TERATOLOGICAL ASPECTS OF DENTAL AMALGAM K.S. LARSSON
Department of Odontological Toxicology Karolinska Institutet Box 4064 S-141 04 Huddinge, Sweden Adv Dent Res 6:114-119, September, 1992
Abstract—The teratogenic effect is determined by four factors: (1) the agent, (2) the dose, (3) the stage of embryonic development, and (4) the genetic constitution of the embryo. The first two factors are of particular interest and warrant further comment. It should be emphasized that the mercury released from dental amalgam is mainly metallic mercury vapor. The dose of mercury vapor from dental amalgam fillings in the order of 5 pg/day is very low compared with the doses in a teratological study and is not likely to exceed the threshold necessary for a teratogenic effect to occur. The concentration of the teratogen at the target tissue is determined not only by the degree of placental transfer but also by other factors, such as the distribution within the maternal organism, the affinity to the fetal liver and blood, the hematocrit value, and the passage through the ductus venosus. These factors might help to explain toxicological mechanisms and species differences and have to be considered if the results of animal experiments are to be extrapolated to human conditions. Neither epidemiological data nor animal experimental data indicate that the release of metallic mercury vapor from dental amalgam therapy should cause teratogenic effects. A comparison with the incorporation of the fetotoxic methyl mercury might be justified.
T
his review will cover the scientific evidence for sideeffects of dental amalgam, with special reference to teratogenic and any other systemic effects. The question of systemic effects has been dealt with in two recent documents. The WHO Environmental Health Criteria 118, Inorganic Mercury (1991) concludes with the following statement: "Recently, there has been an intense debate on the safety of dental amalgams and claims have been made that mercury from amalgam may cause severe health hazards. Reports describing different types of symptoms and signs and the results of the few epidemiological studies produced are inconclusive." The Swedish National Board of Health and Welfare (1991) addressed physicians and dentists recently on the effects on health of mercury from amalgam: "Mercury can cause allergic contact dermatitis. However, this is rare. In some cases oral symptoms such as transient swelling, redness, erosion and socalled oral lichen have been observed. The etiology of oral lichen is uncertain. The changes in the mucous membranes often disappear, however, after amalgam fillings in contact with the lesion are removed. It has not been possible to establish any other connections between long-term exposure to mercury from amalgam in adults and effects on health. Population group studies have not shown any connection between amalgam fillings and symptoms or diseases. This does not, however, preclude connection in individual cases. A number of case reports where such connections are considered to exist have been published." In those European countries in which the question of health hazards from amalgam therapy has arisen, the health authorities have in general taken a similar view. REPRODUCTIVE TOXICOLOGICAL EFFECTS
Pre-natal exposure to potentially harmful drugs and chemicals is an emotionally charged issue a recurrent cause of public concern. It has been inferred that mercury vapor released from dental amalgam is a potential hazard to the fetus because it passes easily over the placenta (Greenwood et aL, 1972; SocialstyrelsensExpertgrupp, 1987). Although this assumption has been questioned (Strubelt etaL, 1988; Larsson and Sagulin, 1990; Larsson, 1991), it has once again attracted the attention of the mass media following publication of a Polish study (Sikorski et aL, 1987) on the outcome of pregnancies in dental personnel, and a recent Canadian study (Vimy et aL, 1990) showing placental transfer of mercury from amalgam fillings in pregnant ewes. This manuscript is published as part of the proceedings of the NIH Technology Assessment Conference on Effects and Sideeffects of Dental Restorative Materials, August 26-28,1991, National Institutes of Health, Bethesda, Maryland, and did not undergo the customary journal peer-review process.
114
Clinical-Epidemiological Reports
The Polish study by Sikorski et aL (1987) on adverse effects on reproduction has frequently been quoted as a key reference in the amalgam debate. The authors claimed a significant, positive association between total mercury concentration in
VOL.
6
115
TERATOLOGICAL ASPECTS
TABLE RELEVANT DATA Dental Personnel
Controls
Interview toxicological examination
81
34
At least one pregnancy before the examination
57
30
Reported pregnancies
117
63
Complicated pregnancies
28
7
19 women
5 women
Spontaneous abortions
19
not known
Stillbirths
3
not known
Malformed children
6
0
5 spina bifida Induced abortions
not known
not known
Compiled from Sikorski et al.9 1987. the hair of female dental personnel with occupational exposure to mercury and their history of reproductive failures. Information on reproductive histories was obtained by questionnaires, and scalp and pubic hair samples were analyzed for total mercury determinations. The relevant data are summarized in the Table. (See "Addendum", page 118.) The scientific merits of this paper warrant critical review. The base population of the women from which the 81 study subjects were randomly selected is not presented, nor are the criteria for inclusion in the control group of the 34 non-exposed women. Of the latter, 87% (30 women) had at least one pregnancy before the day of examination, in contrast to 70% (57 women) in the study group. Thus, the baseline data for the two groups were not comparable. In the group exposed to mercury, out of 117 pregnancies in 57 women, 28 pregnancies in 19 women led to reproductive failure: 19 spontaneous abortions, three stillbirths, and six malformed children. Remarkably, five of these infants had spina bifida. In the non-exposed control group, out of a total of 63 pregnancies in 30 women, no malformations were reported. In five women, there were seven pregnancies with "adverse outcome", but no further details are provided. The presentation of the outcome of pregnancies in the paper is incomplete. Even among cases with the greatest exposure, three out of nine (i.e., 30%) reproductive failures are listed as "not known". A more affirmative statement is needed. Moreover, there is no mention of control of report bias in the study or of induced abortions. Total mercury concentration in scalp and pubic hair was used for determination of the exposure factor, metallic mercury from working with dental amalgam. It is known, however, that hair is not a good indicator of exposure to metallic mercury vapor, due to several factors, including the possibility of external contamination (Environmental Health Criteria 118, 1991). The hair samples moreover represent only the status at registration and are not representative of previous pregnancies.
In the presentation on cases of excessively high mercury concentration in the hair, remarkably high concentrations in scalp hair—59.546 ug/g and 27.344 pg/g, respectively—are noted for one dental assistant and one dentist. The dental assistant also had an outstanding value of 18.166 ug/g in pubic hair. The dental assistant had two abortions and one malformed child, and the dentist had one abortion. The geometric mean for the combined two groups of dental personnel is 0.527 for scalp hair and 0.381 ug/g for pubic hair. Other sources of exposure, such as methylmercury in the diet, are not reported. It is well-known that Eastern Europe is heavily polluted by industrial chemicals, and there are reports in the mass media of a high incidence of adverse health effects on children. With respect to spina bifida, it should also be noted that the etiology is uncertain and that the epidemiology is characterized by peaks in incidence for which there is no satisfactory explanation (Borman and Cryer, 1990). Ericson and Kallen (1989) presented a Swedish study of pregnancy outcomes in female dentists, dental assistants, or dental technicians. They studied a total of 8157 infants born to these subjects in 1976 and 1982-86 with respect to perinatal survival, low birthweight, and malformations. The only deviant finding for these data compared with total births was a significantly lower perinatal death rate. It is of great interest to note that no increase in risk for spina bifida was seen—the upper 95% confidence limit for the risk was 2.1 with respect to neural tube defects. The authors also analyzed other Swedish registers and reported that, among mothers of 220 infants with spina bifida born in 1965-67, none was a dentist. In an analysis of hospitalized spontaneous abortions in female dental personnel in 1980-81, no significant deviations from expected values were found. In a small study of only 84 such pregnancies in 1964-65, no increase in spontaneous abortion rate was seen. However, only one infant was malformed (anencephaly), and both parents were dental technicians. The studies on female dental staff carried out by Heidam
ADV DENT RES SEPTEMBER 1992
LARSSON
116
(1984) and Brodsky et al (1985) revealed no increased abortion rate when compared with age-standardized controls. Studies on occupational hazards are reviewed in the WHO document Environmental Health Criteria 118 (1991) on inorganic mercury. Adverse effects—such as decreased libido and fertility in men, menstrual disturbances, and spontaneous abortion—have been studied in subjects occupationally exposed to metallic mercury vapor. The exposure was extremely high compared with that from amalgam fillings, and the results are considered inconclusive.
TERATOLOGICAL-BIOLOGICAL STUDIES The teratogenic effect is determined by four factors: the agent, the dose, the stage of embryonic development, and the genetic constitution of the embryo. The first two factors are of particular interest and warrant further comment. Agent Accidental cases and studies in laboratory animals have confirmed that methylmercury exposure during pregnancy can cause birth defects (Environmental Health Criteria 101, 1990). It should be emphasized that the mercury released from dental amalgam is mainly metallic mercury vapor. The following paragraph in the Environmental Health Criteria 118 (1991) might cause misunderstanding: "Several investigators have reported abortions following exposure to elemental mercury vapor or inorganic mercury compounds several days after implantation. There are also reports of decreased fetal weight and malformations." Of the six animal studies quoted, five dealt with mercuric acetate, chloride, or oxide. Passage of these compounds through the placenta is limited. In only one study (Steffek et al, 1987) were pregnant rats exposed to elemental mercury vapor. The dosages used in this teratology study will be commented upon below.
Dose Pharmacokinetic studies have demonstrated variability in drug response and have shown that the teratogenic effect is caused by peak exposures exceeding a threshold (Wilson, 1973; Gabrielsson and Larsson, 1990). Restriction of amalgam therapy during pregnancy has been advocated on the grounds that insertion or removal of amalgams causes an acute peak exposure to mercury vapor (Socialstyrelsens Expertgrupp, 1987). It has been suggested (Lorscheider and Vimy, 1990) that chronic low-dose exposure may be a potentially greater risk than peak exposure, but pharmacokinetic calculation on drug placental transfer does not support this speculation (Levy and Hayton, 1973). In the teratology study by Steffek et al. (1987), the dams were exposed to mercury concentrations of 100, 500, and 1000pg/m3ondays 1-20 or days 10-15 of gestation. Exposure at 500 pg/m3 during the organogenic period resulted in an increased frequency of resorption (five out of 41 fetuses). After exposure throughout the gestation period, two fetuses out of 84 examined were malformed. Both short- and longterm exposures to 1000 pg/m3 caused an increase in the frequency of resorptions (eight out of 71, and seven out of
28 fetuses, respectively). The dose of mercury vapor contributed from dental amalgam fillings, of the order of 5 pg/day, is very low compared with the doses in this study and is not likely to exceed the threshold necessary for a teratogenic effect to occur.
Disposition The concentration of the teratogen at the target tissue is determined not only by the dose and the rate of placental transfer but also by other factors, such as the distribution within the maternal organism, the affinity to the fetal liver and blood, the hematocrit value, and the passage through the ductus venosus. These factors might help to explain toxicological mechanisms and species differences, and must be considered if the results of animal experiments are to be extrapolated to human conditions. It should also be emphasized that, although measurable amounts of mercury may be incorporated into certain fetal organs, this is not per se evidence of a toxicological or teratological effect.
Mercury in Fetal and Maternal Blood As an indication of human fetal exposure, the mercury concentration in cord blood and maternal blood gives a F/Mratio slightly over or close to one (Tejning, 1968; Suzuki et al, 1971; Baglanefa/., 1974;WannagandSkjaer&sen, 1975; Pitkin et al., 1976; Hubermont et al., 1978; Lauwerys et al., 1978; KuhnertetaL, 1981; Kuntzetal., 1982; Spencer etal, 1988; Schramel^a/., 1988; Skerfving, 1988; Sikorski^a/., 1989; Klemann et al, 1990). The WHO Environmental Health Criteria 118 (\99\) on inorganic mercury cites one report of transfer of inhaled mercury vapor to the human fetus (Clarkson and Klipper, 1978): Two pregnant women were accidentally exposed to metallic mercury vapor. At the time of delivery, the mercury concentration in the infant blood was similar to that in the maternal blood. The observation that the F/M- ratio B-Hg in late gestation in the sheep exceeded one and was in fact as high as four has attracted a great deal of attention (Vimy etal, 1990). Twelve amalgam fillings containing radioactive mercury (203Hg) were placed in the teeth of five ewes on gestation day 112. The intra-oral air exposure was 44 pg8/Hg/m3 (range, 1398). In other species, such as the guinea pig (Yoshida et al, 1986) and the monkey (Smith etal, 1981), the ratio is found to be less than one after long-term exposure. The sheep has an epithelochorial placenta, with six layers of tissue. The hemochorial placenta—e.g., in the rat, guinea pig, monkey, and man—has only three layers separating fetal and maternal blood (Faber and Thornburg, 1983). The assumption by Vimy et al. (1990), that "on this histological basis alone one might expect a human fetus to receive a greater proportion of a given dose of dental amalgam Hg than would a sheep fetus", is refuted by the numerous clinical observations listed above. A high proportion of mercury bound to the fetal red blood cells might contribute to the high mercury concentration in fetal blood. This assumption is supported by the red-bloodcell-to-plasma ratio of mercury in the ewes of 0.44 and in the fetal lambs of 0.97, which might partly reflect a higher
VOL. 6
TERATOLOGICAL ASPECTS
hematocrit in the newborn than in adults (Vimy et al, 1990). In a study on dental personnel and unexposed controls, the erythrocyte-Hg F/M- ratio was greater than one, and the plasma F/M- ratio was slightly less than one (Wannag and Skjaer&sen, 1975). This distribution of mercury in the blood might reflect a contribution by methylmercury exposure (Suzuki et al, 1971; Wannag and Skjaer&sen, 1975). In order for fetal exposure to be estimated, mercury concentration has also been recorded in the placenta (Suzuki et al, 1971; Wannag and Skjaer&sen, 1975; Pitkin et al, 1976; Hubermont etal, 1978; Kuhnert et al, 1981; Sikorski et al, 1989) and the amnion (Wannag and Skjaer&sen, 1975; Suzuki
Department of Odontological Toxicology Karolinska Institutet Box 4064 S-141 04 Huddinge, Sweden Adv Dent Res 6:114-119, September, 1992
Abstract—The teratogenic effect is determined by four factors: (1) the agent, (2) the dose, (3) the stage of embryonic development, and (4) the genetic constitution of the embryo. The first two factors are of particular interest and warrant further comment. It should be emphasized that the mercury released from dental amalgam is mainly metallic mercury vapor. The dose of mercury vapor from dental amalgam fillings in the order of 5 pg/day is very low compared with the doses in a teratological study and is not likely to exceed the threshold necessary for a teratogenic effect to occur. The concentration of the teratogen at the target tissue is determined not only by the degree of placental transfer but also by other factors, such as the distribution within the maternal organism, the affinity to the fetal liver and blood, the hematocrit value, and the passage through the ductus venosus. These factors might help to explain toxicological mechanisms and species differences and have to be considered if the results of animal experiments are to be extrapolated to human conditions. Neither epidemiological data nor animal experimental data indicate that the release of metallic mercury vapor from dental amalgam therapy should cause teratogenic effects. A comparison with the incorporation of the fetotoxic methyl mercury might be justified.
T
his review will cover the scientific evidence for sideeffects of dental amalgam, with special reference to teratogenic and any other systemic effects. The question of systemic effects has been dealt with in two recent documents. The WHO Environmental Health Criteria 118, Inorganic Mercury (1991) concludes with the following statement: "Recently, there has been an intense debate on the safety of dental amalgams and claims have been made that mercury from amalgam may cause severe health hazards. Reports describing different types of symptoms and signs and the results of the few epidemiological studies produced are inconclusive." The Swedish National Board of Health and Welfare (1991) addressed physicians and dentists recently on the effects on health of mercury from amalgam: "Mercury can cause allergic contact dermatitis. However, this is rare. In some cases oral symptoms such as transient swelling, redness, erosion and socalled oral lichen have been observed. The etiology of oral lichen is uncertain. The changes in the mucous membranes often disappear, however, after amalgam fillings in contact with the lesion are removed. It has not been possible to establish any other connections between long-term exposure to mercury from amalgam in adults and effects on health. Population group studies have not shown any connection between amalgam fillings and symptoms or diseases. This does not, however, preclude connection in individual cases. A number of case reports where such connections are considered to exist have been published." In those European countries in which the question of health hazards from amalgam therapy has arisen, the health authorities have in general taken a similar view. REPRODUCTIVE TOXICOLOGICAL EFFECTS
Pre-natal exposure to potentially harmful drugs and chemicals is an emotionally charged issue a recurrent cause of public concern. It has been inferred that mercury vapor released from dental amalgam is a potential hazard to the fetus because it passes easily over the placenta (Greenwood et aL, 1972; SocialstyrelsensExpertgrupp, 1987). Although this assumption has been questioned (Strubelt etaL, 1988; Larsson and Sagulin, 1990; Larsson, 1991), it has once again attracted the attention of the mass media following publication of a Polish study (Sikorski et aL, 1987) on the outcome of pregnancies in dental personnel, and a recent Canadian study (Vimy et aL, 1990) showing placental transfer of mercury from amalgam fillings in pregnant ewes. This manuscript is published as part of the proceedings of the NIH Technology Assessment Conference on Effects and Sideeffects of Dental Restorative Materials, August 26-28,1991, National Institutes of Health, Bethesda, Maryland, and did not undergo the customary journal peer-review process.
114
Clinical-Epidemiological Reports
The Polish study by Sikorski et aL (1987) on adverse effects on reproduction has frequently been quoted as a key reference in the amalgam debate. The authors claimed a significant, positive association between total mercury concentration in
VOL.
6
115
TERATOLOGICAL ASPECTS
TABLE RELEVANT DATA Dental Personnel
Controls
Interview toxicological examination
81
34
At least one pregnancy before the examination
57
30
Reported pregnancies
117
63
Complicated pregnancies
28
7
19 women
5 women
Spontaneous abortions
19
not known
Stillbirths
3
not known
Malformed children
6
0
5 spina bifida Induced abortions
not known
not known
Compiled from Sikorski et al.9 1987. the hair of female dental personnel with occupational exposure to mercury and their history of reproductive failures. Information on reproductive histories was obtained by questionnaires, and scalp and pubic hair samples were analyzed for total mercury determinations. The relevant data are summarized in the Table. (See "Addendum", page 118.) The scientific merits of this paper warrant critical review. The base population of the women from which the 81 study subjects were randomly selected is not presented, nor are the criteria for inclusion in the control group of the 34 non-exposed women. Of the latter, 87% (30 women) had at least one pregnancy before the day of examination, in contrast to 70% (57 women) in the study group. Thus, the baseline data for the two groups were not comparable. In the group exposed to mercury, out of 117 pregnancies in 57 women, 28 pregnancies in 19 women led to reproductive failure: 19 spontaneous abortions, three stillbirths, and six malformed children. Remarkably, five of these infants had spina bifida. In the non-exposed control group, out of a total of 63 pregnancies in 30 women, no malformations were reported. In five women, there were seven pregnancies with "adverse outcome", but no further details are provided. The presentation of the outcome of pregnancies in the paper is incomplete. Even among cases with the greatest exposure, three out of nine (i.e., 30%) reproductive failures are listed as "not known". A more affirmative statement is needed. Moreover, there is no mention of control of report bias in the study or of induced abortions. Total mercury concentration in scalp and pubic hair was used for determination of the exposure factor, metallic mercury from working with dental amalgam. It is known, however, that hair is not a good indicator of exposure to metallic mercury vapor, due to several factors, including the possibility of external contamination (Environmental Health Criteria 118, 1991). The hair samples moreover represent only the status at registration and are not representative of previous pregnancies.
In the presentation on cases of excessively high mercury concentration in the hair, remarkably high concentrations in scalp hair—59.546 ug/g and 27.344 pg/g, respectively—are noted for one dental assistant and one dentist. The dental assistant also had an outstanding value of 18.166 ug/g in pubic hair. The dental assistant had two abortions and one malformed child, and the dentist had one abortion. The geometric mean for the combined two groups of dental personnel is 0.527 for scalp hair and 0.381 ug/g for pubic hair. Other sources of exposure, such as methylmercury in the diet, are not reported. It is well-known that Eastern Europe is heavily polluted by industrial chemicals, and there are reports in the mass media of a high incidence of adverse health effects on children. With respect to spina bifida, it should also be noted that the etiology is uncertain and that the epidemiology is characterized by peaks in incidence for which there is no satisfactory explanation (Borman and Cryer, 1990). Ericson and Kallen (1989) presented a Swedish study of pregnancy outcomes in female dentists, dental assistants, or dental technicians. They studied a total of 8157 infants born to these subjects in 1976 and 1982-86 with respect to perinatal survival, low birthweight, and malformations. The only deviant finding for these data compared with total births was a significantly lower perinatal death rate. It is of great interest to note that no increase in risk for spina bifida was seen—the upper 95% confidence limit for the risk was 2.1 with respect to neural tube defects. The authors also analyzed other Swedish registers and reported that, among mothers of 220 infants with spina bifida born in 1965-67, none was a dentist. In an analysis of hospitalized spontaneous abortions in female dental personnel in 1980-81, no significant deviations from expected values were found. In a small study of only 84 such pregnancies in 1964-65, no increase in spontaneous abortion rate was seen. However, only one infant was malformed (anencephaly), and both parents were dental technicians. The studies on female dental staff carried out by Heidam
ADV DENT RES SEPTEMBER 1992
LARSSON
116
(1984) and Brodsky et al (1985) revealed no increased abortion rate when compared with age-standardized controls. Studies on occupational hazards are reviewed in the WHO document Environmental Health Criteria 118 (1991) on inorganic mercury. Adverse effects—such as decreased libido and fertility in men, menstrual disturbances, and spontaneous abortion—have been studied in subjects occupationally exposed to metallic mercury vapor. The exposure was extremely high compared with that from amalgam fillings, and the results are considered inconclusive.
TERATOLOGICAL-BIOLOGICAL STUDIES The teratogenic effect is determined by four factors: the agent, the dose, the stage of embryonic development, and the genetic constitution of the embryo. The first two factors are of particular interest and warrant further comment. Agent Accidental cases and studies in laboratory animals have confirmed that methylmercury exposure during pregnancy can cause birth defects (Environmental Health Criteria 101, 1990). It should be emphasized that the mercury released from dental amalgam is mainly metallic mercury vapor. The following paragraph in the Environmental Health Criteria 118 (1991) might cause misunderstanding: "Several investigators have reported abortions following exposure to elemental mercury vapor or inorganic mercury compounds several days after implantation. There are also reports of decreased fetal weight and malformations." Of the six animal studies quoted, five dealt with mercuric acetate, chloride, or oxide. Passage of these compounds through the placenta is limited. In only one study (Steffek et al, 1987) were pregnant rats exposed to elemental mercury vapor. The dosages used in this teratology study will be commented upon below.
Dose Pharmacokinetic studies have demonstrated variability in drug response and have shown that the teratogenic effect is caused by peak exposures exceeding a threshold (Wilson, 1973; Gabrielsson and Larsson, 1990). Restriction of amalgam therapy during pregnancy has been advocated on the grounds that insertion or removal of amalgams causes an acute peak exposure to mercury vapor (Socialstyrelsens Expertgrupp, 1987). It has been suggested (Lorscheider and Vimy, 1990) that chronic low-dose exposure may be a potentially greater risk than peak exposure, but pharmacokinetic calculation on drug placental transfer does not support this speculation (Levy and Hayton, 1973). In the teratology study by Steffek et al. (1987), the dams were exposed to mercury concentrations of 100, 500, and 1000pg/m3ondays 1-20 or days 10-15 of gestation. Exposure at 500 pg/m3 during the organogenic period resulted in an increased frequency of resorption (five out of 41 fetuses). After exposure throughout the gestation period, two fetuses out of 84 examined were malformed. Both short- and longterm exposures to 1000 pg/m3 caused an increase in the frequency of resorptions (eight out of 71, and seven out of
28 fetuses, respectively). The dose of mercury vapor contributed from dental amalgam fillings, of the order of 5 pg/day, is very low compared with the doses in this study and is not likely to exceed the threshold necessary for a teratogenic effect to occur.
Disposition The concentration of the teratogen at the target tissue is determined not only by the dose and the rate of placental transfer but also by other factors, such as the distribution within the maternal organism, the affinity to the fetal liver and blood, the hematocrit value, and the passage through the ductus venosus. These factors might help to explain toxicological mechanisms and species differences, and must be considered if the results of animal experiments are to be extrapolated to human conditions. It should also be emphasized that, although measurable amounts of mercury may be incorporated into certain fetal organs, this is not per se evidence of a toxicological or teratological effect.
Mercury in Fetal and Maternal Blood As an indication of human fetal exposure, the mercury concentration in cord blood and maternal blood gives a F/Mratio slightly over or close to one (Tejning, 1968; Suzuki et al, 1971; Baglanefa/., 1974;WannagandSkjaer&sen, 1975; Pitkin et al., 1976; Hubermont et al., 1978; Lauwerys et al., 1978; KuhnertetaL, 1981; Kuntzetal., 1982; Spencer etal, 1988; Schramel^a/., 1988; Skerfving, 1988; Sikorski^a/., 1989; Klemann et al, 1990). The WHO Environmental Health Criteria 118 (\99\) on inorganic mercury cites one report of transfer of inhaled mercury vapor to the human fetus (Clarkson and Klipper, 1978): Two pregnant women were accidentally exposed to metallic mercury vapor. At the time of delivery, the mercury concentration in the infant blood was similar to that in the maternal blood. The observation that the F/M- ratio B-Hg in late gestation in the sheep exceeded one and was in fact as high as four has attracted a great deal of attention (Vimy etal, 1990). Twelve amalgam fillings containing radioactive mercury (203Hg) were placed in the teeth of five ewes on gestation day 112. The intra-oral air exposure was 44 pg8/Hg/m3 (range, 1398). In other species, such as the guinea pig (Yoshida et al, 1986) and the monkey (Smith etal, 1981), the ratio is found to be less than one after long-term exposure. The sheep has an epithelochorial placenta, with six layers of tissue. The hemochorial placenta—e.g., in the rat, guinea pig, monkey, and man—has only three layers separating fetal and maternal blood (Faber and Thornburg, 1983). The assumption by Vimy et al. (1990), that "on this histological basis alone one might expect a human fetus to receive a greater proportion of a given dose of dental amalgam Hg than would a sheep fetus", is refuted by the numerous clinical observations listed above. A high proportion of mercury bound to the fetal red blood cells might contribute to the high mercury concentration in fetal blood. This assumption is supported by the red-bloodcell-to-plasma ratio of mercury in the ewes of 0.44 and in the fetal lambs of 0.97, which might partly reflect a higher
VOL. 6
TERATOLOGICAL ASPECTS
hematocrit in the newborn than in adults (Vimy et al, 1990). In a study on dental personnel and unexposed controls, the erythrocyte-Hg F/M- ratio was greater than one, and the plasma F/M- ratio was slightly less than one (Wannag and Skjaer&sen, 1975). This distribution of mercury in the blood might reflect a contribution by methylmercury exposure (Suzuki et al, 1971; Wannag and Skjaer&sen, 1975). In order for fetal exposure to be estimated, mercury concentration has also been recorded in the placenta (Suzuki et al, 1971; Wannag and Skjaer&sen, 1975; Pitkin et al, 1976; Hubermont etal, 1978; Kuhnert et al, 1981; Sikorski et al, 1989) and the amnion (Wannag and Skjaer&sen, 1975; Suzuki
