An equivalency iterative algorithm for cancer risk assessment of chemical mixtures with additive effects

A dynamic inventory database for assessing age-, gender-, and route-specific chronic internal exposure to chemicals in support of human exposome research

In this study, we proposed a dynamic inventory database to evaluate chronic internal exposure to chemicals at a population level, which enables users to perform modeling exercises specific to a particular chemical, route of exposure, age group, and gender. The database was built based on the steady-state solution of physiologically based kinetic (PBK) models. The biotransfer factors [BTF, the steady-state ratio between the chemical concentration in human tissues and the average daily dose (ADD) of the chemical] of 931 organic chemicals in major organs and tissues were simulated for a total of 14 population age groups for males and females. The results indicated that infants and children had the highest simulated BTFs of chemicals, and middle-aged adults had the lowest simulated BTFs. The route-specific analysis of the simulated BTFs indicated that the biotransformation half-life and octanol-water partition coefficient of chemicals had a profound impact on the BTFs. Organ- and chemical-specific results indicated that the biotransfer potential of chemicals in human bodies was primarily determined by bio-thermodynamic variables (e.g., lipid contents). In conclusion, the proposed inventory database can be conveniently used to access chronic internal exposure doses of chemicals by multiplying the route-specific ADD values for different population groups. In future studies, we recommend incorporating human biotransformation data, partition coefficients of ionizable chemicals, age-specific vulnerable indicators (e.g., the degree of maturation of immune systems), physiological variations within the same age group (e.g., intensity of daily physical activities), growth rates (i.e., the dilution effect on chemical biotransfer), and all possible target organs of carcinogenicity (e.g., bladder) into the proposed dynamic inventory database to help promote human exposome research.

A disease-specific screening-level modeling approach for assessing the cancer risks of pesticide mixtures

As humans are always exposed to multiple pesticides, it is necessary to conduct risk assessments for pesticide mixtures. Due to data limitations, in this study, we introduced a disease-specific screening-level modeling framework to simulate the cumulative cancer risk (CR) of carcinogenic pesticides, which was developed based on the lognormal dose-response (LDR) curve of chemicals with disease-specific modes of action (MOAs). The simulated results of a case study indicate that the cumulative CR can be at least two orders of magnitude higher than the simulated CRs of individual pesticides. The comparison between the LDR model and the linear extrapolation (or cancer slope factor, CSF) model indicates that the CSF model can greatly overestimate population cancer risks. In addition, we applied our model to evaluate current regulatory standards of carcinogenic pesticide mixtures, and the results indicate that current standards for the selected jurisdictions can control the cumulative cancer risks within the acceptable level. However, the CSF model suggests that all selected jurisdictions cannot protect population health against the carcinogenic pesticide mixture, which is due to the nature of the low-dose linear extrapolation that triggers an initial slope when the effect dose is close to zero. Thus, we concluded that although the MOAs of pesticides in human bodies must be evaluated in future studies, our disease-specific model can be a useful and practical tool for cancer risk assessment and regulatory management of pesticide mixtures.

Improving screening model of pesticide risk assessment in surface soils: Considering degradation metabolites

As pesticides can be degraded to toxic metabolites in the soil, metabolite toxicity should be considered in human health risk assessments. In this study, a screening-level modeling framework was developed to manage pesticides in surface soil, which was discussed under discrete and continuous emission scenarios. In addition, we selected glyphosate and its major metabolite (aminomethylphosphonic acid or AMPA) as examples to conduct screening-level risk management at regional, national, and global scales. The results indicated that if soil AMPA were not considered, human health risks could be significantly underestimated because of the large half-life of AMPA in the soil. For example, the added concentration factors of AMPA were simulated as 0.19 and 6.72 considering all major elimination pathways and considering the degradation pathway alone, respectively, indicating that AMPA formation could lead to severe extra health burdens. Furthermore, the evaluation of current glyphosate soil standards suggested that toxic metabolites should be considered in the regulatory process; otherwise, many standards could theoretically trigger high levels of soil AMPA, which could result in serious human health damage. Our proposed screening-level model can help to improve risk assessment and regulatory management of pesticides in surface soils.

Backward modeling of urinary test reliability for assessing PAH health risks: An approximation solution for naphthalene

Urine sample tests are one of the most common methods of estimating human exposure to polycyclic aromatic hydrocarbons (PAHs) and assessing population health risks. To evaluate the reliability of the urine test and the impact of other PAH elimination routes on the health risk estimated by this test, we proposed a backward modeling framework integrating other common elimination routes of PAH metabolites to calculate the overall intake rate of the parent PAH based on the levels of corresponding main metabolites in urine. Due to limited biotransformation data, we selected naphthalene as an example to evaluate model performance and collected urine samples from 234 random adults in Shenzhen. The overall intake rates of naphthalene were then simulated and compared to current literature data. The simulated intake rates of naphthalene ranged from 3.70 × 10-3 mg d-1 to 1.95 mg d-1 and followed a lognormal distribution with a median value of 6.51 × 10-2 mg d-1. The results indicated that, if naphthalene exposure occurred only via food for the population of Shenzhen, the literature data fell within the most frequent interval [3.70 × 10-3, 4.45 × 10-2] but were lower than the simulated median value. However, if other exposure routes were considered, the allocation factor-adjusted literature data were close to the simulated median values. In addition, under normal physiological conditions, the simulated results were more sensitive to 1-hydroxynaphthalene (1-OHN) and 2-hydroxynaphthalene (2-OHN) levels in urine than other biometric variables, which is due to the limited load of 1-OHN and 2-OHN in human elimination routes. Furthermore, the suggested safety levels of 1-OHN and 2-OHN in urine to protect 99% of the general population of Shenzhen were 6.40 × 10-6 and 3.75 × 10-5 mg L-1, which could be used as regulatory indicators based on the high reliability of the model.