. This is not surprising, given the historical linear no-threshold default model (generally based on cumulative exposure) used in regulatory risk assessments. With newly proposed US EPA rule language, fuller consideration of alternative exposure and dose-response metrics will be supported. "Peak" exposure has not been consistently defined and rarely has been evaluated in epidemiologic studies of cancer risks. We recommend developing uniform definitions of "peak" exposure to facilitate fuller evaluation of dose response for environmental chemicals and cancer risks, especially where mechanistic understanding indicates that the dose response is unlikely linear and that short-term high-intensity exposures increase risk.

The epidemiologic evidence for elongate mineral particle (EMP)-related human cancer risk.

Epidemiologic research on the role of fibers and other elongate mineral particles (EMP) and human diseases including cancers has generated a large body of literature over the last decades: nevertheless, there remain some questions for which the scientific community appears unable to reach consensus. Reasons for genuine differences in opinion include (i) ways in which exposures have been classified; (ii) methodological limitations of the available studies, (iii) criteria for the interpretation of study results, including potential underlying biological mechanisms, and (iv) methodology for integrating the evidence. Various approaches have been proposed in recent years to address these issues, which will be illustrated using examples from asbestos, talc, taconite, synthetic mineral fibers and silicon carbide, with emphasis on potential carcinogenic effects. Potential solutions include improved exposure and outcome assessment - including use of biomarkers and other molecular approaches, consideration of potential confounding and other sources of bias, implementation of guidelines for study quality assessment and evidence evaluation and integration.

Six years after the NRC review of EPA's Draft IRIS Toxicological Review of Formaldehyde: Regulatory implications of new science in evaluating formaldehyde leukemogenicity.

Shortly after the International Agency for Research on Cancer (IARC) determined that formaldehyde causes leukemia, the United States Environmental Protection Agency (EPA) released its Draft IRIS Toxicological Review of Formaldehyde ("Draft IRIS Assessment"), also concluding that formaldehyde causes leukemia. Peer review of the Draft IRIS Assessment by a National Academy of Science committee noted that "causal determinations are not supported by the narrative provided in the draft" (NRC 2011). They offered recommendations for improving the Draft IRIS assessment and identified several important research gaps. Over the six years since the NRC peer review, significant new science has been published. We identify and summarize key recommendations made by NRC and map them to this new science, including extended analysis of epidemiological studies, updates of earlier occupational cohort studies, toxicological experiments using a sensitive mouse strain, mechanistic studies examining the role of exogenous versus endogenous formaldehyde in bone marrow, and several critical reviews. With few exceptions, new findings are consistently negative, and integration of all available evidence challenges the earlier conclusions that formaldehyde causes leukemia. Given formaldehyde's commercial importance, environmental ubiquity and endogenous production, accurate hazard classification and risk evaluation of whether exposure to formaldehyde from occupational, residential and consumer products causes leukemia are critical.

Does occupational exposure to formaldehyde cause hematotoxicity and leukemia-specific chromosome changes in cultured myeloid progenitor cells?

Several cross-sectional studies of a single population of workers exposed to formaldehyde at one of two factories using or producing formaldehyde-melamine resins in China have concluded that formaldehyde exposure induces damage to hematopoietic cells that originate in the bone marrow. Moreover, the investigators interpret observed differences between groups as evidence that formaldehyde induces myeloid leukemias, although the mechanisms for inducing these diseases are not obvious and recently published scientific findings do not support causation. Our objective was to evaluate hematological parameters and aneuploidy in relation to quantitative exposure measures of formaldehyde. We obtained the study data for the original study (Zhang et al. 2010 ) and performed linear regression analyses. Results showed that differences in white blood cell, granulocyte, platelet, and red blood cell counts are not exposure dependent. Among formaldehyde-exposed workers, no association was observed between individual average formaldehyde exposure estimates and frequency of aneuploidy, suggested by the original study authors to be indicators of myeloid leukemia risk.

Quantitative estimated exposure to vinyl chloride and risk of angiosarcoma of the liver and hepatocellular cancer in the US industry-wide vinyl chloride cohort: mortality update through 2013.

OBJECTIVE: To evaluate mortality risks of angiosarcoma of the liver (ASL), primary hepatocellular carcinoma (HCC) and other cancers among 9951 men employed between 1942 and 1972 at 35 US vinyl chloride (VC) or polyvinyl chloride plants followed for mortality through 31 December 2013. METHODS: SMR and time-dependent Cox proportional hazards analyses were used to evaluate mortality risks by cumulative VC exposure. RESULTS: Liver cancer mortality was elevated (SMR=2.87, 95% CI 2.40 to 3.40), and ASL and HCC were strongly associated with cumulative VC exposure ≥865 parts per million-years (ppm-years) (ASL: HR=36.3, 95% CI 13.1 to 100.5; and HCC: HR=5.3, 95% CI 1.6 to 17.7 for ≥2271 ppm-years). Excess deaths due to connective and soft tissue cancers (SMR=2.43, 95% CI 1.48 to 3.75), mesothelioma (SMR=2.29, 95% CI 1.18 to 4.00) and explosions (SMR=3.43, 95% CI 1.25 to 7.47) were seen. Mortalities due to melanoma, brain cancer, lung cancer and non-Hodgkin's lymphoma were not increased or associated with VC exposure. CONCLUSION: The association between VC and ASL first reported in this cohort 44 years ago persisted and was strongest among workers most highly exposed. VC exposure also was associated with HCC mortality, although it remains possible that misdiagnosis of early ASLs influenced findings.

The need for transparency and reproducibility in documenting values for regulatory decision making and evaluating causality: The example of formaldehyde.

Reproducibility and transparency in scientific reporting is paramount to advancing science and providing the foundation required for sound regulation. Recent examples demonstrate that pivotal scientific findings cannot be replicated, due to poor documentation or methodological bias, sparking debate across scientific and regulatory communities. However, there is general agreement that improvements in communicating and documenting research and risk assessment methods are needed. In the case of formaldehyde, the peer-review conducted by a National Academy of Sciences (NAS) Committee questioned the approaches used by the Integrated Risk Information System (IRIS) in developing draft unit risk values. Using the original data from the key study (Beane Freeman et al., 2009) and documentation provided in the draft IRIS profile, we attempted to duplicate the reported inhalation unit risk values and address the NAS Committee's questions regarding application of the appropriate dose-response model. Overall, documentation of the methods lacked sufficient detail to allow for replication of the unit risk estimates, specifically for Hodgkin lymphoma and leukemias, the key systemic endpoints selected by IRIS. The lack of apparent exposure-response relationships for selected endpoints raises the question whether quantitative analyses are appropriate for these endpoints, and if so, how results are to be interpreted.

Wind turbines and health: An examination of a proposed case definition.

Renewable energy demands have increased the need for new wind farms. In turn, concerns have been raised about potential adverse health effects on nearby residents. A case definition has been proposed to diagnose "Adverse Health Effects in the Environs of Industrial Wind Turbines" (AHE/IWT); initially in 2011 and then with an update in 2014. The authors invited commentary and in turn, we assessed its scientific merits by quantitatively evaluating its proposed application. We used binomial coefficients to quantitatively assess the potential of obtaining a diagnosis of AHE/IWT. We also reviewed the methodology and process of the development of the case definition by contrasting it with guidelines on case definition criteria of the USA Institute of Medicine. The case definition allows at least 3,264 and up to 400,000 possibilities for meeting second- and third-order criteria, once the limited first-order criteria are met. IOM guidelines for clinical case definitions were not followed. The case definition has virtually no specificity and lacks scientific support from peer-reviewed literature. If applied as proposed, its application will lead to substantial potential for false-positive assessments and missed diagnoses. Virtually any new illness that develops or any prevalent illness that worsens after the installation of wind turbines within 10 km of a residence could be considered AHE/IWT if the patient feels better away from home. The use of this case definition in the absence of a thorough medical evaluation with appropriate diagnostic studies poses risks to patients in that treatable disorders would be overlooked. The case definition has significant potential to mislead patients and its use cannot be recommended for application in any health-care or decision-making setting.

Triangulation of epidemiological evidence and risk of bias evaluation: A proposed framework and applied example using formaldehyde exposure and risk of myeloid leukemias.

Evidence triangulation may help identify the impact of study design elements on study findings and to tease out biased results when evaluating potential causal relationships; however, methods for triangulating epidemiologic evidence are evolving and have not been standardized. Building upon key principles of epidemiologic evidence triangulation and risk of bias assessment, and responding to the National Academies of Sciences, Engineering, and Medicine (NASEM) call for applied triangulation examples, the objective of this manuscript is to propose a triangulation framework and to apply it as an illustrative example to epidemiologic studies examining the possible relationship between occupational formaldehyde exposure and risk of myeloid leukemias (ML) including acute (AML) and chronic (CML) types. A nine-component triangulation framework for epidemiological evidence was developed incorporating study quality and ROB guidance from various federal health agencies (i.e., US EPA TSCA and NTP OHAT). Several components of the triangulation framework also drew from widely used epidemiological analytic tools such as stratified meta-analysis and sensitivity analysis. Regarding the applied example, fourteen studies were identified and assessed using the following primary study quality domains to explore potential key sources of bias: 1) study design and analysis; 2) study participation; 3) exposure assessment; 4) outcome assessment; and 5) potential confounding. Across studies, methodological limitations possibly contributing to biased results were observed within most domains. Interestingly, results from one study - often providing the largest and least-precise relative risk estimates, likely reflecting study biases, deviated from most primary study findings indicating no such associations. Triangulation of epidemiological evidence appears to be helpful in exploring inconsistent results for the identification of study results possibly reflecting various biases. Nonetheless, triangulation methodologies require additional development and application to real-world examples to enhance objectivity and reproducibility.

Systematic review of the association between talc and female reproductive tract cancers.

Talc is a hydrous magnesium sheet silicate used in cosmetic powders, ceramics, paints, rubber, and many other products. We conducted a systematic review of the potential carcinogenicity of genitally applied talc in humans. Our systematic review methods adhere to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and incorporated aspects from the US Institute of Medicine (IOM, now the National Academy of Medicine) and several US EPA frameworks for systematic reviews, evaluating and integrating the epidemiological, animal, and mechanistic literature on talc and cancer. We conducted a comprehensive literature search. Detailed data abstraction and study quality evaluation, adapting the Toxic Substances Control Act (TSCA) framework, were central to our analysis. The literature search and selection process identified 40 primary studies that assessed exposure to talc and female reproductive cancer risks in humans (n = 36) and animals (n = 4). The results of our evaluation emphasize the importance of considering biological plausibility and study quality in systematic review. Integrating all streams of evidence according to the IOM framework yielded classifications of suggestive evidence of no association between perineal application of talcum powders and risk of ovarian cancer at human-relevant exposure levels. We also concluded that there is suggestive evidence of no association between genital talc application and endometrial cancer, and insufficient evidence to determine whether a causal association exists between genital talc application and cervical cancer based on a smaller but largely null body of literature.

Critical review and synthesis of the epidemiologic evidence on formaldehyde exposure and risk of leukemia and other lymphohematopoietic malignancies.

PURPOSE: Recent epidemiologic studies indicate elevated risks for some lymphohematopoietic malignancies (LHM) related to formaldehyde exposure. We performed a systematic review of literature to assess the strength and consistency of associations. METHODS: We summarized published literature in the PubMed database of the National Library of Medicine during 1966-2012. Literature was categorized according to study design and population: industrial cohort studies, professional cohort studies, and population-based case-control studies. RESULTS: Findings from occupational cohort and population-based case-control studies were very inconsistent for LHM, including myeloid leukemia. Apart from some isolated exceptions, relative risks were close to the null, and there was little evidence for dose-response relations for any of the LHM. CONCLUSIONS: At present, there is no consistent or strong epidemiologic evidence that formaldehyde is causally related to any of the LHM. The absence of established toxicological mechanisms further weakens any arguments for causation. To be informative, future epidemiologic research should improve on formaldehyde exposure assessment and apply modern diagnostic schemes for specific LHM.

Lost in the woods: Finding our way back to the scientific method in systematic review.

Systematic review has become the preferred approach to addressing causality and informing regulatory and other decision-making processes, including chemical risk assessments. While advocates of systematic reviews acknowledge that they hold great potential for increasing objectivity and transparency in assessments of chemicals and human health risks, standardizing and harmonizing systematic review methods have been challenging. This review provides a brief summary of the development of systematic review methods and some of the frameworks currently in use in the US and Europe. We also provide an in-depth evaluation and comparison of two "competing" US EPA systematic review frameworks, informed by the constructively critical recommendations from the US National Academies of Science, Engineering and Medicine. We conclude with suggestions for moving forward to harmonize systematic review methods, as we believe that further criticism of individual available frameworks likely will be unproductive. Specifically, we issue a call to action for an international collaboration to work toward a blueprint that embraces the most scientifically critical elements common to most systematic review frameworks, while necessarily accommodating adaptations for specific purposes. Despite the array of available systematic review methods, it is clear that there is a shared goal and desire to promote objective assessment and synthesis of scientific evidence informing globally important issues regarding disease causality and human health risk evaluation.

Carcinogenicity of Poorly Soluble Low Toxicity Particles: Commentary on Epidemiology as a Risk Assessment "Reality Check".

Inhaled particles that are poorly soluble or insoluble and of low toxicity ("poorly soluble low toxicity" or "PSLT" particles), can accumulate in the lung and at lung overload levels induce lung cancers in rats. The question of whether PSLT particles increase lung cancer risk in humans is complicated by large differences between rats and humans and the relatively large particle doses administered in animal studies even when compared with heavy human occupational exposures. We review the findings of epidemiological studies on occupational exposure to each of three different PSLT particles (carbon black, talc and taconite). The epidemiological evidence indicates that at even very high occupational exposure levels at which non-malignant respiratory diseases including pneumoconiosis and even talcosis are observed, lung cancer risks appear not to be elevated. Although positive human cancer risks might be predicted based on extrapolation from overload doses in rats to relevant exposures in humans, the epidemiological "reality check" based on the three examples indicates that these PSLT particles are unlikely to increase lung cancer risk in humans even at high occupational levels of exposure. Therefore, we propose that careful evaluation of the epidemiological evidence can serve as a "reality check" for human risk assessment and help balance the risk evaluation process.