9.5. DISPOSITION AND PHARMACOKINETICS

The disposition and pharmacokinetics of 2,3,7,8-TCDD and related compounds have been investigated in several species and under various exposure conditions. These data and models derived from them are critical in understanding the sequelae of human exposure. Data related to disposition and pharmacokinetics of dioxin and related compounds and efforts to develop models to further understand tissue dosimetry are described in detail in Chapter 1 of the Health Assessment Document.

The gastrointestinal, dermal, and transpulmonary absorptions of these compounds represent potential routes for human uptake. Findings of studies in experimental animals indicate that oral exposure to 2,3,7,8-TCDD in the diet or in an oil vehicle results in the absorption of >50%, and often closer to 90%, of the administered dose. Gastrointestinal absorption of related compounds is variable, incomplete, and congener specific. More soluble congeners, such as 2,3,7,8-TCDF, are almost completely absorbed, while the extremely insoluble OCDD is very poorly absorbed. In some cases, absorption has been found to be dose dependent, with increased absorption occurring at lower doses (2,3,7,8-TBDD, OCDD). The limited data base also suggests that there are no major interspecies differences in the gastrointestinal absorption of these compounds among mammals. Limited data (Poiger and Schlatter, 1986) from a single human volunteer suggest a high level (>87%) of absorption of 2,3,7,8-TCDD in corn oil from the gastrointestinal tract. Following absorption, a half-life for elimination was estimated to be 2,120 days (5.8 years). It should be noted that this estimate of half-life is for a single individual and that longer median half-lives for 2,3,7,8-TCDD have been estimated (7.1 and 11.3 years) in other studies described in this chapter and in Chapter 1.

Additional data also indicate the importance of the formulation or vehicle containing the toxicant(s) on the relative bioavailability of 2,3,7,8-TCDD and related compounds after exposure. For instance, rodent feeding studies indicate that the bioavailability of 2,3,7,8-TCDD from soil varies between sites and 2,3,7,8-TCDD content alone may not be indicative of potential human hazard from contaminated environmental materials. Although data indicate that substantial absorption may occur from contaminated soil, soil type and duration of contact may substantially affect the absorption of 2,3,7,8-TCDD from soils obtained from different contaminated sites. This uncertainty should be kept in mind as intake values and the assumption of 50-100% absorption are often used to estimate potential risk from environmental samples.

In experiments measuring dermal absorption for 2,3,7,8-TCDD and several CDFs, the percentage of administered dose absorbed decreased with increasing dose while the amount absorbed increased with dose. Results also suggest that the majority of the compound remaining at the skin exposure site was associated with the outer skin layer (the stratum corneum) and did not penetrate through to the dermis. Together, these results on dermal absorption indicate that at £ 0.1 m mol/kg, a greater percent of this administered dose of 2,3,7,8-TCDD and three CDFs was absorbed. Nonetheless, even following a low-dose dermal application of 200 pmol (1 nmol/kg), the rate of absorption of 2,3,7,8-TCDD is still very slow (rate constant of 0.005 hour^-1). Dermal exposure of humans to 2,3,7,8-TCDD and related compounds usually occurs as a complex mixture of these contaminants in soil, oils, or other mixtures that would be expected to alter absorption. Available data suggest that the dermal absorption of 2,3,7,8-TCDD depends on the formulation (vehicle or adsorbent) containing the toxicant. Although no data are available to directly evaluate human dermal absorption, the data available from in vitro and animal studies suggest slow dermal absorption of these compounds, which is likely to be dependent on the vehicle or adsorbent containing the compounds and the duration of the contact.

The use of incineration as a means of solid and hazardous waste management results in the emission of vapors and contaminated particles that may contain TCDD and related compounds into the environment. Thus, exposure to TCDD and related compounds may result from inhalation of contaminated fly ash, dust, and soil or from ingestion if air-transported particles are deposited on fruits and vegetables. Direct exposure by the inhalation route is usually relatively low as a percentage of overall intake. Systemic effects occur in animals after pulmonary exposure to TCDD, suggesting that transpulmonary absorption of TCDD does occur. Further results suggest that the transpulmonary absorption of 2,3,7,8-TCDD and 2,3,7,8-TBDD was similar to that observed following oral exposure. These limited data provide evidence of efficient transpulmonary absorption after intratracheal instillation in laboratory animals. No data from humans or primates are available to address this issue. However, these data provide support for the inference that efficient absorption will occur when vapors and particles containing dioxin and related compounds are inhaled by humans.

Once absorbed into blood, 2,3,7,8-TCDD and related compounds readily distribute to all organs. Tissue distribution within the first hour after exposure reflects physiological parameters such as blood flow to a given tissue and relative tissue size. There do not appear to be major species or strain differences in the tissue distribution of 2,3,7,8-TCDD and 2,3,7,8-TCDF in mammals, with the liver and adipose tissue being the primary disposition sites although human data to address this issue are quite limited. The tissue distribution of the coplanar PCBs and PBBs also appears to be similar to that of 2,3,7,8-TCDD and 2,3,7,8-TCDF based on evaluation in experimental animals.

Multiple studies suggest that distribution of this class of compounds to internal organs is dose dependent. At low doses in animal studies, adipose tissue serves as the major depot; at high doses, a major fraction is sequestered in the liver. The biochemical basis for this observation is under investigation. Induction of a hepatic binding protein has been hypothesized to play a major role.

As discussed above, levels of 2,3,7,8-TCDD averaging 5-10 pg/g lipid (ppt) have been reported for background populations. Sielken (1987) evaluated these data and concluded that the levels of 2,3,7,8-TCDD in human adipose are log-normally distributed and positively correlated with age. Among the observed U.S. background levels of 2,3,7,8-TCDD in human adipose tissue, more than 10% were >12 pg/g (ppt).

Human body-burden measurements on dioxins were initially conducted using adipose tissue, which required surgical samples, or, occasionally for women, using breast milk. Patterson et al. (1988) showed that human serum was an accurate and more practical surrogate for human adipose tissue. They found that the partitioning ratio of 2,3,7,8-TCDD between adipose tissue and serum was approximately 1.09 when the concentrations were adjusted for lipid content. This relationship appears to hold for at least a thousandfold concentration range in excess of background levels. This correlation indicates that serum 2,3,7,8-TCDD, coupled with measurement of serum lipid content, provides a valid estimate of the 2,3,7,8-TCDD concentration in adipose tissue under steady-state, low-dose conditions.

In a study of potentially heavily exposed Vietnam veterans, the Centers for Disease Control and Prevention (MMWR, 1988) reported an Air Force study of Ranch Hand veterans who were either herbicide loaders or herbicide specialists in Vietnam. The herbicide 2,4,5-T (Agent Orange) that was used in Vietnam was contaminated with a low percentage of 2,3,7,8-TCDD. The mean serum 2,3,7,8-TCDD level of 147 Ranch Hand personnel was 49 pg/g (ppt) in 1987, based on total lipid-weight, while the mean serum level of the 49 controls was 5 pg/g (ppt). In addition, 79% of the Ranch Hand personnel and 2% of the controls had 2,3,7,8-TCDD levels ³ 10 pg/g (ppt). The distribution of 2,3,7,8-TCDD levels in this phase of the Air Force health study indicates that Ranch Hand veterans have had higher lifetime exposures than controls and that a small number of Ranch Hand personnel had unusually heavy 2,3,7,8-TCDD exposure. Pirkle et al. (1989) estimated the median half-life of 2,3,7,8-TCDD in humans to be approximately 7 years on the basis of 2,3,7,8-TCDD levels in serum samples taken in 1982 and 1987 from 36 of the Ranch Hand personnel who had 2,3,7,8-TCDD levels >10 pg/g (ppt) in 1987. Similar tissue concentrations were obtained by Kahn et al. (1988) in a report comparing 2,3,7,8-TCDD levels in blood and adipose tissue of moderately exposed Vietnam veterans who handled herbicides regularly while in Vietnam and matched controls. Although this study can distinguish moderately exposed men from others, the data do not address the question of the difficulty of characterizing the exposures of persons whose exposures are relatively low and who constitute the bulk of the population, both military and civilian, who may have been exposed to greater than background levels of 2,3,7,8-TCDD. Despite the fact that their exposures may result in slightly elevated levels of 2,3,7,8-TCDD, these individuals are indistinguishable from the general population with similar blood levels spanning a range of nondetect to >10 ppt. Recently, a follow-up analysis to the Ranch Hand study described above has been published. This study (Wolfe et al., 1994) describes half-life measurements based on 337 Ranch Hand veterans. The estimate of the median half-life of TCDD is predicted to be 11.3 years. The implications of this longer half-life on our understanding of TCDD kinetics and on the back-calculations of historic intake values and body burdens will need to be fully described in future versions of this report.

The metabolism of 2,3,7,8-TCDD and related compounds is required for urinary and biliary elimination and therefore plays a major role in regulating the rate of excretion of these compounds and determining their half-life. Although early in vivo and in vitro investigations were unable to detect the metabolism of 2,3,7,8-TCDD, there is now evidence that a wide range of mammalian and aquatic species are capable of slowly biotransforming 2,3,7,8-TCDD to polar metabolites. Although metabolites of 2,3,7,8-TCDD have not been directly identified in humans, recent analytic data from feces samples from an individual in a self-dosing experiment suggest that humans can slowly metabolize 2,3,7,8-TCDD (Wendling and Orth, 1990). Direct intestinal excretion of the parent compound is another route for excretion of 2,3,7,8-TCDD and related compounds that is not regulated by metabolism.

Some investigators have questioned whether the parent compound or metabolites are responsible for dioxin toxicity. Structure-activity studies of 2,3,7,8-TCDD and related compounds support the widely accepted principle that the parent compound is the active species, and the relative lack of biological activity of readily excreted monohydroxylated metabolites of 2,3,7,8-TCDD and 3,3¢ ,4,4¢ -TCB suggests that metabolism is a detoxification process necessary for the biliary and urinary excretion of these compounds. This concept has also been generally applied to 2,3,7,8-TCDD-related compounds, although data are lacking on the structure and toxicity of metabolites of other CDDs, BDDs, CDFs, BDFs, PCBs, and PBBs. It is still possible, however quite unlikely, that low levels of unextractable and/or unidentified metabolites may contribute to one or more of the toxic responses of 2,3,7,8-TCDD and related compounds.

Physiologically based pharmacokinetic (PB-PK) models have been developed for 2,3,7,8-TCDD in mice, rats, and humans. PB-PK models incorporate known or estimated anatomical, physiological, and physicochemical parameters to describe quantitatively the disposition of a chemical in a given species. PB-PK models can assist in the extrapolation of high-to-low dose kinetics within a species, estimating exposures by different routes of administration, calculating effective doses, and extrapolating these values across species. These models are particularly important given the limited empirical data on individual dioxin-like congeners.

Chapter 8 contains a review of biologically based models of dioxin pharmacokinetics. The early studies in rodents have recently been extended to describe protein induction and tissue distribution data in the mouse (Leung et al., 1990b) and rat (Leung et al., 1990a). Andersen et al. (1993) refined the model to include induction of CYP1A1 and diffusion-limited tissue distribution. CYP1A1 is one of a family of proteins involved in the activation and detoxification of both endogenous and exogenous chemicals. The model described by Kedderis et al. (1993) for 2,3,7,8-tetrabromodibenzo-p-dioxin extended the use of PBPK models to the brominated congener of TCDD. Portier et al. (1993) modeled the steady-state induction of CYP1A1 and CYP1A2 using Hill equations. Their analysis stressed the importance of the mechanism of endogenous protein expression on the shape of the dose-response curve in the low-dose region. Kohn and Portier (1993) extended this result to a general class of models and discussed implications of these models for risk assessment. Kohn et al. (1993) used approaches to describe tissue dosimetry of TCDD and additionally incorporated dioxin-mediated effects on growth factors, induction of the Ah-receptor, and several models for endogenous induction of CYP1A1, CYP1A2, and the EGF receptor. Other models have been proposed recently to describe effects of TCDD on lipid metabolism (Roth et al., 1993).

An empirical dose-dependent model by Carrier (1991) relates the varying fraction of the body burden of TCDD associated with the liver in humans to the total body burden of TCDD. This model is consistent with the animal results described by the PB-PK models of Andersen et al. (1993) and Kohn et al. (1993).

Our uncertainty in the validity of predictions from PB-PK models is primarily driven by the limited availability of congener and species-specific data that accurately describe the dose- and time-dependent disposition of 2,3,7,8-TCDD and related compounds. As additional data become available, particularly on the dose-dependent disposition of these compounds, more accurate models can be developed. In developing a suitable model in the human, it is also important to consider that the half-life estimate of 7.1 years for 2,3,7,8-TCDD was based on two serum values taken 5 years apart, with the assumption of a single compartment, and assuming a first-order elimination process (Pirkle et al., 1989). It is likely that the excretion of 2,3,7,8-TCDD in humans is more complex, involving several compartments, tissue-specific binding proteins, and a continuous daily background exposure. Furthermore, changes in body weight and body composition should also be considered in developing PB-PK models for 2,3,7,8-TCDD and related compounds in humans. Data contained in the recently reported, expanded study of half-life of 2,3,7,8-TCDD in Ranch Hand Veterans (Wolfe et al., 1994) and additional follow-up studies using blood level information from the 1992-1993 physical examination should allow for better estimates of TCDD half-life, provide important additional data to evaluate whether TCDD follows single-compartment, first-order kinetics, and provide additional information with which to study the influence of percent body fat on TCDD elimination in these veterans.

It is known that exposure occurs to the developing fetus through placental transfer of dioxin-like compounds in maternal blood via the placenta. In addition, exposure is likely to increase in the early postnatal period through intake of mother’s milk containing dioxin-like compounds. Redistribution of body burdens is likely to occur with growth and development, depending on relative intakes and changes in body fat content. Fasting, aging, and disease are all thought to alter steady-state levels of dioxin during life. These changes complicate standard pharmacokinetic models and present the possibility for temporary but potentially important increases in blood or tissue levels of dioxin-like compounds during critical periods of development, growth, and aging. Additional data on both distribution and dose to target organs and response to the tissue-specific dose in relation to development and growth will be required to refine our perspectives on the importance of these issues in evaluating dioxin hazards and risks.