Use of a toxicokinetic model in the analysis of cancer mortality in relation to the estimated absorbed dose of dioxin (2,3,7,8-tetrachlorodibenzo-p-dioxin, TCDD)
Introduction
Dioxins are chlorinated aromatic hydrocarbons of which the 2,3,7,8-isomer (tetrachlorodibenzo-p-dioxin or TCDD) is known for its multisite animal carcinogenicity. The International Agency for Research on Cancer (IARC) reviewed the available evidence for carcinogenicity and assigned TCDD to group one (human carcinogen) (IARC, 1997). In its review, IARC considered the evidence from human epidemiological data to be limited. However, the multisite cancer excess observed in occupationally exposed populations was found to be consistent with the animal evidence, based on likely similarities of the underlying biological mechanism (interaction with the aryl hydrocarbon (Ah) receptor, which is present in both animals and humans) (IARC, 1997). Exposed workers in the chemical industry provide exposure levels comparable to those observed in animal experiments. It is, therefore, of interest to estimate cancer risk as a function of TCDD dose in these occupational cohorts. The largest of these cohorts was assembled by the US National Institute for Occupational Safety and Health (NIOSH) and consisted of 5172 workers from 12 chemical plants in the United States. Cancer mortality in this cohort was first analyzed by Fingerhut et al. (1991), the observed cancer mortality excess for all cancer and lung cancer was found to be related to duration of exposure (time spent in TCDD-exposed jobs). All cancer mortality in this cohort was later analyzed by quantitative exposure scores and the excess for all cancer appeared to be restricted to workers with the highest exposures, at levels 100–1000 times those observed in the general population (Steenland et al., 1999).
In this paper we describe an analysis of all cancer and lung cancer mortality in the NIOSH cohort in relation to the estimated absorbed dose of TCDD.
Analyses based on estimated serum levels have been conducted on other cohorts with occupational exposure to TCDD (Ott and Zober, 1996, Becher et al., 1998, Hooiveld et al., 1998). In these analyses, past TCDD exposure levels were estimated based on statistical models of TCDD kinetics. These models rely on statistical adjustments via covariate control to account for deviation of TCDD kinetics from a fixed half-life (first-order) time behavior. In the analysis presented in this paper, we adopted a minimal physiological toxicokinetic model (MPTK) to estimate TCDD dose for individual members of the cohort. The model addresses long-term behavior of TCDD in humans and it was first proposed by Dankovic et al. (1995) and further analyzed by Thomaseth and Salvan, 1998, Salvan et al., 1999 and Bortot et al., 1999, Bortot et al., 2000. The model accounts for variations in lipid volume over time and how they affect the predicted lipid-adjusted serum concentration of TCDD. No covariate control is necessary to account for body mass and its changes or age. Unlike statistical models of TCDD kinetics (e.g. Flesch-Janys et al., 1996, Michalek et al., 1996), its parameters have a direct physiological meaning, however, without the complexity of mechanistic models for TCDD (e.g. Kohn et al., 1993), which typically require animal data as a source of parameter values.
In the following sections, we summarize our work on the MPTK model for TCDD and describe a mortality analysis for all cancer and lung cancer in the NIOSH cohort based on TCDD dose estimates obtained from the kinetic model.
Section snippets
The TCDD model
A detailed account and analysis of the MPTK model for TCDD first proposed by Dankovic et al. (1995) is reported elsewhere (Thomaseth and Salvan, 1998, Salvan et al., 1999, Bortot et al., 1999, Bortot et al., 2000). In the model, TCDD concentration is assumed to be in a dynamic equilibrium between three compartments of body lipids: blood; liver; and adipose tissue. Basic model assumptions are: (1) all TCDD elimination occurs in the liver with a fractional clearance rate which is proportional to
TCDD kinetic model
Parameter estimates for the TCDD kinetic model obtained under different methods are reported in Table 1. All estimation methods produced similar point and interval estimates of the liver-elimination parameter. A higher uncertainty seemed to be associated with the estimation of background input, probably because of the low amount of information on background levels contained in the Ranch Hand data. This difficulty was overcome in the Bayesian analysis (Bortot et al., 2000) which used an
Discussion
This mortality analysis of occupational cancer in relation to estimated absorbed dose of TCDD was based on dose estimates obtained from a minimal physiologic toxicokinetic model for TCDD. A special feature of the MPTK model is its ability to account for intra- and inter-individual variations of BMI over time and how they affect TCDD kinetics. This approach is a departure from that followed by others (e.g. Michalek et al., 1996, Flesch-Janys et al., 1996). The model is, however, built upon a
Conclusions
We presented a cancer mortality analysis in relation to estimated absorbed dose of TCDD. This study provided an example of interfacing an exposure model with a risk estimation model. Several limitations may have affected our analysis at different stages, as discussed above. However, our results appear to be generally consistent with those of other analyses of the same and of similar cohorts. Specifically, a sizable elevation (doubling) in risk for all cancer is obtained at TCDD doses that are
Acknowledgements
This work was partially supported by the National Institute for Occupational Safety and Health, Cincinnati, USA and by a grant from the Italian Ministero del Lavoro e della Previdenza Sociale. We thank J. Michalek for providing data from the Vietnam Veterans Health Study and K. Steenland, D. Dankovic and L. Stayner for comments and discussions at several stages of this project.
References (31)
- et al.
Modeling of the toxicokinetics of polychlorinated dibenzo-p-dioxins and dibenzofurans in mammalians, including humans. II. Kinetics of absorption and disposition of PCDDs/PCDFs
Toxicol Appl Pharmacol
(1995) - et al.
A mechanistic model of effects of dioxin on gene expression in the rat liver
Toxicol Appl Pharmacol
(1993) - et al.
Serum levels of PCDDs and PCDFs among workers exposed to 2,3,7,8-TCDD contaminated chemicals
Chemosphere
(1992) - et al.
The estimation of elimination rates of persistent compounds: a re-analysis of 2,3,7,8-tetrachlorodibenzo-p-dioxin levels in Vietnam veterans
Chemosphere
(1998) - et al.
Quantitative cancer risk assessment for dioxins using an occupational cohort
Environ Health Perspect
(1998) - Bortot P, Thomaseth K, Salvan A. Population analysis of a toxicokinetic model for 2,3,7,8-tetrachlorodibenzo-p-dioxin...
- Bortot P, Thomaseth K, Salvan A. Population toxicokinetic analysis of 2,3,7,8–tetrachlorodibenzo-p-dioxin using...
- et al.
A simplified model describing the kinetics of TCDD in humans [abstract]
Toxicologist
(1995) - et al.
BMI as a measure of body fatness: age and sex-specific prediction formulas
Br J Nutr
(1991) - et al.
Cancer mortality in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin
N Engl J Med
(1991)
Elimination of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/F) in occupationally exposed persons
J Toxicol Environ Health
Second follow-up of a Dutch cohort occupationally exposed to phenoxy herbicides, chlorophenols, and contaminants
Am J Epidemiol
International commission on radiation protection. The reference man
Height, weight, percent body fat, and indices of adiposity in young men and women entering the US army
Aviat Space Environ Med
Cited by (0)
- 1
Present address: Department of Statistics, University of Bologna, Via Belle Arti 41, Bologna 40126, Italy.