Probabilistic methods for non-cancer health effects.

Noncancer risk assessment methods and harmonization with cancer assessment methods have advanced from the simple divide a No Observed Adverse Effect Level (NOAEL) by a default safety factor or a linear extrapolation to background of the early 1980's. This advance is due in part due to groups such as the American Industrial Health Council, the National Institute of Environmental Health Sciences, the Society for Risk Analysis, the Society of Toxicology, and the U.S. Environmental Protection Agency, the National Academy of Sciences (NAS), the International Programme on Chemical Safety, and to many independent researchers outside of and within a workshop series sponsored by the Alliance for Risk Assessment prompted by the NAS \. Several of the case studies from this workshop series, and earlier work such as Bogdanffy et al., demonstrate that the dose response assessment of non-cancer toxicity and the harmonization of cancer and non-cancer methods are more than just a simple reflection of treating all non-cancer toxicity as if it has a threshold, or all cancer toxicity as if it did not. Moreover, one recommendation of NAS \ was to develop a problem formulation with risk managers prior to conducting any risk assessment. If the development of this problem formulation only necessitates the determination of a safe, or virtually safe dose, then the estimation of a Reference Dose (RfD) or virtually safe dose (VSD) or similar constructs should be encouraged. Not all of our environmental problems need a precise quantitative solution.

Range of the perfluorooctanoate (PFOA) safe dose for human health: An international collaboration.

Many government agencies and expert groups have estimated a dose-rate of perfluorooctanoate (PFOA) that would protect human health. Most of these evaluations are based on the same studies (whether of humans, laboratory animals, or both), and all note various uncertainties in our existing knowledge. Nonetheless, the values of these various, estimated, safe-doses vary widely, with some being more than 100,000 fold different. This sort of discrepancy invites scrutiny and explanation. Otherwise what is the lay public to make of this disparity? The Steering Committee of the Alliance for Risk Assessment (2022) called for scientists interested in attempting to understand and narrow these disparities. An advisory committee of nine scientists from four countries was selected from nominations received, and a subsequent invitation to scientists internationally led to the formation of three technical teams (for a total of 24 scientists from 8 countries). The teams reviewed relevant information and independently developed ranges for estimated PFOA safe doses. All three teams determined that the available epidemiologic information could not form a reliable basis for a PFOA safe dose-assessment in the absence of mechanistic data that are relevant for humans at serum concentrations seen in the general population. Based instead on dose-response data from five studies of PFOA-exposed laboratory animals, we estimated that PFOA dose-rates 10-70 ng/kg-day are protective of human health.

The Conundrum of the PFOA human half-life, an international collaboration.

The Steering Committee of the Alliance for Risk Assessment (ARA) opened a call for scientists interested in resolving what appeared to be a conundrum in estimating of the half-life of perfluorooctanoate (PFOA) in humans. An Advisory Committee was formed from nominations received and a subsequent invitation led to the development of three small independent working groups to review appropriate information and attempt a resolution. Initial findings were shared among these groups and a conclusion developed from the ensuing discussions. Many human observational studies have estimated the PFOA half-life. Most of these studies note the likely occurrence of unmonitored PFOA exposures, which could inflate values of the estimated PFOA half-life. Also, few of these studies estimated the half-life of PFOA isomers, the branched chains of which likely have shorter half-lives. This could deflate values of the estimated linear PFOA half-life. Fortunately, several studies informed both of these potential problems. The majority opinion of this international collaboration is that the studies striking the best balance in addressing some of these uncertainties indicate the likely central tendency of the human PFOA half-life is less than 2 years. The single best value appears to be the geometric mean (GM) of 1.3 years (Zhang et al., 2013, Table 3), based on a GM = 1.7 years in young females (n = 20) and GM = 1.2 years in males of all ages and older females (n = 66). However, a combined median value from Zhang et al. (2013) of 1.8 years also adds value to this range of central tendency. While the Collaboration found this study to be the least encumbered with unmonitored PFOA exposures and branched isomers, more studies of similar design would be valuable. Also valuable would be clarification around background exposures in other existing studies in case adjustments to half-life estimates are attempted.

The Future of Uncertainty Factors With In Vitro Studies Using Human Cells.

New approach methodologies (NAMs), including in vitro toxicology methods such as human cells from simple cell cultures to 3D and organ-on-a-chip models of human lung, intestine, liver, and other organs, are challenging the traditional "norm" of current regulatory risk assessments. Uncertainty Factors continue to be used by regulatory agencies to account for perceived deficits in toxicology data. With the expanded use of human cell NAMs, the question "Are uncertainty factors needed when human cells are used?" becomes a key topic in the development of 21st-century regulatory risk assessment. M.D., PhD, the coauthor of an article detailing uncertainty factors within the U.S. EPA, and L.E., PhD., Executive Vice President, Science, Emulate, who is involved in developing organ-on-a-chip models, debated the topic. One important outcome of the debate was that in the case of in vitro human cells on a chip, the interspecies (animal to human) uncertainty factor of 10 could be eliminated. However, in the case of the intraspecies (average human to sensitive human), the uncertainty factor of 10, additional toxicokinetic and/or toxicodynamic data or related information will be needed to reduce much less eliminate this factor. In the case of other currently used uncertainty factors, such as lowest observable adverse effect level to no-observed adverse effect level extrapolation, missing important toxicity studies, and acute/subchronic to chronic exposure extrapolation, additional data might be needed even when using in vitro human cells. Collaboration between traditional risk assessors with decades of experience with in vivo data and risk assessors working with modern technologies like organ chips is needed to find a way forward.

A regulatory relic: After 60 years of research on cancer risk, the Delaney Clause continues to keep us in the past.

The Delaney Clause of the Federal Food, Drug, and Cosmetic Act became law in 1958 because of concerns that potentially harmful chemicals were finding their way into foods and causing cancer. It states, "[n]o additive shall be deemed to be safe if it is found to induce cancer when ingested by man or animal, or if it is found, after tests which are appropriate for the evaluation of the safety of food additives, to induce cancer in man or animal." The United States Food and Drug Administration (US FDA) and United States Environmental Protection Agency (US EPA, prior to implementation of the Food Quality Protection Act) were charged with implementing this clause. Over 60 years, advances in cancer research have elucidated how chemicals induce cancer. Significant advancements in analytical methodologies have allowed for accurate and progressively lower detection limits, resulting in detection of trace amounts. Based on current scientific knowledge, there is a need to revisit the Delaney Clause's utility. The lack of scientific merit to the Delaney Clause was very apparent when recently the US FDA had to revoke the food additive approvals of 6 synthetic flavoring substances because high dose testing in animals demonstrated a carcinogenic response. However, US FDA determined that these 6 synthetic flavoring substances do not pose a risk to public health under the conditions of intended use. The 7th substance, styrene, was de-listed because it is no longer used by industry. The scientific community is committed to improving public health by promoting relevant science in risk assessment and regulatory decision making, and this was discussed in scientific sessions at the American Association for the Advancement of Science (AAAS) 2020 Annual Meeting and the Society of Toxicology (SOT) 2019 Annual Meeting. Expert presentations included advances in cancer research since the 1950s; the role of the Delaney Clause in the current regulatory paradigm with a focus on synthetic food additives; and the impact of the clause on scientific advances and regulatory decision making. The sessions concluded with panel discussions on making the clause more relevant based on 21st-century science.

The Dilemma of perfluorooctanoate (PFOA) human half-life.

Disparity in the results from human observational and clinical studies is not uncommon, but risk assessment efforts often judge one set of data more relevant with the loss of valuable information. The assessment for perfluorooctanoate (PFOA) is a good example of this problem. The estimation of its safe dose is disparate among government groups due in part to differences in understanding of its half-life in humans. These differences are due in part to incomplete information on sources of exposure in the human observational half-life studies, which have been routinely acknowledged, but until recently not well understood. Exposure information is thus critical in understanding, and possibly resolving, this disparity in PFOA safe dose, and potentially for disparities with similar chemistries when both human observational and clinical findings are available. We explore several hypotheses to explain this disparity in PFOA half-life from human observational studies in light of findings of a clinical study in humans and relevant exposure information from a recent international meeting of the Society of Toxicology and Environmental Chemistry (SETAC). Based on information from both human observational studies and clinical data, we proposed a range for the half-life for PFOA of 0.5-1.5 years, which would likely raise many existing regulatory safe levels if all other parameters stayed the same.

Analysis for data-derived extrapolation factors for procymidone.

The derivation of Chemical Specific Adjustment Factors (CSAFs) (IPCS, 2005; U.S. EPA, 2014) depends on the choice of appropriate dose metric. EPA and IPCS guidance was applied to derive a CSAF for developmental toxicity for procymidone (PCM). Although kinetic data were not available in humans at any dose, sufficient toxicokinetic data are available in a surrogate species, primates, and from chimeric mice with both rat and human liver cells to offer insights. Alternative approaches were explored in the derivation of the CSAG based on review of the available kinetic data. The most likely dosimetric adjustment is the Cmax based on the character of the critical effect - reduced anogenital distance and increased incidence of hypospadias in male rats, which likely occurs during a small window of time during development of the rat fetus. Cmax is also the default dosimeter from U.S. EPA (1991). However, in this case, the use of Cmax is also likely more conservative than the use of area under the curve (AUC), which otherwise is the default recommendation of the IPCS (2005). Despite human data, estimated tentative CSAF value is 0.48 (range, 0.22 to 0.74). The use of any of these values would be supported by the available data.

Suggestions for Improving the Characterization of Risk from Exposures to Per and Polyfluorinated Alkyl Substances.

Many state and federal environmental and health agencies have developed risk-based criteria for assessing the risk of adverse health effects of per- and polyfluorinated alkyl substances (PFAS) exposure to humans and the environment. However, the criteria that have been developed vary; drinking water criteria developed for perfluorooctanoic acid, for example, can vary by up to 750 fold. This is due to differences and variability in the data and information used, study/endpoint selection, assumptions and magnitude of uncertainty factors used in the absence and extrapolation of critical effect data, differences in underlying approaches to addressing exposure within criteria development, and/or policy decisions on levels of acceptable risk. We have critically evaluated the methods used to develop these criteria while focusing on derivation and application of drinking water criteria and discuss a range of improvements to risk-characterization practice recently presented at a Focused Topic Meeting on PFAS conducted by the Society of Environmental Toxicology and Chemistry in Durham, North Carolina, USA, 12 to 15 August 2019. We propose methods that consider maximizing the use of disparate data streams, seeking patterns, and proposing biologically based approaches to evidence integration toward informed criteria development. Environ Toxicol Chem 2021;40:883-898. © 2020 SETAC.

A commentary on some epidemiology data for chlorpyrifos.

Extensive animal and human studies on chlorpyrifos (CPF) point to changes in a blood enzyme as its first biological effect, and governments and health groups around the world have used this effect in the determination of its safe dose. Preventing this first biological effect, referred to in risk assessment parlance as the critical effect, is part and parcel of chemical regulation in general and of CFP specifically. Rauh et al. (2011), one of the published studies from the Columbia Center for Children's Environmental Health (CCCEH), reported evidence of deficits in Working Memory Index and Full-Scale IQ in children at 7 years old as a function of prenatal CPF exposures that are much lower than levels causing cholinesterase inhibition. Since the raw data on which Rauh et. al. (2011) publicly-funded (in part) findings were based have not been made available despite repeated requests, we show extracted data in Fig. 1A and 1E of Rauh et al. (2011), and plotted these extracted data as response versus log dose, a common risk assessment approach. Surprisingly, a significant portion of the data stated to be available in Rauh et al. (2011) were not found in these published figures, perhaps due to data point overlay. However, the reported associations of chlorpyrifos levels with Working Memory and Full Scale IQ were also not replicated in our analysis due perhaps to this missing data. Multiple requests were made to Rauh et al. (2011) for access to data from this, in part, publicly funded study, so that confirmation could be attempted. This general lack of data and inconsistency with cholinergic responses in other researches raises concerns about the lack of data transparency.

Bioelution, Bioavailability, and Toxicity of Cobalt Compounds Correlate.

Based on the wide use of cobalt substances in a range of important technologies, it has become important to predict the toxicological properties of new or lesser-studied substances as accurately as possible. We studied a group of 6 cobalt substances with inorganic ligands, which were tested for their bioaccessibility (surrogate measure of bioavailability) through in vitro bioelution in simulated gastric and intestinal fluids. Representatives of the group also underwent in vivo blood kinetics and mass balance tests, and both oral acute and repeated dose toxicity (RDT) testing. We were able to show a good correlation between high in vitro bioaccessibility with high in vivo bioavailability and subsequent high in vivo toxicity; consequently, low in vitro bioaccessibility correlated well with low in vivo bioavailability and low in vivo toxicity. In vitro bioelution in simulated gastric fluid was the most precise predictor of the difference in the oral RDT lowest observed adverse effect levels of 2 compounds representing the highly and poorly bioaccessible subset of substances. The 2 compounds cobalt dichloride hexahydrate and tricobalt tetraoxide differed by a factor of 440 in their in vitro bioaccessibility and by a factor of 310 in their RDT lowest observed adverse effect level. In summary, this set of studies shows that solubility, specifically in vitro bioelution in simulated gastric fluid, is a good, yet conservative, predictor of in vivo bioavailability and oral systemic toxicity of inorganic cobalt substances. Bioelution data are therefore an invaluable tool for grouping and read across of cobalt substances for hazard and risk assessment.

Data derived Extrapolation Factors for developmental toxicity: A preliminary research case study with perfluorooctanoate (PFOA).

Guidelines of the United States Environmental Protection Agency (EPA, 1991) and the International Programme on Chemical Safety (IPCS, 2005) suggest two different default positions for dosimetric extrapolation from experimental animals to humans when the dosimetry of the critical effect is not known. The default position of EPA (1991) for developmental toxicity is to use peak concentration (or Cmax) for this dosimetric extrapolation. In contrast, IPCS (2005, page 39) states its default position for dosimetric choice in the absence of data is to use the area under the curve (or AUC). The choice of the appropriate dose metric is important in the development of either a Chemical Specific Adjustment Factor (CSAF) of IPCS (2005) or a Data Derived Extrapolation Factor (DDEF) of EPA (2014). This research shows the derivation of a DDEF for developmental toxicity for perfluorooctanoate (PFOA), a chemical of current interest. Here, identification of the appropriate dosimetric adjustment from a review of developmental effects identified by EPA (2016) is attempted. Although some of these effects appear to be related to Cmax, most appear to be related to the average concentration or its AUC, but only during the critical period of development for a particular effect. A comparison was made of kinetic data from PFOA exposure in mice with newly available and carefully monitored kinetic data in humans after up to 36 weeks of PFOA exposure in a phase 1 clinical trial by Elcombe et al. (2013). Using the average concentration during the various exposure windows of concern, the DDEF for PFOA was determined to be 1.3 or 14. These values are significantly different than comparable extrapolations by several other authorities based on differences in PFOA half-life among species. Although current population exposures to PFOA are generally much lower than both the experimental animal data and the clinical human study, the development of these DDEFs is consistent with current guidelines of both EPA (2014) and IPCS (2005).

Let the IRIS Bloom:Regrowing the integrated risk information system (IRIS) of the U.S. Environmental Protection Agency.

The Integrated Risk Information System (IRIS) of the U.S. Environmental Protection Agency (EPA) has an important role in protecting public health. Originally it provided a single database listing official risk values equally valid for all Agency offices, and was an important tool for risk assessment communication across EPA. Started in 1986, IRIS achieved full standing in 1990 when it listed 500 risk values, the effort of two senior EPA groups over 5 years of monthly face-to-face meetings, to assess combined risk data from multiple Agency offices. Those groups were disbanded in 1995, and the lack of continuing face-to-face meetings meant that IRIS became no longer EPA's comprehensive database of risk values or their latest evaluations. As a remedy, a work group of the Agency's senior scientists should be re-established to evaluate new risks and to update older ones. Risk values to be reviewed would come from the same EPA offices now developing such information on their own. Still, this senior group would have the final authority on posting a risk value in IRIS, independently of individual EPA offices. This approach could also lay the groundwork for an all-government IRIS database, especially needed as more government Agencies, industries and non-governmental organizations are addressing evolving risk characterizations.

Benchmark dose (BMD) modeling: current practice, issues, and challenges.

Benchmark dose (BMD) modeling is now the state of the science for determining the point of departure for risk assessment. Key advantages include the fact that the modeling takes account of all of the data for a particular effect from a particular experiment, increased consistency, and better accounting for statistical uncertainties. Despite these strong advantages, disagreements remain as to several specific aspects of the modeling, including differences in the recommendations of the US Environmental Protection Agency (US EPA) and the European Food Safety Authority (EFSA). Differences exist in the choice of the benchmark response (BMR) for continuous data, the use of unrestricted models, and the mathematical models used; these can lead to differences in the final BMDL. It is important to take confidence in the model into account in choosing the BMDL, rather than simply choosing the lowest value. The field is moving in the direction of model averaging, which will avoid many of the challenges of choosing a single best model when the underlying biology does not suggest one, but additional research would be useful into methods of incorporating biological considerations into the weights used in the averaging. Additional research is also needed regarding the interplay between the BMR and the UF to ensure appropriate use for studies supporting a lower BMR than default values, such as for epidemiology data. Addressing these issues will aid in harmonizing methods and moving the field of risk assessment forward.

Derivation of a no-significant-risk-level for tetrabromobisphenol A based on a threshold non-mutagenic cancer mode of action.

A no-significant-risk-level of 20 mg day-1 was derived for tetrabromobisphenol A (TBBPA). Uterine tumors (adenomas, adenocarcinomas, and malignant mixed Müllerian) observed in female Wistar Han rats from a National Toxicology Program 2-year cancer bioassay were identified as the critical effect. Studies suggest that TBBPA is acting through a non-mutagenic mode of action. Thus, the most appropriate approach to derivation of a cancer risk value based on US Environmental Protection Agency guidelines is a threshold approach, akin to a cancer safe dose (RfDcancer ). Using the National Toxicology Program data, we utilized Benchmark dose software to derive a benchmark dose lower limit (BMDL10 ) as the point of departure (POD) of 103 mg kg-1  day-1 . The POD was adjusted to a human equivalent dose of 25.6 mg kg-1  day-1 using allometric scaling. We applied a composite adjustment factor of 100 to the POD to derive an RfDcancer of 0.26 mg kg-1  day-1 . Based on a human body weight of 70 kg, the RfDcancer was adjusted to a no-significant-risk-level of 20 mg day-1 . This was compared to other available non-cancer and cancer risk values, and aligns well with our understanding of the underlying biology based on the toxicology data. Overall, the weight of evidence from animal studies indicates that TBBPA has low toxicity and suggests that high doses over long exposure durations are needed to induce uterine tumor formation. Future research needs include a thorough and detailed vetting of the proposed adverse outcome pathway, including further support for key events leading to uterine tumor formation and a quantitative weight of evidence analysis.

Update: Mode of action (MOA) for liver tumors induced by oral exposure to 1,4-dioxane.

Previous work has shown that the weight of evidence supports the hypothesis that 1,4-dioxane causes liver tumors in rodents through cytotoxicity and subsequent regenerative hyperplasia. Questions regarding a lack of concordant findings for this mode of action (MOA) in mice have not been resolved, however. In the current work, a reanalysis of data from two chronic mouse cancer bioassays on 1,4-dioxane, one 13-week mouse study, seven rat cancer bioassays, coupled with other data such as 1,4-dioxane's negative mutagenicity, its lack of up-regulated DNA repair, and the appearance of liver tumors with a high background incidence, support the conclusion that rodent liver tumors, including those in mice, are evoked by a regenerative hyperplasia MOA. The initiating event for this MOA is metabolic saturation of 1,4-dioxane. Above metabolic saturation, higher doses of the parent compound cause an ever increasing toxicity in the rodent liver as evidenced by higher blood levels of enzymes indicative of liver cell damage and associated histopathology that occurs in a dose and time related manner. Importantly, alternative modes of action can be excluded. The observed liver toxicity has a threshold in the dose scale at or below levels that saturate metabolism, and generally in the range of 9.6-42 mg/kg-day for rats and 57 to 66 mg/kg-day for mice. It follows that threshold approaches to the assessment of this chemical's toxicity are supported by the non-mutagenic, metabolic saturation kinetics, and cytotoxicity-generated regenerative repair information available for 1,4-dioxane promoted rodent liver tumors.

Quantitative weight of evidence to assess confidence in potential modes of action.

The evolved World Health Organization/International Programme on Chemical Safety mode of action (MOA) framework provides a structure for evaluating evidence in pathways of causally linked key events (KE) leading to adverse health effects. Although employed globally, variability in use of the MOA framework has led to different interpretations of the sufficiency of evidence in support of hypothesized MOAs. A proof of concept extension of the MOA framework is proposed for scoring confidence in the supporting data to improve scientific justification for MOA use in characterizing hazards and selecting dose-response extrapolation methods for specific chemicals. This involves selecting hypothesized MOAs, and then, for each MOA, scoring the weight of evidence (WOE) in support of causality for each KE using evolved Bradford Hill causal considerations (biological plausibility, essentiality, dose-response concordance, consistency, and analogy). This early proof of concept method is demonstrated by comparing two potential MOAs (mutagenicity and peroxisome proliferator activated receptor-alpha) for clofibrate, a rodent liver carcinogen. Quantitative confidence scoring of hypothesized MOAs is shown to be useful in characterizing the likely operative MOA. To guide method refinement and future confidence scoring for a spectrum of MOAs, areas warranting further focus and lessons learned, including the need to incorporate a narrative discussion of the weights used in the evaluation and an overall evaluation of the plausibility of the outcome, are presented.