Introduction

Human exposures to relatively high levels of PCBs have occurred primarily in persons working in plants that extensively manufactured and used PCBs and PCB-containing equipment. Occupational exposure to PCBs can result in a broad spectrum of effects that includes

• Increased levels of some liver enzymes, with possible hepatic damage,
• Chloracne and related dermal lesions, and
• Respiratory problems [Alvares et al. 1977; Chase et al. 1982; Emmett and Emmett 1985; Lawton et al. 1985; Meigs 1954; Ouw et al. 1976; Safe S 1993; Warshaw et al. 1979].

Potential adverse human health effects of low-level environmental exposure to PCBs are complex and still need further validation [Safe SH 2007].

In animal studies, commercial PCBs elicit a broad range of toxic responses including:

• Acute lethality,
• Body weight loss,
• Carcinogenesis,
• Dermal toxicity,
• Fatty liver,
• Genotoxicity,
• Hepatomegaly,
• Immunosuppressive effects,
• Neurotoxicity,
• Porphyria,
• Reproductive and developmental toxicity,
• Thymic atrophy, and
• Thyroid hormone-level alterations.

This adverse health effects section addresses PCBs as a whole.

Mechanism of PCB Toxicity

PCBs are metabolized by the microsomal monooxygenase system catalyzed by cytochrome P-450 to phenols (via arene oxide intermediates), which can be conjugated or further hydroxylated to form a catechol. Arene oxide intermediates are electrophilic in nature. They can covalently bind to nucleophilic cellular macromolecules (e.g., protein, DNA, RNA) and induce DNA strand breaks and DNA repair, which can contribute to the toxic response of PCBs. Additionally, arene oxide intermediates can be conjugated with glutathione and further metabolized to form methylsulfonyl metabolites, which have been identified in human serum and tissue samples and in laboratory animals. Binding of methylsulfonyl metabolites to some proteins may contribute to some of the toxic effects of PCBs. It has also been hypothesized that hydroxylated PCB metabolites could contribute to the toxicity of PCBs [ATSDR 2000; Safe SH 2007].

Dermatologic Effects

Chloracne and related dermal lesions have been reported in workers occupationally exposed to PCBs. Mild to moderate chloracne was observed in 7 of 14 workers exposed to 0.1 mg/m³ Aroclors for an average duration of 14.3 months [Meigs 1954]. Among 80 workers who manufactured capacitors in Italy, 10 cases of acne or folliculitis, or both, and 5 cases of dermatitis were reported. All of the workers with chloracne were employed in high exposure jobs. Their blood PCB concentrations ranged from 41 to 1319 µg/kg [Maroni et al. 1981].

In a person with PCB-induced chloracne, the acne-like lesions arise as a result of inflammatory responses to irritants in the sebaceous glands. Chloracne usually begins with the formation of keratin plugs in the pilosebaceous orifices. The resulting inflammatory folliculitis stimulates keratinization of the sebaceous gland ducts and outer root sheath of the hair, leading to the formation of keratin cysts.

The chin, periorbital, and malar areas are most often involved, although lesions might also appear in areas not usually affected by acne vulgaris (e.g., the chest, arms, thighs, genitalia, and buttocks). The most distinctive lesions are cystic and measure 1-10 mm, although comedonal lesions can also be present. The cysts and comedones can become inflamed and secondarily infected, and papules and cysts can be surrounded by edema and erythema [Crow 1970; Letz 1983].

Chloracne generally indicates systemic toxicity and can be caused by not only dermal contact but also ingestion of PCBs. However, the absence of chloracne does not rule out exposure [Kimbrough 1980; Letz 1983]. No reliable dose-response model exists for chloracne in exposed populations, and the dose-response relationship might be dependent on individual predisposition. Chloracne typically develops weeks or months after exposure. The lesions are often refractory to treatment and can last for years or decades.

In addition to chloracne, other dermal effects noted in some PCB-exposed workers include pigmentation disturbances of skin and nails, erythema and thickening of the skin, and burning sensations [Fischbein et al. 1982; Fischbein et al. 1979; Ouw et al. 1976; Smith et al. 1982].

Skin effects were reported widely among victims of the Yusho (Japan) and Yu-Cheng (Taiwan) poisoning episodes in 1968 and in 1978, respectively. In these episodes, persons were exposed to PCBs and their heat-degradation products, mainly polychlorinated dibenzofurans (PCDFs). Exposure to PCBs occurred by consuming rice oil that had become contaminated by heat-degraded PCBs during processing. Unlike usual PCB mixtures, the Yusho and Yu-Cheng mixtures were heated in thermal heat exchangers during the cooking process, resulting in contamination of the oil by chlorinated dibenzofurans as well as PCBs. This co-contamination created controversy [Anonymous 1997; Kimbrough et al. 2003; Ross 2004] about the extent to which the health effects observed in the Yusho and Yu-Cheng populations can be attributed to PCBs legitimately, as opposed to the dibenzofuran co-contaminants.

No adverse dermal effects have been reported in persons who consume large amounts of Great Lakes fish contaminated with PCBs and other environmentally persistent chemicals, or in other cohorts from the general population. However, whether this outcome was systematically studied in these cohorts is unknown [ATSDR 2000].

A skin lesion exactly like chloracne in humans has been observed in several species of animals experimentally exposed to PCBs [Allen 1975]. After monkeys incur long-term oral exposure to commercial PCB mixtures, related dermal effects are well characterized and generally are similar to those observed in humans [ATSDR 2000].

Reproductive and Developmental Effects

Courval, DeHoog et al. [1999] conducted a study of 626 married couples in Michigan. The relative risk of conception failure (defined as inability to conceive after 12 months) rose in men but not in women with increasing consumption of PCB-contaminated fish. Some evidence shows that increased intake of PCB-contaminated fish can shorten menstrual cycle length, but no adverse association was found between the duration of fish consumption and time-to-pregnancy in the same population.

In a study of 1,820 multigravida women, no significant association was found between low-to-moderate PCB intake and clinically recognized spontaneous fetal death [Mendola et al. 1995].

A recent oocupational cohort study examined the data from 2595 live births of female workers from three capacitor plants and found no evidence of altered sex ratio among children born to PCB-exposed female workers [Rocheleau et al. 2011].

The first epidemiologic investigation to demonstrate an association between the amounts of PCB-contaminated fish eaten by pregnant women and behavioral deficits in their newborns was the Michigan Maternal Infant Cohort Study, published in 1984 [Fein et al. 1984; Jacobson SW et al. 1985]. In this study, developmental and cognitive deficits were observed in the children of mothers who had eaten moderate to high amounts of contaminated fish during the six years preceding pregnancy and who continued to do so during pregnancy. Developmental effects in this population included statistically significant decreases in

• Gestational age (4.9 days),
• Birth weight (160-190 grams), and
• Head circumference (0.6 centimeters).

In addition, infants born to mothers who had eaten the greatest amount of contaminated fish during pregnancy exhibited weaker reflexes, greater motor immaturity, and more pronounced startle responses compared with infants born to women who had consumed less fish.

It is essential that women of childbearing age be aware of fish advisories to ensure they not only limit their consumption of fish with elevated PCB levels but also learn how to prepare fish to limit their PCB ingestion.

Follow-up studies of the children from this cohort have demonstrated that the effects of perinatal exposure to PCBs are persistent. At four years of age, these children still had deficits in

• Weight gain,
• Depressed responsiveness, and
• Reduced performance on the visual recognition-memory test.

At 11 years of age, the children of highly exposed mothers were

• Three times more likely than controls to have low full-scale verbal IQ scores,
• Twice as likely to lag behind at least 2 years in reading comprehension, and
• More likely to have difficulty paying attention [Jacobson JL et al. 1990a, 1990b].

Recent studies indicate that maternal consumption of PCB-contaminated fish can cause disturbances in reproductive parameters and neurobehavioral and developmental deficits in newborns and older children. Prenatal exposure to PCBs from the mother's body burden, rather than exposure through human milk, is believed to account for the developmental effects of these compounds [Jacobson JL et al. 1996; Longnecker et al. 2003; Ribas-Fito et al. 2001; Schantz et al. 2003].

Similar reproductive, developmental, and neurobehavioral deficits have been reported in children born to women who were pregnant during the Yusho and Yu-Cheng incidents [Hsu et al. 2003; Hsu et al. 2005; Yang et al. 2005].

Developmental delays were seen at all ages and were greater in children who were smaller and had neonatal signs of intoxication or nail deformities, or both. Follow-up testing indicated that effects on cognitive development persisted for several years after exposure [Guo et al. 1995].

In rhesus monkeys, exposure to PCBs is associated with alterations in the menstrual cycle, decreases in fertility, increases in spontaneous abortion, and a reduced number of conceptions [Arnold et al. 1990; Barsotti et al. 1976].

Endocrine Effects

Limited but corroborative occupational data indicate a potential for toxic effects in the thyroid system in humans. Studies that have examined relationships between exposure to PCBs and thyroid hormone status have reported a variety of results. Findings include both negative and positive significant correlations between exposure to PCBs and circulating levels of thyroid-stimulating hormone (TSH), T4, or T3. These findings are dependent on the:

• Specific type of analysis for exposure to PCBs,
• Age of the cohort, and
• Specific exposure scenario [Emmett et al. 1988; Koopman-Esseboom et al. 1994; Langer et al. 1998; Longnecker et al. 2003; Nagayama et al. 1998; Osius et al. 1999].

In a Dutch population, elevated levels of PCBs correlated with lower maternal levels of circulating triiodothyronine and total thyroxine and with higher plasma levels of TSH in infants during the second week and third month after birth. Infants exposed to higher levels of PCBs also had lower plasma levels of free thyroxine and total thyroxine in the second week after birth [Koopman-Esseboom et al. 1994].

In addition, a significantly elevated odds ratio for goiter was found among the Yu-Cheng cohort [Guo et al. 1999], suggesting the possibility of excessive thyroid disease in a population that experienced relatively high exposures to mixtures of PCBs and PCDFs.

Thyroid hormones are essential for normal behavioral, intellectual, and neurologic development. Thus, the deficits in learning, memory, and attention processes among the offspring of women exposed to PCBs are partially or predominantly mediated by alterations in hormonal binding to the thyroid hormone receptor [ATSDR and EPA 1998]. Some PCB congeners are capable of competing with endogenous hormone for binding to this receptor, suggesting a possible mechanism of thyroid toxicity. Hydroxylated PCB metabolites appear to be particularly potent in this regard [ATSDR 2000].

Studies in animals, including rodents and primates, provide evidence of thyroid hormone involvement in PCB toxicity. The most convincing evidence that PCBs can exert toxicity by disrupting thyroid hormone system derives from two studies in rats [Cooke et al. 1996; Goldey et al. 1998].

The contribution of persistent organic pollutants (POPs) exposure to the incidence of diabetes has received little attention until recently. Recent studies in populations exposed to PCBs and chlorinated pesticides found a dose-dependent elevated risk of diabetes [Carpenter 2008].

Hepatic Effects

Evidence for liver effects of occupational exposure to PCBs is essentially limited to elevation of serum liver enzymes that are routinely examined in clinical assays. These serum liver enzymes include aspartate aminotransferase (AST), alanine aminotransferase (ALT), glutamyl transpeptidase (GGT) and other biochemical indices (e.g., bilirubin). No overt hepatotoxicity has been seen in workers exposed to PCBs [ATSDR 2000].

A cross-sectional survey found no significant differences in liver function test results between workers who manufacture capacitors with low-level chronic exposure and non-exposed controls [Fischbein et al. 1979]. However, in another cross-sectional study, liver function tests showed abnormalities that seemed to correlate with serum PCB levels [Maroni et al. 1981].

Increases in urinary excretion of porphyrins appear to be associated with occupational exposure to PCBs, an effect that is believed to be secondary to the induction of hepatic microsomal enzymes. Total bilirubin levels exhibit a positive correlation with serum PCB levels [Colombi et al. 1982; Maroni et al. 1984; Smith et al. 1982].

PCBs are more potent enzyme inducers than phenobarbital, a drug that occasionally causes clinical problems due to its microsomal enzyme-inducing effects. The health implications of enzyme induction include the occurrence of disease secondary to increased metabolism of endogenous or exogenous substances and interference in medical therapy due to increased metabolism of administered drugs. The enzyme-inducing effects of PCBs can persist long after cessation of exposure [Letz 1983].

In the Yu-Cheng population, the incidence of chronic liver disease and cirrhosis was significantly higher than the incidence of these conditions in the general population of Taiwan. Asymptomatic hepatomegaly has been reported in exposed workers, many of whom had concomitant elevated serum PCB levels. Due to the mixed chemical nature of the exposure, the results cannot be attributed solely to PCBs [ATSDR 2000].

Liver damage is a consistent and prominent finding among animals exposed to PCBs, particularly rats and monkeys, which are the species tested most extensively. Liver effects are similar in nature among species and appear to be reversible when mild. Liver effects characteristically include

• Fat deposition,
• Fibrosis,
• Hepatic microsomal enzyme induction,
• Increased serum levels of liver-related enzymes indicative of possible hepatocellular damages,
• Liver enlargement, and
• Necrosis [ATSDR 2000].

Carcinogenic Effects

Epidemiologic studies have raised concerns about the potential carcinogenicity of PCBs.

A retrospective analysis of a study of two plants that manufactured electrical capacitors in the United States found a significant increase in the incidence of cancer. The primary target tissues for the cancers were the liver, gallbladder, and biliary tract [Brown 1987].

Likewise, an increased incidence of melanomas associated with exposure to PCBs has also been observed for workers who manufactured capacitors [Bahn et al. 1976; Ruder et al. 2006; Sinks et al. 1992]. Sinks et al. [1992] observed the increased risks for brain cancer among workers exposed to PCBs in an electrical capacitor manufacturing plant in Indiana, and this finding has been further confirmed by a recent study from Ruder et al. [2006].

One study suggests that exposure to electrical insulating fluids, for which the main constituent is PCBs, may cause malignant melanoma of the skin [Loomis et al. 1997].

The results of a mortality study of workers employed between 1944 and 1977 at an electrical capacitor manufacturing plant were recently reported. The report pointed out that PCBs alone or in combination with other chemicals could be associated with increased risks for

• Liver or biliary,
• Stomach, intestinal, and
• Thyroid cancers [Mallin et al. 2004].

A recent analysis of a cohort of 24,865 capacitor-manufacturing workers exposed to PCBs at three plants showed evidence of associations between cumulative exposure to PCBs and increased total cancer and intestinal cancer mortality among female long-term workers and excess myeloma for male long-term workers [Ruder et al. 2014].

In contrast, increased cancer incidence was not observed in male workers who manufactured capacitors in Sweden exposed to PCBs for an average of 6.5 years [Gustavsson et al. 1986]. The results from the Swedish study, however, cannot rule out the possibility of a carcinogenic risk from PCB exposure because of the small size of the cohort and relatively brief follow-up period.

Different mixtures of PCBs had different potencies and, thus, different toxicity. As noted previously, PCB mixtures found in the environment are different from commercial PCB mixtures. EPA agreed that some mixtures of PCBs are more likely to cause cancer than others, and found that all PCBs mixtures can cause cancer [Cogliano 1998; EPA 1996c].

In environmental case-control studies that compared PCB concentrations in breast tissue in both women with (case patients) and without (case controls) breast cancer, some studies reported higher levels of total PCBs among case patients than control patients [Falck et al. 1992; Guttes et al. 1998; Wassermann et al. 1976]. Other studies found no elevated PCB levels in breast tissue in patients with breast cancer [Aronson et al. 2000; Liljegren et al. 1998; Unger et al. 1984]. A recent occupational cohort study found no overall elevation in breast cancer risk after occupational exposure to PCBs [Silver et al. 2009].

In persons without known occupational exposure to PCBs, elevations of PCB level in the adipose tissue and serum have been associated with an increased risk of non-Hodgkin lymphoma (NHL) [De Roos et al. 2005; Engel et al. 2007; Hardell E et al. 2001; Hardell L et al. 1996; Rothman et al. 1997].

After registering as Yusho victims, 887 male and 874 female patients were observed for an average 11 years. A retrospective study found statistically significant increased liver cancer mortality rates among the males compared to national liver cancer mortality rates [Kuratsune et al. 1987].

A retrospective mortality study of 1940 Yu-Cheng cases found no statistically significant increased mortality from liver and intrahepatic bile duct cancers [Hsieh et al. 1996].

Before the comprehensive study conducted by Mayes et al. [1998], only commercial mixtures 60% chlorinated had been tested, and controversy existed about whether mixtures with lower chlorine content were carcinogenic. The Mayes et al. study [Mayes et al. 1998] supported the position that all PCB mixtures can cause cancer. Data from animal studies have shown that PCBs cause gastrointestinal tract tumors, hepatocarcinomas, leukemia, lymphomas, and pituitary tumors [ATSDR 2000].

On the basis of these laboratory data, EPA has determined that PCBs are probable human carcinogens and has assigned them the cancer weight-of-evidence classification B2 [IRIS 2012]. DHHS concluded that PCBs are reasonably anticipated to be carcinogenic in humans based on sufficient evidence of carcinogenicity in animals [NTP 2011].

In February 2013, 26 experts from 12 countries met at the International Agency for Research on Cancer (IARC), Lyon, France, to reassess the carcinogenicity of PCBs. The Working Group considered more than 70 independent epidemiological studies with informative data for carcinogenicity of PCBs in human beings. On the basis of sufficient evidence of carcinogenicity in humans and experimental animals, the IARC classified PCBs as carcinogenic to humans (Group 1). The classification is based on consistent association between exposure to PCBs and increased risk of melanoma in humans [IARC 2013].

Other Effects

Occupational and epidemiologic studies have suggested or demonstrated other adverse health effects from exposure to PCBs. These health effects can involve the:

• Cardiovascular,
• Gastrointestinal,
• Immune,
• Musculoskeletal, and
• Neurological systems.

In southwest Quebec, adults who ate fish from PCB-contaminated waters had

• Significantly greater motor retardation,
• Poorer results on certain memory and attention tests, and
• Higher scores on a standardized confusion scale than did control adults.

These neurological deficits were directly related to the frequency of fish consumption [Mergler et al. 1998].

Immune system effects reported in PCB-exposed populations include alterations in the ratio of helper to killer (CD4+/CD8+) T-cells, decreases in IgA and IgM antibody levels, decreases in monocyte and granulocyte counts, and decreases in natural killer cell count [Svensson et al. 1994].

In the Yusho and Yu-Cheng populations, the immunosuppressive effects of PCB exposure were associated with an increased incidence of persistent respiratory infection and enhanced responsiveness to mitogens [Guo et al. 1995].

Appetite loss has been reported in transformer and electrical equipment manufacturing workers exposed to various PCB-containing mixtures. Other nonspecific gastrointestinal symptoms experienced by workers exposed to PCBs include nausea, epigastric distress and pain, and intolerance to fatty foods [Emmett et al. 1988; Smith et al. 1982].

A recent study has indicated that several PCB metabolites induce gene mutations, chromosome breaks, chromosome loss and polyploidization in cells in culture and even provided the first evidence that a PCB congener is mutagenic in vivo [Robertson and Ludewig 2011].

Progress Check