Information is gathered from animal studies and human data (epidemiological studies, case studies, accidental/intentional poisonings).
Information indicates if a chemical is associated with health effects like cancer, kidney toxicity, or birth defects.
Analysis uses a "weight of evidence" approach to determine toxicity consistently produced at lower doses.
The primary hazard of the chemical is identified (e.g., lead's primary hazard is neurotoxicity).
Dose-Response Assessment
Qualitative and quantitative toxicity information is collected.
It's a quantification of the hazard assessment.
Empirical data from animal studies usually provides information for quantitative risk assessment.
Animals are exposed to chemicals (individually or in combination) under controlled conditions to generate quantitative exposure-response data.
Quantitative dose-response relationships in humans are difficult to obtain, but human exposures provide the strongest evidence for causality between exposure and injury.
Lethality is the least desirable toxicological endpoint, but may sometimes be the only endpoint available.
For noncarcinogens, a threshold response is expected.
Extrapolating animal data to humans has limitations, especially when adjusting high experimental doses to those more representative of human exposures. This involves "adjusting" the experimental data to account for the use of a surrogate for humans and to address additional uncertainties concerning human safety.
For carcinogens, risk estimations differ; they are considered to have risk at any dose or concentration (no threshold dose).
A dose vs. response plot yields a straight line through the origin, indicating risk at any dose.
Exposure Assessment
Critical because there is no risk without exposure.
Example: A hazardous underground storage tank poses no risk unless it's leaking or there's a spill, and chemicals are identified in the soil, water, or air above background levels.
Determines who, what, when, where, and how exposure occurs.
What are the contaminants?
Nature of chemical (single or mixture).
Where are the contaminants located?
Complex, considers physicochemical properties and biological factors:
Partitioning between media (air-to-soil, water-to-soil, air-to-water).
Vapor pressures.
Bioconcentration.
Degradation.
Biotransformation.
What are the concentrations?
Requires adequate sampling and quantitative analysis.
Who is at risk?
Workers, residents, vulnerable groups.
How would an exposure result?
Air, soil, water.
When would exposure occur?
Limited to direct contact with soil, or every time water is consumed.
What are the pathways of exposure?
Oral, respiratory, dermal.
Environmental media (air, soil, breast milk).
Noncancer Risk Assessment
Acceptable Daily Intake (ADI): Amount of chemical an individual may be exposed to daily without apparent toxicity (used by FDA).
Minimal Risk Levels (MRL): Used by ATSDR.
Reference Dose (RfD): Used by EPA.
All express the NOAEL (No Observed Adverse Effect Level) or LOAEL (Lowest Observed Adverse Effect Level) divided by a value accounting for uncertainty (safety factor).
Uncertainty Factors
Modifying factors provide a safety element to compensate for the quality of published studies.
F_H: Human variability factor (10X).
F_A: Animal data extrapolation to humans (10X).
F_L: LOAEL use in place of the NOAEL (10X).
F_S: Subchronic data (10X).
F_D: Doubt as a modifying factor (0.1 – 10X), used if studies are limited or the risk assessor is doubtful about the study quality.
Level of modification is subjective.
ADI can be expressed as: ADI = \frac{NOAEL}{cumulative \, safety \, factors}
General formula for RfD: RfD = \frac{NOAEL \, or \, LOAEL}{(FH \times FA \times F_L)}
A value ≥ 1 indicates a risk of noncancer adverse health effects.
Indicates exposure is higher than acceptable under current public policy.
Does not guarantee illness but suggests an increased likelihood of adverse effects.
Thallium Levels and Bioaccumulation example
Thallium (Tl) is a highly toxic heavy metal.
Exposure from natural contamination and copper mining.
CDI_i = estimate of chronic daily intake from exposure assessment.
RfD_i = chronic reference dose for administered or absorbed dose.
A quotient under 1 is assumed as safe.
Cancer Risk Assessment
Post-WWII, views on cancer and its causes were reevaluated.
Link between radiation and cancer was found.
By the 1950s, carcinogens were known to produce toxicity through cellular mechanisms different from other toxicants.
Idea of “non-threshold” mechanisms emerged; any exposure increased cancer risk, and no safe limits could be established.
1958 Delaney Clause to the Food and Color Additive Amendments:
Forbade any chemical classified as a carcinogen from being added to the American food supply.
Similar regulatory positions were taken by the EPA through the Safe Drinking Water Act.
Cancer Risk Assessment Cont.
More chemicals were found to be carcinogenic and present in environmental media.
Regulators needed a way to assess cancer risk.
Mathematical models were used to draw inferences about cancer risk.
Most models assume a linear-dose response with a zero threshold dose.
Different cancer risk assessment models can yield different lifetime cancer risks.
Cancer Risk Assessment Models
Example: For the pesticide chlordane, the results of a lifetime risk of 1 \times 10^{-6} (one cancer death per 1 million individuals) from its consumption in drinking water were calculated using several models.
One-hit model (0.03 μg/l):
Assumes one molecule of a carcinogen can produce cancer.
Not widely accepted.
Inconsistent with the multistage hypothesis.
Multihit model (2 μg/l):
Consistent with the multistage process.
Assumes several mutagenic events are required.
Probit model (50 μg/l):
Infrequently used.
Assumes a log-normal distribution of susceptibility.
Linearized multistage model (0.07 μg/l):
Assumes linear extrapolation with a zero-dose threshold from the upper confidence level of the lowest dose that produced cancer.
Produces a cancer slope factor to predict cancer risk at a specific dose.
Used by the EPA and other state agencies.
Risk Characterization
Cancer and noncancer risks are estimated.
Uncertainty and modifying factors are addressed.
Information is combined and summarized to inform decision-making for managing the risk.
Quantitative exposure assessments are often difficult, so estimations are made based on exposure scenarios and assumptions about chemical intake through multiple routes.
Risk characterization becomes an estimation under different exposure scenarios.
Assumptions are made to address variations in human populations:
20 m3/day is the average inhalation rate of an adult.
2 l/day is the average water consumption rate of an adult.
70 kg is the average weight of an adult male.
Risk Management
Not part of the risk assessment process.
Application of risk characterization informs interaction between risk managers, risk assessors, the community, industry, academia, and government to produce rational decisions for public health concerns.
Rational management decisions must be made without a complete picture.
Overestimation of risk could lead to very costly and time-consuming decisions.
Underestimation of risk may leave communities or sensitive individuals vulnerable to excessive exposures.
Toxicant Interaction and Chemical Mixtures
Most toxicological studies focus on individual chemical agents (an unrealistic scenario but achievable in the lab).
Newer focus on chemical mixtures in toxicology and regulatory policy.
EPA defines a chemical mixture as either:
Complex: Too many components to estimate toxicity based on individual toxicities.
Simple: Contains few enough components that mixture toxicity can be characterized by combining individual toxicities and interactions.
Human health risk assessments generally consider multiple chemicals as simple mixtures, with combined toxicity resulting from the summation of individual toxicities derived from single chemical testing in lab animals.
Different chemicals may interact to produce effects in the body:
Independent effects.
Physiologically similar or additive effects.
Antagonistic effects.
Synergistic effects.
Potentiation effects.
Types of Toxicant Interactions
Independent effects: Chemicals produce dose-dependent effects through different physiological mechanisms (e.g., cadmium and organophosphate pesticide malathion).
Physiologically similar or additive effects: Chemicals interact to produce physiologically similar actions, resulting in combined toxicity.
Antagonistic effects: One chemical reduces the toxicity of another (important in developing antidotes).
Synergistic effects: Combination of two chemicals is more toxic than the sum of the separate toxicities.
Potentiation effects: A chemical with no toxic effect increases in toxicity when another chemical is present.