This chapter discusses the immunologic effects attributed to persistent organochlorines and other hormonally active agents (HAAs). Effects of specific HAAs (such as the halogenated aromatic hydrocarbons (HAHs), dichlorodiphenyltrichloroethane (DDT), chlordecone (Kepone), endrin, aldrin, dieldrin, lindane, chlordane, toxaphene, endosulfan, and hexachlorobenzene) observed in laboratory studies and to a lesser extent in field and human studies are compared and correlated with information on exposure. Because these agents are postulated to act by means of hormonally mediated mechanisms, a brief discussion of how hormones affect the immune system is presented below. As more information becomes available, this will provide a context for evaluating the immunologic effects of HAAs.
It has been well documented that immunity can be modulated by hormones (Grossman 1984, 1985). The presence of steroid hormone receptors in a strikingly wide variety of immunologic tissues is a strong indication that cells and tissues of the immune system must be targets for steroid hormones, and that steroid hormones elicit regulatory effects in these cells and tissues. The heterogeneity of responses caused by interactions of various steroid hormones (such as corticoids, estrogens, androgens, and progestins) between the immune and endocrine systems has been documented (Grossman 1984, 1985, 1989, 1994: Grossman et al. 1991; Berczi 1994; Chapman and Michael 1994; Dardeene and Savino 1994; Fabris 1994; McCruden and Stimson 1994; Rivier 1994; Wira et al. 1994). Thus, a variety of stimuli (including exposure to environmental toxicants) could mediate nonspecific, stress like effects.
In the case of steroid hormones, it is clear that the actions of estrogens and androgens are important in the reported differences in immune response between male and female laboratory animals (Batchelor and Chapman 1965: Terres et al.continue
1968; Grossman 1984, 1985, 1989, 1994; Grossman et al. 1991). Termed "immunologic sexual dimorphism," the results of the actions are a general increase in humoral immunity in females compared with males (Batchelor and Chapman 1965; Terres et al. 1968) and differences in some cell-mediated immune responses (Graff et al. 1966, 1969; Kittas and Henry 1979, 1980) in females and males. For example, females tend to be far more susceptible than males to such autoimmune diseases as Hashimoto's disease (Tunbridge et al. 1977: Grossman et al. 1991), Grave's disease (Grossman et al. 1991), rheumatoid arthritis (Vandenbroucke 1982; Vandenbroucke et al. 1982; Lotz and Vaughan 1988; Grossman et al. 1991), systemic lupus erythematosus (Roubinian et al. 1979a,b; Lahita 1985; Grossman et al. 1991), thyroid disease (Tunbridge et al. 1977), and demyelinating disease (Arnason and Richman 1969).
Immunologic differences are also observed between pregnant and nonpregnant females. Immune responses in pregnant women are depressed compared with nonpregnant women. This depression of the immune system might be necessary during pregnancy to prevent fetal rejection and abortion before term (Grossman and Roselle 1987), and might be due, in part, to the presence of sex steroids that are elevated during pregnancy. However, it also could be partly responsible for the reported increases in the susceptibility of pregnant women to such infectious diseases as smallpox, polio, viral hepatitis, varicella-zoster, influenza, cytomegalovirus, and pulmonary and systemic mycoses (Grossman and Roselle 1987).
Diethylstilbestrol (DES) has been studied extensively in humans, and this estrogen has been shown to alter immunity (Dodds et al. 1938; Ablin et al. 1974; Korach et al. 1978; Dean et al. 1980; Kalland and Forsberg 1981; Haukaas et al. 1982; Fugmann et al. 1983; Luster et al. 1984; Morahan et al. 1984; Pung et al. 1984, 1985; Noller et al. 1988). These studies provide examples of how hormones can affect the immune system, but whether all HAAs act in this manner remains to be determined.
HAAs and Steroid Hormones
HAAs and Lymphatic Tissue Structure
Some reports of architectural changes in primary and secondary lymphatic tissues exposed to HAAs are available. Of particular interest are studies that describe thymic atrophy, since the thymus is a major site of early T-lymphocyte development, as well as a source of immunologic regulatory hormones (Dardeene and Savino 1994) in the adult. Thus, it follows that disorganization of thymic structure in the embryo could also result in immunologic abnormalities. In mammalian species, thymic atrophy and disruption of the secondary lymphatic organs has been generally observed as a result of PCB exposure (McKinney et al. 1976: Safe 1985; Thomas and Faith 1985), and these compounds can alter lymphoidcontinue
development of thymus and bursa (Andersson et al. 1991). Notably, TCDD promotes thymic involution in fish (Spitsbergen et al. 1986) and mice (Luster and Rosenthal 1986; Kerkvliet et al. 1990) and cellular depletion in thymus, spleen, and lymph nodes (Clark et al. 1981).
The action of some HAAs, such as TCDD and some PCBs, are mediated through the aryl hydrocarbon (Ah) receptor mechanism (Kerkvliet et al. 1990: Andersson et al. 1991; Kerkvliet and Burleson 1994; K. White et al. 1994), and may not be construed as direct acting HAAs. It is important to keep in mind that the regulatory pathways between the endocrine and immune systems are complex. Alterations in thymic structure and function can affect sex-and adrenal-hormone regulation of immunity, as mediated by the various thymic-hypothalamic-pituitary axes (Grossman 1984, 1985, 1989; Grossman et al. 1991).
Halogenated Aromatic Hydrocarbon Compounds
It has been well documented that HAHs such as TCDD, polychlorinated dibenzofurans (PCDFs), and PCBs, affect immune response, and they appear to affect all functional arms of the immune system (innate immunity and host resistance, cell-mediated immunity, and humoral immunity) (Table 7-1).
Specific examples of the immunologic effects of PCBs and TCDD on laboratory animals are detailed in Table 7-2. These HAHs cause atrophy of the thymus, the primary lymphoid organ in which stem cells are selected and differentiated into T-cells. Thymic atrophy has been induced in adrenalectomized animals (Vos and Luster 1989; Lundberg 1991). HAHs cause thymic involution and decrease the number of colony-forming stem cells in animals-and these effects are more dramatic when exposures occur either perinatally or postnatally (Lundberg et al. 1990; Holladay et al. 1991; Lundberg 1991; De Waal et al. 1992: De Heer et al. 1994). This suggests that HAHs target the developing immune system (Fine et al. 1990). Because TCDD directly affects the thymic cortical epithelium, it has been hypothesized that the hormonal factors necessary for lymphocyte maturation are not produced and that thymocytes are pushed into premature terminal differentiation (Greenlee et al. 1985; Lundberg et al. 1990).
Exposure to HAHs decreases cell-mediated immune (CMI) responses against bacteria and viruses. This is shown in the decreased host resistance reported in Table 7-1. Treatment with HAHs before immunization with sheep red blood cellsa T-dependent antigenresults in dose- and structure-dependent suppression of immune response in mice (Silkworth et al. 1986; Davis and Safe 1988, 1990; Dickerson et al. 1990; Kerkvliet et al. 1990; Tomar and Kerkvliet 1991). TCDD also has been shown to suppress delayed-type hypersensitivity (Vos and Luster 1989) and to suppress generation and lytic activity of cytotoxic T-cells incontinue
a dose- and strain-dependent manner (Clark et al. 1981. 1983: Nagarkatti et al. 1984). The suppression of cytotoxic T-cells could be related to a concurrent increased number of T-suppressor cells and increased suppressor activity (Clark et al. 1981, 1983; Holsapple et al. 1986); however, that hypothesis is the subject of controversy.
Although the mechanism by which HAHs alter CMI responses is unknown, studies have shown that HAHs affect these responses without decreasing T-cell proliferation, IL-2 production, or the number of IL-2 receptors (Dooley et al. 1990). There is evidence that TCDD targets activated lymphocytes rather than resting cells and that TCDD specifically inhibits the activation of antigen-specific T-cells (Dooley et al. 1990; Lundberg et al. 1992).
HAHs have been shown to affect humoral immune response. This response is characterized by B-cell antibody production, and it requires B-cell interaction with T-cells and interleukins, which are necessary for B-cell activation and differentiation. B-cells produce antibodies to specific antigens presented by T-cells (T-dependent antigen) or to antigens that cross-link surface immunoglobulins on the B-cell membrane (T-independent antigen). Exposure to HAHs followed by immunization with either T-dependent or T-independent antigens results in a dose- and structure-dependent decrease in antibody production without affecting B-cell proliferation (Davis and Safe 1988, 1990; Kerkvliet et al. 1990; Holsapple et al. 1991; Harper et al. 1995). Although the mechanism by which HAHs suppress the humoral immune response is unknown, it appears that HAHs act by means of the Ah receptor (Silkworth and Grabstein 1982; Lubet et al. 1984; Silkworth et al. 1984, 1986; Kerkvliet et al. 1985, 1990: Davis and Safe 1988. 1990; Howie et al. 1990; Tomar and Kerkvliet 1991; Howie 1992) and that immunotoxic potency correlates with binding affinity. However, it has been reported that components of immunosuppression induced by some HAHs act independently of the Ah receptor (Howie et al. 1990; Kerkvliet et al. 1990: Howie 1992).
An observational study was conducted between 1992 and 1994 to determine whether contaminant-associated immunosuppression occurs in prefledgling Caspian terns and herring gulls of the Great Lakes (Grasman et al. 1996). The phytohemagglutinin skin test for T-cell mediated immunity was conducted on 3-wk-old chicks at colonies distributed across a broad gradient of organochlorine contamination (primarily PCBs). In both species, there was a strong exposure-response relationship between organochlorines and suppressed T-cell-mediated immunity. Suppression was most severe (30-45%) in colonies in Lake Ontario and Saginaw Bay for Caspian terns and herring gulls, and in western Lake Erie for herring gulls. Although there were significant differences in total antibody and IgG titers among sites, there was no consistent exposure-response relationship with organochlo-soft
(table continued on next page)break
(table continued from previous page)break
(table continued on next page)break
(table continued from previous page)break
rines. In 1992, altered white blood cell numbers were associated with elevated organochlorine concentrations in Caspian terns but not herring gulls. Although the identity of the specific organochlorine(s) responsible for the suppression of T-cell-mediated immunity could not be determined, the researchers noted that PCBs were the most closely associated with immunosuppression.
Field studies of the immunologic effects of HAHs also have been conducted with seals, whales, and dolphins. de Swart et al. (1994, 1996) reported that innate and acquired immune responses were functionally impaired in harbor seals fed herring from PCB-contaminated waters (Baltic Sea) for 126 wk. The estimated intake of PCBs was 1,460 µg/d. Ross et al. (1995) reported similar findings in captive harbor seals fed fish from the PCB-contaminated Baltic Sea. These seals demonstrated impaired ability to mount a delayed hypersensitivity response when challenged with ovalbumin, and they generated 37% less antibody to ovalbumin after antigen challenge than did seals fed fish from the relatively uncontaminated Atlantic Ocean. In exposed seals, the combined concentration of mono-PCB (International Union of Pure and Applied Chemistry (IUPAC) numbers 118, 156, and 189) and diortho-PCB (IUPAC number 180), measured in nanograms of toxic equivalent per kilogram of lipid, was 140.0; it was 35.5 in unexposed seals.
De Guise et al. (1994, 1995) studied beluga whales living in the highly contaminated St. Lawrence estuary of Quebec, Canada, and compared them with belugas living in the much less contaminated arctic. They observed that belugas from the St. Lawrence had numerous severe and disseminated infections caused by mildly pathogenic bacteria. They suggested that the generalized immunosuppression was caused by organochlorine contamination. In addition, 75 tumors have been reported in whales worldwide, 28 (37%) of which were found in 18 St. Lawrence beluga whales. The researchers tentatively concluded that this could result from depression in immunosurveillance caused by exposure to environmental contaminants or carcinogens, or both. However, because the reported results were obtained through highly selective (nonrandom) sampling, selection bias might also skew these conclusions.
In a study of 15 bottlenose dolphins along the west coast of Florida, peripheral blood lymphocyte responses to Concanavalin A (ConA) and phytohemagglutinin were determined in vitro and compared by regression analysis with contaminant concentrations in whole blood from five of the dolphins (Lahvis et al. 1995). Reduction in ConA-induced lymphocyte responses was correlated with increasing whole blood concentrations of tetrachloro-PCBs (1-18 ng/g), pentachloro-PCBs (4-44 ng/g), hexachloro-PCBs (13-322 ng/g), heptachloro-PCBs (7293 ng/g), and octachloro-PCBs (2-81 ng/g), and similar correlations were also found with DDT (see below). Immunosuppression caused by exposure to environmental contaminants also could account for the severity and extent of morbilli-virus epizootics observed among seals and dolphins (de Swart et al. 1995). However, a direct cause-and-effect relationship has not been proven.break
Alterations in immune responses caused by exposure to HAHs have been documented in a few human studies. Lu and Wu (1985) describe the acnegenic and hepatotoxic effects in residents of Yu-Cheng, Taiwan, who ingested high concentrations of PCBs that were accidentally leaked into rice oil. The resulting effects were related primarily to increased respiratory infection; decreased serum concentrations of IgA and IgM; decreased CD4+ T-cells and increased CD8+ T-cells; suppressed dermal delayed hypersensitivity responses to a combination of streptokinase and streptodormase and to tuberculosis antigens; and augmentation of the in vitro lymphocyte mitogen stimulation to phytohemagglutinin (PHA) and pokeweed mitogen (PWM), but not to ConA. Average blood concentrations of PCBs in the affected individuals were 89 + 6.9 ppb.
In a study of Wisconsin infants whose mothers ate PCB-contaminated fish, maternal serum PCB levels were positively associated with the number and type of infectious illnesses, such as colds, earache, and flu symptoms, that occurred in infants during the first 4 mo of life (Smith 1984). The authors concluded that prenatal exposure to PCBs was the cause of the increased infections. However, as noted by Swain (1991) in a critique of the study, these results should be interpreted carefully because blood concentrations of PCBs were only measured after birth and not during pregnancy.
In a study of Dutch infants, 105 breast-fed and 102 formula-fed infants were evaluated from birth until 18 mo to determine whether prenatal and postnatal exposure to background concentrations of PCBs and dioxins had an effect on the incidences of rhinitis, bronchitis, tonsillitis, and otitis (Weisglas-Kuperus et al. 1995). Humoral immunity was also measured by detecting antibody levels to mumps, measles, and rubella as a result of vaccinations. Prenatal exposure was estimated by PCBs in maternal blood and the total toxic equivalent (TEQ) level in breast milk (measured as pg TEQ/g milk fat), and postnatal exposure was calculated as a product of the total TEQ level in breast milk multiplied by the weeks of breast feeding. Umbilical cord and venous blood was taken from a subgroup of 55 infants at 3 and 18 mo for white blood cell counts and immunologic marker analysis. No relationship was found between pre- and postnatal PCB/dioxin exposure and upper or lower respiratory symptoms or humoral antibody production. However, higher prenatal and postnatal exposures to PCBs/ dioxins were associated with lower monocyte and granulocyte counts at 3 mo, and increases in the total number of T-cells and in the number of cytotoxic T-cells were observed at 18 mo.
Recent studies of Inuit people exposed to organochlorines in their diet via sea-mammal fat have reported serum lipid concentrations of 4.1 mg/kg lipids PCBs and 184.2 ng/kg lipids 2,3,7,8-TCDD (Ayotte et al. 1997). Health risk assessments for newborns in these populations indicate a correlation between PCB/dioxin exposures in breast milk and suppressed levels of white blood cellscontinue
in infants (Ayotte et al. 1996). The breast milk of Inuit women contained 7 times more PCBs (sum of the PCB congeners = 1,052 ng/g, lipid) than the milk from women from urban, industrialized areas south of Quebec (Dewailly et al. 1993b). Researchers are investigating the possible connection between unusually high rates of infectious disease, particularly acute ear infections, among Inuit children and exposure to PCBs. Such studies must be interpreted carefully because comparisons of organochlorine concentrations over time are unreliable.
Webb et al. (1989) found that humans exposed to TCDD had increased CD8+ T-lymphocyte populations; no change in CD4+ T-cells; no change in lymphocyte response to the mitogens ConA, PHA, and PWM; no change in cytotoxic T-cells; and increased serum IgA. Of the 41 individuals studied, 16 had TCDD concentrations below 20 ppt in their adipose tissue, 13 had concentrations of 20-60 ppt, and 12 had concentrations above 60 ppt (the maximum was 750 ppt).
DDT has been reported to possess estrogenic and antiandrogenic properties (Kupfer and Bulger 1980), supporting the hypothesis that it acts by binding to steroid receptors in immunologic target tissues. Laboratory studies have demonstrated that DDT can alter both the primary and the secondary humoral immune response, immunoglobulin production, splenic plaque-forming cell (PFC) response, histamine concentrations, and mast cell numbers. Specific examples of DDT's immunologic effects in laboratory animals are detailed in Table 7-3. DDT has been demonstrated to trigger some immunologic effector mechanisms in animal and bird models (Barnett and Rodgers 1994).
In studies of harbor seals, a diet of DDT-contaminated fish from the Baltic Sea was shown to impair immune response, as measured by delayed hypersensitivity in the skin to ovalbumin and in vitro lymphocyte assays (de Swart et al. 1994; Ross et al. 1995). In a study of bottlenose dolphins, Lahvis et al. (1995) reported that ConA mitogen assays of peripheral lymphocytes demonstrated a correlation between reduced immune response and increasing concentrations of p,p'-DDT (0-24 ng/g) and p,p'-DDE (13-536 ng/g). l,l-Dichloro-2,2-bis(p-chlorophenyl)ethylene (p,p'-DDE) is a metabolite of DDT. It has been suggested that, with elevated concentrations of these contaminants, a reduction in immune response can be correlated with an increased incidence of infection (Svensson et al. 1994).
Overall, the data are suggestive of DDT-mediated immunosuppression in animals and birds being detected at about 10 mg per kilogram of body weight (BW) per day. Additional laboratory studies are needed to identify the functional parameters and immunologic effector cells involved in DDT-mediated immunotoxicity. Until human data are available for comparison, no conclusions can be made about the effects of DDT on the human immune system.break
Chlordecone binds to estrogen receptors (Hammond et al. 1979) and could thus modulate immunity via binding to them. The only hard evidence in support of an immunomodulatory effect of chlordecone is from a study of the immunologic effects of malnutrition and administration of chlordecone in rats. Chetty et al. (1993) reported that 95-100% of malnourished rats fed 5 mg/kg/d died; that 0.5 mg/kg/d decreased body weight and increased spleen weight; that both malnutrition and chlordecone increased PFC; and that 0.5 and 5 mg/kg/d plus a calcium diet increased PFC response, but that 5 mg/kg/d plus a protein diet decreased PFC response.
Endrin, Aldrin, and Dieldrin
Data from laboratory studies with animals show that organochlorine pesticides, such as endrin, aldrin, and dieldrin, can be immunotoxic at very low doses (0.065-36 ppm) (depending on the model system and the route and duration of exposure). Organochlorine pesticides affect immune system functions either because they directly interact with immune effector cells or because their metabolic products do so. Specific examples of the immunologic effects of endrin and dieldrin in laboratory animals are shown in Table 7-4. The primary action appears to be in the macrophage processing of antigen (Loose et al. 1981; Loose 1982).
Limited information is available about the immunologic consequences of human exposure to organochlorine pesticides. In one case study (Muirhead et al. 1959), a pesticide sprayer developed immunohemolytic anemia after multiple exposures to dieldrin, heptachlor, and toxaphene. Specifically, the individual had circulating antibodies against dieldrin-coated erythrocytes and heptachlor-coated erythrocytes. However, because the patient also was exposed to other pesticides, including DDT, the usefulness of the information is limited. Immunohemolytic anemias also have been found after multiple dieldrin exposures in workers (Hamilton et al. 1978), or after consumption of dieldrin-contaminated fish (Hamilton et al. 1978). Loose et al. (1981) suggest that the human threshold for dieldrin immunotoxicity is 7 x 10-5 mg/kg/d. Given the limited amount of human data, definitive conclusions cannot be drawn about the effects of this family of pesticides on human health.
Lindane (primarily the Y-isomer of hexachloro-cyclohexane or benzene hexachloride) has been reported to perturb immune function after prenatal and postnatal exposure. Specific examples of the immunologic consequences of in vitro and laboratory exposures to animals are presented in Table 7-5. These effects appear to encompass both the nonspecific and the specific arms of thecontinue
(table continued on next page)break
(table continued from previous page)break
immune system, including effects on humoral and cell-mediated immunity. Notably, lindane at concentrations of 0.012-10 mg/kg stimulated antibody production; production was inhibited at 100 mg/kg. Mitogen response and delayed hypersensitivity displayed similar biphasic effects (stimulation at 10 mg/kg; inhibition at 100 mg/kg). Although lindane at 1 mg/kg inhibited phagocytosis, higher concentrations were needed to inhibit other nonspecific elements of immunity (10 mg/kg for lysozyme concentrations; 300 mg/kg for natural killer-cell activity). Given the various concentrations of lindane tested, diverse routes of administration, differences in length of treatment, and variety of animal models, the exact sites of action of this pesticide and its mechanisms of action on the immune system remain clouded in speculation. Furthermore, with just one limited in vitro laboratory study on the effects of lindane on human immunity, it is premature to extrapolate broad conclusions about its effects in humans.
Some studies that use laboratory animal models suggest that exposure to chlordane (also under the trade names Octochlor and Velsicol 1068) leads to moderate immunotoxicity. Specific examples of the effects of this pesticide in laboratory animals are presented in Table 7-6. In inhalation studies, chlordane administered at 1 or 10 mg/m3 for 90 d increased lymphocyte numbers in female rats; in vitro tests that used 10 M chlordane showed mitogenic activity. In addition, prenatal exposure to chlordane was reported to alter immune responsessuch as delayed hypersensitivity, macrophage activation, and colony-forming unit activityin offspring of exposed rats. The majority of studiesregardless of the concentration, route of administration, or time of applicationfound no histologic or functional changes that could be attributed to chlordane treatment.
Very limited information is available on the possible immunotoxic effects of chlordane exposure for humans. McConnachie and Zahalsky (1992) report significant changes in cell-mediated immunity and humoral immunity in humans exposed to chlordane aerosols for 3-15 mo. They report impairment in lymphoproliferation to mitogens ConA and phytohemagglutinin A (PHA) and increased titers of autoantibodies in 11 of the 12 subjects tested. These tests were performed from 4 mo to 10 yr after exposure, implying long-term immunotoxicity for this agent. However, the studies are too limited to support any conclusions about the effects of chlordane on human health.
There are few data to describe the immunologic effects of toxaphene in laboratory animal models. In one study (Trottmann and Desaiah 1980), thymus weight was reduced in mice after oral administration of toxaphene at 22.5 and 30 mg/kg/d for 14 d. Koller et al. (1983) observed depressed IgG antibody produc-soft
tion in rats treated with 0.5, 1.5, and 10 mg/kg/d for 6 wk. Allen et al. (1983) treated mice with toxaphene at 15 and 30 mg/kg/d for 8 wk and report depressed IgG antibody production. No effect on delayed hypersensitivity was reported.
In a prenatal exposure study (Chernoff et al. 1990), pregnant rats were administered 32 mg/kg/d toxaphene by gavage from the onset of pregnancy until gestation d 8, 12, or 16. Rats were killed at these time points or on d 20 of gestation. Spleen weight was significantly reduced in fetuses from rats sacrificed on d 8, 16, and 20; thymus weight was reduced in fetuses from rats sacrificed on d 8 and 20.
The studies above are limited, and additional studies are required before a reliable assessment of the possible immunologic effect of toxaphene on animals or humans can be made.
Exposure to endosulfan has been reported to produce immunotoxic changes in nonspecific immunity and in humoral and cell-mediated responses in laboratory animals. Immunologic studies with endosulfan are summarized in Table 7-7. Oral exposure to endosulfan induced immunologic effects at 1-5 mg/kg/d (from 6-22 wk), but inhalation and dermal routes did not. The endosulfan studies are limited, and there is little information on the immunologic effects of this compound in humans.
Exposure to hexachlorobenzene has been reported to produce histologic changes in lymphoid tissue architecture in laboratory animals. Vos et al. (1983) report that prenatal and postnatal exposure to hexachlorobenzene at 4 mg/kg/d in feed enhanced humoral and cellular immune responses in rats. Hexachlorobenzene also promoted accumulation of macrophages in the rats' lungs. The high endothelial venules present in the lymph nodes underwent abnormal proliferation, accompanied by lymphoid hyperplasia in the splenic white pulp, in rats fed this compound at 25-100 mg/kg/d for 3 wk (Vos et al. 1979). Hyperplasia of lymphoid tissue in the stomach has been induced in dogs fed hexachlorobenzene at 6.5-10 mg/kg (1 mg/d) for 12 mo (Gralla et al. 1977).
No information is available about the possible immunotoxicity of hexachlorobenzene in humans, and the data from laboratory studies are too limited to support any conclusions about how this compound affects human immune response.
Summary and Conclusions
There are very few studies of the immunologic effects of human exposure to HAAs, but for some chemicals there are adequate data for laboratory animals.continue
(table continued on next page)break
(table continued from previous page)break
Animal studies have identified immunotoxic and immunomodulatory effects. The most extensively studied compounds are the HAHs. Laboratory studies have shown that HAHs affect the functional arms of the immune system. Field studies of birds show a strong exposure-response relationship between organochlorines and immune suppression. Experimental studies have also shown that innate and acquired immune responses were impaired in seals fed fish from the contaminated Baltic Sea. This immunosuppression is believed to be the reason for increased incidences of bacterial and viral infections in seal populations found in contaminated waters. There have only been a few studies of the effects of HAAs in humans, but the results of laboratory and wildlife studies suggest that HAAs have the potential to affect human immune functions. Certainly, additional clinical immunologic end points must be studied. As noted by Kerkvliet and Burleson (1994) ''massive retrospective studies on poorly defined exposure groups cannot be justified to try to 'prove' that immune modulation has occurred in these people." The authors state that "research must focus on the definition of sensitive end points (i.e., biomarkers) of immune dysfunction in humans. . . . In particular it is important to determine in animal models how well changes in immune function in the lymphoid organs (e.g., spleen and lymph nodes) correlate with changes in the expression of lymphocyte subset/activation markers in peripheral blood. Until such correlations are established, the interpretation of changes observed in subset/activation markers in human peripheral blood lymphocytes in terms of health risk will be limited to speculation." HAHs are thought to act through Ah receptor binding, but some components of HAH-mediated immune suppression could function through other, independent mechanisms.
The data available on DDT suggest that DDT mediates immunosuppression in laboratory animals and in birds. Certainly, DDT possesses estrogenic properties, and it could act by binding to steroid receptors to modulate immunity. Chlordecone also could modulate immunity through steroid-receptor-binding pathways. Studies of endrin, aldrin, and dieldrin are limited, but the immunologic effects reported for these chemicals in laboratory studies appear to involve macrophage processing of antigen. Reported effects from lindane encompass both the specific and the nonspecific arms of the immune system, and in laboratory animals such responses have included effects on humoral and cell-mediated immunity. Exposure to endosulfan in laboratory studies also has produced immunotoxic effects on nonspecific immunity and altered humoral and cell-mediated responses. In studies of chlordane, moderate immunotoxicity has been observed in laboratory animals, but most studies have not identified histologic or functional changes. Histologic changes in lymphoid tissue architecture were found in laboratory animals after exposure to hexachlorobenzene. There is little information available on the immunologic effects of toxaphene.
Generally, the available data neither support nor refute the premise that thecontinue
actions of HAAs is mediated either directly or indirectly through endocrine pathways. It can be stated with some degree of certainty that some HAAs affect one or more aspects of immune function, at least in animal models. Field studies, especially in marine mammals, generally support this view, although there is intrinsic uncertainty for such studies because the conditions of exposure to the environmental contaminants responsible are not known and mixtures of the compounds are not always clearly defined.
In human studies, the cause-and-effect relationship between HAAs and immunotoxicity is not clear cut. Thomas (1995) considers all the available data on Great Lakes residents exposed to potentially immunotoxic agents through the food chain and concludes that, based on "uncertainties with regard to exposure levels, predictability of tests, suitability of the animal models, and immune reserve . . . there is no definite evidence as yet that environmental [exposure] to these xenobiotics poses a significant threat to the human immune system." In the few human studies available, exposure to mixtures, extended delays between time of exposure and performance of immunologic tests, and the effects of other confounding variablesage, sex, lifestyle, underlying diseaseall tend to limit support for any definitive conclusion. In addition, immunosuppressive effects of background exposures have not been determined. Thus, although animal studies suggest that HAAs can cause immunologic effects, underlying mechanisms are not clear. It is also unclear whether they have similar effects on the human immune system.
Based on the committee's review of the extensive laboratory animal data on immunologic effects of HAAs, as well as the limited information from wildlife and human studies, the following are recommended:
Comprehensive epidemiologic studies that evaluate a variety of health effects, including immunologic effects, of human populations suspected of being affected by HAAs should be initiated. Especially needed are studies of cohorts established either through registries or directed effort to assess the prevalence of autoimmune problems in offspring whose mothers were exposed during pregnancy. To address the potential problem of measuring exposure in a case-control or cohort design, it is suggested that populations known to have been heavily exposed to HAAs (such as the Seveso population) be used for cohort studies. Ideally, such cohorts should be followed throughout their lifetime.
Epidemiologic studies should use clinically relevant immunologic assays, such as those for monitoring concentrations of circulating antibodies to thymus-dependent antigens; antigen test banks to monitor delayed-hypersensitivity skin reactions; quantitative lymphocyte subclass identifications; in vitro measurementscontinue
of lymphocyte cytokine production and possibly mitogenic responsiveness: and the lytic action attributed to cytotoxic lymphocytes and natural killer cells in exposed populations to clarify the relationship between HAA exposure and human health.
Because much of the available immunologic laboratory data on HAAs is on chemicals that have been regulated and, in most cases, are no longer used in the United States, future studies should focus on chemicals that are being used, such as endosulfan and lindane.break