Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
4 Animal Studies THE quality and relevance of animal studies that have evaluated the effects of perchlorate exposure have been debated. Accordingly, the committee was tasked with evaluation of those published and unpublished studies. In this chapter, the committee reviews the animal studies with particular attention to those critiqued and used to derive the reference dose in the 2002 U.S. Environmental Protection Agency (EPA) draft risk assessment. Because rats have been used as the primary animal model, the committee first compares thyroid function in rats and humans and then discusses animal studies that investigated the effects of perchlorate exposure on serum thyroid hormone concentrations, thyroid histopathology, brain mor- phometry, neurobehavior, and thyroid tumors. Physiologically based pharmacokinetic models are also reviewed briefly. Because concerns have been raised about the possibility of effects other than those resulting from altered thyroid function, the chapter concludes with a discussion of general toxicologic evaluations with emphasis on immunologic studies. The toxico- logic implications of perchlorateâs interaction with the sodium (Na+)/iodide (I2) symporter (NIS) of other tissues and organs are also discussed. COMPARISON OF THYROID FUNCTION IN RATS AND HUMANS The fundamental mechanisms involved in the function and regulation of the pituitary-hypothalamus-thyroid system in rats are qualitatively similar to those in humans (Bianco et al. 2002). However, the dynamics of the two systems differ substantially. The biochemical and physiologic differences between rats and humans related to thyroid function are de- scribed in the following paragraphs. 115
116 Health Implications of Perchlorate Ingestion Thyroid hormones in serum are extensively bound to plasma proteins. The proteins that bind thyroxine (T4) and triiodothyronine (T3) vary widely among species, and their binding affinities for the thyroid hormones also differ. In humans and other primates, thyroxine-binding globulin (TBG) is the principal protein that binds T4 (Dohler et al. 1979). It has a very high affinity for T4 : only about 0.03% of the T4 in serum is in the free unbound form (Hill et al. 1989). Binding sharply reduces clearance of T4 from serum. Rats do not have TBG, and most T4 in rat serum is bound to albumin and transthyretin. The binding affinity of T4 for TBG is more than a 100 times greater than that of albumin or transthyretin (Hill et al. 1989), and the difference contributes to the higher rate of T4 clearance in rats. The in- creased clearance contributes to the need for a higher rate of production of T4 per unit of body weight in rats to maintain normal concentrations of T4 (Dohler et al. 1979). The higher production rate is reflected in the histo- logic appearance of the rat thyroid, which has small thyroid follicles that contain much less colloid than those of primates (McClain 1995). Those features give the rat thyroid a more âfunctionally activeâ histologic appear- ance than that of primates, including humans. The follicular epithelium in rats is cuboidal; that of monkeys appears flattened in comparison. The change in the height of the follicular cells from flattened to cuboidal to columnar represents follicular-cell hypertrophy and is characteristic of the increased functional activity. There appear to be some differences in the metabolism of T4 by the liver between rats and humans. Some 50% of T4 is eliminated via bile in rats, but only 10-15% in humans (Hill et al. 1989). The difference does not reflect a qualitative difference in metabolism, because the major metabolite in bile (glucuronide conjugate) is the same in both species (Hard 1998). The biochemical and physiologic differences between rats and humans related to the thyroid affect their responses to goitrogens, such as perchlor- ate. For example, Yu et al. (2002) evaluated inhibition of radioiodide uptake by the thyroid in rats exposed to perchlorate in drinking water at 0, 1.0, 3.0, and 10.0 mg/kg of body weight for 1, 5, and 14 days. After 1 day of perchlorate administration, inhibition of iodide uptake was about 15%, 55%, and 65% at 1.0, 3.0, and 10 mg/kg, respectively. After 5 days, inhibi- tion of iodide uptake was 0, 10%, and 30%. After 14 days, inhibition of iodide uptake was observed only at 10 mg/kg. The data show that the initial inhibition of radioiodide uptake by perchlorate in rats is similar to that in humans. However, rats compensated for the inhibition within 5 days of perchlorate administration, most likely by increasing the expression of NIS in the thyroid. A similar response was not observed in a 14-day human study with perchlorate administration (Greer et al. 2002). The data suggest
Animal Studies 117 that compensation occurs more quickly in rats because rats have a smaller reserve capacity of thyroid hormones than humans. Another example of different responses to perchlorate is related to changes in serum concentrations of thyroid hormones and thyrotropin (thyroid-stimulating hormone, TSH). For example, Siglin et al. (2000) treated male and female rats with ammonium perchlorate at 0.01-10 mg/kg per day. At 14 days, serum T4 concentrations were significantly decreased at 10 mg/kg per day in both male and female rats. Serum T3 concentrations were significantly decreased in males at 0.01 mg/kg per day or higher. No significant decreases were observed in serum T3 concentrations at any dose in female rats. Serum TSH concentrations were significantly increased at 0.20 mg/kg per day or higher in males and at 0.05 mg/kg per day or higher in females. Studies in adult humans have not found increases in serum TSH or decreases in serum T4 or T3 with potassium perchlorate exposure over a similar period. For example, administration of 10 mg of potassium per- chlorate per day (0.1 mg/kg of perchlorate per day assuming a 70-kg hu- man) to healthy men for 14 days resulted in no changes in serum thyroid hormone or TSH concentrations during the exposure period (Lawrence et al. 2000). Similarly, Greer et al. (2002) found no decreases in serum thyroid hormones or increases in serum TSH in healthy men and women given perchlorate at up to 0.5 mg/kg per day for 14 days. There are important differences between rats and humans in pituitary- thyroid function during pregnancy. In humans, serum total T4 and T3 concentrations rise progressivelyâon the average, about 50% during the first trimester of pregnancyâand remain increased during the remainder of pregnancy (Glinoer 1997). That response is due to an increase in serum TBG, which is stimulated by an increase in estrogen production (see Chap- ter 2). During the first trimester, serum free T4 and T3 concentrations also increase slightly because of stimulation of the thyroid gland by chorionic gonadotropin, a hormone produced by the placenta. The primary action of chorionic gonadotropin is to sustain pregnancy, but it also has weak thyroid-stimulating activity. The increases in serum free T4 and T3 concen- trations decrease TSH secretion slightly in pregnant women. Later in pregnancy, the decrease in production of chorionic gonadotropin results in a return of serum free T4, T3, and TSH to concentrations comparable with those in nonpregnant women and in men. Thyroid hormone concentrations change during pregnancy in rats, but detailed studies are limited to gestation days 17-22. During gestation days 17-22, serum T4 concentrations in pregnant rats are significantly lower than those in nonpregnant female rats (Fukuda et al. 1980; Calvo et al. 1990;
118 Health Implications of Perchlorate Ingestion Versloot et al. 1994). The differences in serum thyroid hormone concentra- tions between pregnant rats and women are related largely to the lack of TBG and absence of chorionic gonadotropin production in rats. There are also differences between rats and humans in the timing of development of thyroid function. Thyroid function in rats at birth is rela- tively immature, equivalent to that of a third-trimester human fetus. The human fetus is protected by maternal thyroid hormone for a longer period of development. In rats, serum total T4 and T3 concentrations increase between postnatal days 5 and 15 in association with an increase in a serum thyroid hormone-binding globulin (Obregon et al. 1991). Thereafter, the production of that binding protein decreases, and therefore, serum T4 and T3 concentrations fall (Vranckx et al. 1994). Thus, thyroid function and regulation are qualitatively similar in rats and humans, but important differences in serum thyroid hormone binding and clearance rates and thyroid stimulation by a placental hormone in pregnant women lead to important quantitative differences between the two species. The species differences must be carefully considered in interpret- ing serum thyroid hormone, TSH, and thyroid histopathology data in studies that use rats to assess human health risk associated with perchlorate exposure. THYROID HORMONES AND THYROID HISTOPATHOLOGY The committee reviewed published literature and laboratory study reports on the effects of perchlorate exposure on thyroid hormones, TSH, and thyroid histopathology in animals. The discussion here focuses on the Argus (2001) study because it provides a comprehensive evaluation of those entities in the most sensitive populationsâpregnant females, fetuses, and neonates. The findings of the Argus study are generally consistent with those of previous studies. Inconsistencies are most likely due to small sample sizes that were used in some of the other studies or to differences in study design, such as use of animals at different stages of development. In Argus (2001), female rats (dams) were exposed to perchlorate through gestation and lactation. Ammonium perchlorate was administered in drinking water at concentrations that provided doses of 0, 0.01, 0.1, 1.0, and 30 mg/kg per day. Administration began 2 weeks before mating and extended through postnatal day 22. The offspring were exposed to perchlor- ate in utero, through their motherâs milk, and through any consumption of the perchlorate-containing drinking water provided to their mothers. Serum thyroid hormones and TSH were measured on gestation day 21 in the dams
Animal Studies 119 and fetuses, on postnatal days 10 and 22 in the dams, and on postnatal days 5, 10, and 22 in the pups. Histologic thyroid evaluations were conducted on the same schedule as the hormone analyses. Standard toxicologic, reproductive, and developmental end points also were evaluated. Findings regarding neurodevelopmental measures are discussed in the next section. Changes in thyroid hormones and TSH are illustrated in Figures 4-1 and 4-2. Statistically significant differences between control and perchlorate dose groups as determined by Argus (2001) are indicated in the figures. The changes were consistent with those of previous rat studies and with perchlorateâs known mode of action in inhibiting the NIS. There were dose-related increases in serum TSH and dose-related decreases in serum total T4 and T3 in the dams, fetuses, and pups. Overall, the dams appeared to be more sensitive to perchlorate administration than the fetuses or pups, and the most dramatic changes in the dams were observed on gestation day 21. Although a downward trend was observed in serum T3 in the dams, it appeared to be a less sensitive marker than T4. Serum T3 was decreased significantly only during gestation and only at the highest dose (30 mg/kg per day), whereas serum T4 was decreased significantly at all doses during gestation and at the highest dose on postnatal days 10 and 22. An important point is that the serum T4 levels of control rats on gestation day 21 were substantially lower than those of female rats on postnatal days 10 and 22. Those data are consistent with the literature, as discussed previously, and suggest a low thyroid hormone reserve in pregnant rats. Another way to evaluate the data is to calculate the percentage change from the control value. Figure 4-3 illustrates the changes from controls in serum TSH, T4, and T3 in dams, fetuses, and pups at the different evaluation times. Several points should be noted. First, the greatest changes from control values were observed in TSH in the dams, particularly in the highest-dose group. For example, on gestation day 21, when the greatest changes were observed, TSH was increased by 35% at 0.01 mg/kg per day, 50% at 0.1 mg/kg per day, 64% at 1.0 mg/kg per day, and 146% at 30 mg/kg per day. Second, T4 and T3 in the dams differed by only about 10% or less from that in controls at the different evaluation times, with the notable exception of T4 on gestation day 21, which was decreased by 11% at 0.01 mg/kg per day, 45% at 0.1 mg/kg per day, 48% at 1.0 mg/kg per day, and 54% at 30 mg/kg per day. Third, presentation of the data as percentage changes from the control more clearly illustrates the greater sensitivity of T3 in the pups than in the dams and the greater sensitivity of perchlorate-associated decreases in serum T3 than in T4 in the pups. Fourth, the data from postnatal day 22 indicate that males have a greater reduction than females in total T4 and T3 compared with controls.
120 Health Implications of Perchlorate Ingestion 5 Control 4.5 0.01 mg/kg-d * 4 0.1 mg/kg-d ** 3.5 1 mg/kg-d 30 mg/kg-d 3 T4 (ug/dL) 2.5 *** 2 *** *** 1.5 *** 1 0.5 0 GD 21 PND 10 PND 22 Day of Sample 140 Control 0.01 mg/kg-d 0.1 mg/kg-d 120 1 mg/kg-d 30 mg/kg-d 100 ** 80 T3 (ng/dL) 60 40 20 0 GD 21 PND 10 PND 22 Day of Sample Control 18 *** 0.01 mg/kg-d 16 0.1 mg/kg-d 14 1 mg/kg-d *** 30 mg/kg-d *** 12 *** *** TSH (ng/mL) 10 *** *** *** *** *** 8 ** 6 4 2 0 GD 21 PND 10 PND 22 Day of Sample
Animal Studies 121 Histologic evaluations of the thyroid gland were conducted at the same times as those of thyroid hormones and TSH. Absolute and relative thyroid weights were significantly increased in the dams at 30 mg/kg per day at all evaluation times. A similar trend was noted in absolute thyroid weight in male and female pups over the course of the study. Statistically significant increases in absolute thyroid weights also were observed in all groups of male pups on postnatal day 10 and in females at 1.0 mg/kg per day on postnatal day 22. Histologic examination of the thyroid gland revealed colloid depletion, follicular-cell hypertrophy, and follicular-cell hyperplasia in the dams (see Table 4-1). Those effects were mainly restricted to the highest-dose group (30 mg/kg per day), although colloid depletion and follicular-cell hyperplasia were increased at 1.0 mg/kg per day on postnatal days 10 and 22, respectively. The predominant effect in the fetuses and pups was colloid depletion (see Table 4-2). Colloid depletion was present primarily in the highest-dose group (30 mg/kg per day) but was somewhat increased on several evaluation days in the 1.0-mg/kg group. Follicular-cell hyperplasia was noted occasionally in a few pups, but no clear dose-re- sponse trends were noted. The committee has concerns about the reliability of the thyroid histo- pathology data, particularly those on the dams. For example, hyperplasia was observed on postnatal day 22 at the lower doses in the absence of hypertrophy, which typically does not occur in rats. The data at the highest dose appeared to be more reliable inasmuch as the expected TSH-mediated morphologic changes in the thyroid were observed: colloid depletion, follicular-cell hypertrophy, and follicular-cell hyperplasia. The histologic data on the fetuses and pups were more consistent. Colloid depletion of thyroid follicles, although a subjective morphologic end point, was the most consistent histologic finding in rat fetuses and pups on all evaluation days. It was observed consistently in the highest-dose animals and, to a smaller extent, in the 1.0-mg/kg group. The thyroid morphology of the two lower-dose groups of animals (0.01 and 0.1 mg/kg per day) was similar to that of control animals. FIGURE 4-1 (Top) Serum thyroxine (T4) concentrations in dams treated with ammonium perchlorate at indicated doses in drinking water. (Middle) Serum triiodothyronine (T3) concentrations in same animals. (Bottom) Serum thyroid- stimulating hormone (TSH) in same animals. Values presented as mean Â± SD of 14-16 animals (data from Argus 2001). Abbreviations: GD, day of gestation; PND, postnatal day; *, significantly different from control, p # 0.05; **, significantly different from control, p # 0.01; *** significantly different from control, p # 0.001.
5 Control 250 Control 122 0.01 mg/kg-d 0.01 mg/kg-d 4.5 0.1 mg/kg-d *** 0.1 mg/kg-d * ** 4 200 *** 1 mg/kg-d 1 mg/kg-d *** *** * 3.5 30 mg/kg-d 30 mg/kg-d 3 150 2.5 ** ** T4 (ug/dL) T3 (ng/dL) 2 100 ** *** *** 1.5 * * 1 50 ** *** 0.5 *** *** 0 0 GD 21 PND 5 PND 10 PND 22 M PND 22 F GD 21 PND 5 PND 10 PND 22 M PND 22 F Day of Sam ple and Sex for PND 22 Day of Sam ple and Sex for PND 22 14 Control 0.01 mg/kg-d *** 12 0.1 mg/kg-d 1 mg/kg-d *** 10 30 mg/kg-d FIGURE 4-2 (Upper Left) Serum thyroxine (T4) concentrations ** *** in fetus and pups of dams treated with ammonium perchlorate at 8 indicated doses in drinking water. (Upper Right) Serum triiodo- *** *** thyronine (T3) concentrations in same animals. (Lower Left) Serum *** 6 *** TSH (ng/mL) *** thyroid-stimulating hormone (TSH) in same animals. Values re- ** ported as the mean Â± SD of 11-17 animals except for T3 measures 4 at day 21 of gestation, when two to eight animals were used to 2 derive values (data from Argus 2001). F, female; GD, gestation day; M, male; PND, postnatal day; *, significantly different from 0 control, p # 0.05; **, significantly different from control, p # 0.01; GD 21 PND 5 PND 10 PND 22 M PND 22 F Day of Sample and Sex at PND 22 ***, significantly different from control, p # 0.001.
Animal Studies 123 TABLE 4-1 Thyroid Histopathology in Control Dams and Dams Given Four Doses of Perchlorate Dose (mg/kg per day) Evaluation Day and Effect 0 0.01 0.1 1.0 30.0 Gestation day 21 Colloid depletion 0/16 0/16 0/16 0/15 16/16 Follicular-cell hypertrophy 0/16 0/16 0/16 0/15 14/16 Follicular-cell hyperplasia 0/16 0/16 0/16 0/15 2/16 Postnatal day 10 Colloid depletion 0/16 1/16 1/16 5/16 16/16 Follicular-cell hypertrophy 0/16 0/16 1/16 0/16 16/16 Follicular-cell hyperplasia 0/16 0/16 1/16 2/16 9/16 Postnatal day 22 Colloid depletion 0/16 0/16 1/15 1/16 16/16 Follicular-cell hypertrophy 0/16 0/16 0/15 0/16 14/16 Follicular-cell hyperplasia 3/16 4/16 5/15 10/16 10/16 Source: Data from Argus 2001. TABLE 4-2 Thyroid Histopathology in Control and Perchlorate- Exposed Fetuses and Pups Dose (mg/kg per day) Evaluation Day and Effect 0 0.01 0.1 1.0 30.0 Gestation day 21âFetuses Colloid depletionâmales 0/16 2/16 0/16 12/16 16/16 Colloid depletionâfemales 0/16 1/16 1/16 13/16 16/16 Postnatal day 5âPups Colloid depletionâmales 0/16 2/16 0/16 4/16 16/16 Colloid depletionâfemales 0/16 0/16 0/16 6/16 16/16 Postnatal day 10âPups Colloid depletionâmales 0/16 1/16 1/16 1/16 16/16 Colloid depletionâfemales 0/16 0/16 1/16 4/16 15/15 Postnatal day 22âPups Colloid depletionâmales 0/16 0/16 0/15 0/16 11/16 Colloid depletionâfemales 0/16 0/16 0/15 0/15 12/16 Source: Data from Argus 2001.
0 0 a b 124 -10 -10 -20 -20 0.01 mg/kg-d 0.01 mg/kg-d -30 0.1 mg/kg-d -30 0.1 mg/kg-d 1 mg/kg-d 1 mg/kg-d 30 mg/kg-d -40 30 mg/kg-d Percent change from control -40 Percent change from control -50 -50 T4, Fetus/Pup T4, Dams -60 -60 GD 21 PND 5 PND 10 PND 22 M PND 22 F GD 21 PND 10 PND 22 Day of Sam ple and Sex for PND 22 Day of Sample 0 c 0 d -5 -5 -10 -10 -15 -15 0.01 mg/kg-d -20 0.01 mg/kg-d 0.1 mg/kg-d -20 1 mg/kg-d Percent change from control 0.1 mg/kg-d -25 30 mg/kg-d Percent change from control 1 mg/kg-d -25 30 mg/kg-d -30 -30 T3, Fetus/Pup -35 T3, Dams GD 21 PND 5 PND 10 PND 22 M PND 22 F -35 GD 21 PND 10 PND 22 Day of Sam ple and Sex for PND 22 Day of Sample
150 150 e f 0.01 mg/kg-d TSH, Dams TSH, Fetus/Pup 130 130 0.1 mg/kg-d 0.01 mg/kg-d 1 mg/kg-d 110 110 0.1 mg/kg-d 30 mg/kg-d 1 mg/kg-d 90 90 30 mg/kg-d 70 70 50 50 Percent change from contro Percent change from control 30 30 10 10 -10 -10 GD 21 PND 10 PND 22 GD 21 PND 5 PND 10 PND 22 M PND 22 F Day of Sample Day of Sample and Sex for PND 22 FIGURE 4-3 Changes in serum T4, T3, and TSH in dams (a, c, e) and fetuses and pups (b, d, f) presented as percent change from control (data for calculations were those of Argus 2001). Abbreviations: F, female; GD, gestation day; M, male; PND, postnatal day. 125
126 Health Implications of Perchlorate Ingestion Whether changes of the magnitude observed in the Argus (2001) study can cause adverse effects in the animals, particularly the offspring, is the critical issue. EPA (2002a) provided no discussion of the probable implications of the changes observed. The pregnant dams had decreases greater than 10% in serum T4 at 0.1 mg/kg per day and above. However, it is important to emphasize that the variations in serum T4 in the control animals at different stages were larger than the changes induced by perchlorate. Serum TSH was substantially increased in the pregnant dams at all doses. Although statistically significant changes in the thyroid hormones and TSH were noted in the fetuses and pups, the changes tended to be more modest than those in the dams. BRAIN MORPHOMETRY Because thyroid hormones play a critical role in brain development, two Argus (1998, 2001) studies included measures of brain morphometry. In the 1998 study, female Sprague-Dawley rats were exposed to ammonium perchlorate at 0, 0.1, 1.0, 3.0 or 10.0 mg/kg per day in the drinking water beginning on gestation day 0 and continuing until postnatal day 10. One male and one female pup from each of six control and six high-dose litters were sacrificed for brain morphometry on postnatal day 10-12 and postnatal day 82-85. The brains taken on day 10-12 were immersion-fixed in 10% neutral buffered formalin, embedded in paraffin, and sectioned coronally. Linear measurements were taken from three sections. A section through the optic chiasm was used to measure the thickness of the frontal and parietal cortex, the diagonal width of the striatum, and the thickness of the corpus callosum. The thickness of the hippocampal gyrus was measured in a section taken just posterior to the mammillary body. A section taken just anterior to the midpoint of the cerebellum was used to measure the height of the cerebellum at the level of the deep cerebellar nuclei, including lobes 1-6 and extending from the roof of the fourth ventricle to the dorsal surface. Initially, only the brains from the 0- and 10.0-mg/kg groups were sectioned and measured. The only statistically significant difference reported by Argus for the day 10-12 brains was an increase in the thickness of the corpus callosum in the female pups. An increase of similar magnitude was observed in the male pups, but the data on the male pups were more variable, and the difference was not statistically significant. After reviewing the data, EPA requested that sections from the 3.0-mg/kg group be assessed. Apparently, brains from the 0.1- and 1.0-mg/kg groups were never assessed. EPA then performed statistical analyses to compare
Animal Studies 127 the 0-, 3.0-, and 10.0-mg/kg groups. EPA reported a significant increase in the thickness of the corpus callosum in the 10.0-mg/kg group, significant decreases in the thickness of the hippocampal gyrus and striatum in the 3.0- mg/kg group, and significant increases in the thickness of the anterior and posterior cerebellum in the 3.0-mg/kg group. No alterations in the thickness of the corpus callosum were observed at 3.0 mg/kg per day. It does not appear that those effects were sex-specific, although the EPA report is not clear on this point (EPA 2002a; p. 5-37). The brains from control and 10-mg/kg group on postnatal day 82-85 were perfusion-fixed in 10% neutral buffered formalin and later sectioned and measured according to the same method used with the postnatal day 10- 12 brains. Statistically significant increases in the thickness of the corpus callosum and frontal cortex and in the cross-sectional width of the striatum were observed in the exposed males, but not in the females. Apparently, brains from the 3.0-mg/kg group were not assessed on postnatal day 82-85. Significant changes in brain morphometry at 10-12 and 82-85 days of age are summarized in Table 4-3. The neuromorphologic findings from Argus (1998) are difficult to interpret because of the absence of a dose-effect relationship and the inconsistency of effects across age and sex. A possible exception is the increase in the thickness of the corpus callosum, which was present in the highest-dose male rats (10 mg/kg) at both ages, although the difference did not reach statistical significance on postnatal day 10-12. Because of the difficulties in interpreting the data from the 1998 study, Argus (2001) conducted additional neuromorphometric analyses. Perchlo- rate exposure of female Sprague-Dawley rats started 2 weeks before cohabitation with male rats and continued until they were sacrificed. Doses of 0, 0.01, 0.1, 1.0, and 30.0 mg/kg per day were used, and pups were sacrificed for brain morphometry on postnatal days 10 and 22. The brains were immersion-fixed, embedded, and sectioned with the same methods used in the 1998 study. Argus (2001) reported a number of statistically significant differences between treatment groups, including increases in the thickness of the corpus callosum of male rats on postnatal days 10 and 22, finding that are consistent with those in the earlier study. However, there was not a clear dose-response relationship. In the 1998 study, an increase in the thickness of the corpus callosum was seen at 10 mg/kg per day but not at 3 mg/kg. In contrast, in the 2001 study, increases in corpus callosum thickness were seen at 0.1 and 1.0 mg/kg per day but not at 30 mg/kg. There is some confusion as to how the data were statistically analyzed by Argus. EPA states that the differences were determined by performing multiple t tests (EPA 2002a; p. 5-65). However, the Argus report (Argus
TABLE 4-3 Summary of Morphometric Findings in Rat Pups Exposed to Perchlorate a 128 Neuroanatomic Argus 1998 Argus 2001 EPA 2003 Region Day 10-12 Day 82-85 Day 10 Day 22 Day 22 Frontal cortex No change Increase at high dose Increase at low or Increase at high dose only Not measured only in males intermediate doses but not in males at high dose Parietal cortex No change No change Increase at low or No change Not measured intermediate doses but not at high dose Striatum Decrease at low or Increase at high dose Increase at low or Decrease at low or All dose groups intermediate doses only in males intermediate doses but not intermediate doses but not increased but not at high dose at high dose at high dose Corpus callosum Increase at high dose Increase at high dose Increase at low or Increase at low or Increase at low only only in males intermediate doses but not intermediate doses but not or intermediate at high dose in males; at high dose doses but not at Decrease at low or high dose intermediate doses but not at high dose in females Hippocampus Decrease at low or No change No change No change Not measured gyrus intermediate doses but not at high dose Dentate gyrus Not measured Not measured No change No change Not measured CA1 portion Not measured Not measured Decrease at low or No change Not measured intermediate doses but not at high dose in females; Increase at low or intermediate doses but not at high dose in males
CA3 portion Not measured Not measured No change Decrease at low or Not measured intermediate doses but not at high dose External germinal Not measured Not measured Decrease at low or No change Not measured layer of intermediate doses but not cerebellum at high dose Anterior/posterior Increase at low or No change No change Increase at low or Not measured cerebellum intermediate doses intermediate doses but not but not at high dose at high dose a Doses assessed at different evaluation points in Argus (1998) â day 10-12 (0, 3, and 10 mg/kg per day); day 82-85 (0 and 10 mg/kg per day) Argus (2001) â day 10 (all dose groups); day 22 (all dose groups). 129
130 Health Implications of Perchlorate Ingestion 2001, pp. 43-44 and p. 820) states that one-way ANOVAs were performed and were followed by Dunnettâs comparisons of each treatment group with the control group when the overall ANOVA was significant. In either case, a large number of statistical comparisons were run, so Type I error is a concern. EPA later reanalyzed the same data using two statistical approaches. In the first, it used a multivariate approach known as profile analysis. Briefly, profile analyses make comparisons between groups by creating a vector composed of all the measurements taken from an individual animal. The primary testâor test of parallelism of vectorsâestablishes whether the pattern of results across brain regions differs between treatment groups. Data from the right and left hemispheres were averaged for those analyses. Male and female pups were included in the same analysis with sex nested within a litter, but separate profile analyses were run on the data from postnatal days 10 and 22. In the second approach, univariate ANOVAs and Dunnettâs comparisons were used to determine which brain regions and dose groups differed from controls. That approach was similar to the one used by Argus, except that data on the two hemispheres were averaged rather than analyzed separately and the litter was considered the unit of variance with sex as a nested variable within a litter. The profile analysis indicated a significantly different pattern of results across treatment groups. The finding was due primarily to changes in the perchlorate-exposed groups compared with the control group in thickness of the striatum (decrease), posterior corpus callosum (increase), and cere- bellum (increase). For several of the observed effects, the dose-effect relationship was an inverted U-shaped function with the low or intermediate dose groups showing the largest deviations from controls (see EPA 2002a, Figure 5-15, p. 5-69). The results of the univariate analysis of the same data are summarized in Table 4-3 (see column labeled Argus 2001, Day 22). That analysis revealed a pattern of effects similar to that revealed by the profile analysis. However, several additional significant findings were observed, including an inverted U-shaped dose-effect function in region CA3 of the hippocampus and an increase in the thickness of the frontal cortex that was present mainly in the male pups. A number of outside experts have reviewed the Argus studies and raised substantial questions and concerns regarding the study design, the method- ology, and the biologic plausibility of the findings. Methodologic problems that have been discussed include method of fixation, time since fixation, variation in the plane of section, and the conduct of some measurements in a nonblinded fashion. Various reviewers have also pointed out the relative insensitivity of linear measurements of thickness relative to mea-
Animal Studies 131 surements of the area or volume of a structure and have raised questions about the appropriateness of using coronal sections to measure the corpus callosum. Finally, reviewers have raised concerns about the lack of a clear dose-response relationship and questioned whether reductions in thyroid hormones that were not large enough to reduce the growth rate or overall brain weight of the pups would be large enough to produce gross alterations in neurodevelopment. The Argus studies also have been criticized for using immersion fixa- tion rather than perfusion to fix pup brains for sectioning (TERA 2001; Elberger 2003). There does not appear to be a clear consensus among experts on that point. Harry (2001) notes that immersion fixation is appro- priate for brains of young pups because the decreased blood volume in the pup can result in less than optimal perfusion. However, Juraska (J. Juraska, University of Illinois, Champaign, personal commun., 2004) states that although that is true for pups under 5 days old, perfusion would be the most appropriate method of fixation for brains of 10- or 22-day-old pups. TERA (2001) notes that with immersion fixation distortion of the tissue can occur as the brain fixes from the outside in. Because it takes longer, immersion fixation can also lead to degeneration of cellular components. That makes the brain more susceptible to shrinkage and swelling. TERA (2001) notes that brains fixed with any method shrink during the first several weeks in fixative and swell later. Therefore, it is critical that the time between fixa- tion and sectioning be held constant across treatment groups in a study, particularly if immersion fixation is used. All tissues from the Argus studies were embedded in paraffin at the same time, but tissues from some groups were held longer than others before being sectioned and stained. Because all tissues from each Argus study were embedded in paraffin at the same time, differences in how long the tissues were held before sectioning should not be an issue. In addition to questions about fixation of the brains, reviewers have noted what appears to be considerable variability in the plane of sec- tionâboth between animals and between hemispheres in the same ani- malâin both Argus studies (Harry 2001; Wahlsten 2002). Variability, if equally distributed across dose groups, will increase Type II error, making it more difficult to detect treatment-related effects, but it should not increase the likelihood of spurious effects (Type I error). No evidence has been presented to suggest that the anterior-posterior coordinates of the sections varied systematically by dose group in the 1998 study. However, both Harry (2001) and Wahlsten (2002) have reviewed brain sections from the 2001 study in detail and have concluded that differences in the plane of section varied systematically across groups in that study. Wahlsten (2002)
132 Health Implications of Perchlorate Ingestion has noted (1) that brains with a thicker corpus callosum appear to have been sectioned more posteriorly near the splenium of the corpus callosum, which is its thickest part, and (2) that there were more of these posterior ânear spleniumâ sections in the three low-dose groups. Harry (2001) reviewed the sections used to measure the thickness of the corpus callosum with refer- ence to a stereotaxic atlas of the rat brain and noted that although only 10% of the sections from the control group were taken above anterior-posterior coordinate 3.5, 23% and 56% of the sections from the 0.1- and 1.0-mg/kg groups were cut at anterior-posterior coordinates above 3.5. Those were the groups that showed increases in corpus callosum thickness relative to the control group. Harry (2001) has noted that the differences in the plane of section could also influence measurements of the striatum and frontal cortex. A further criticism of Argus (2001) is that morphometric measurements of the sections from the three low-dose groupsâwhich were sectioned and measured later than the control and high-dose groupsâwere not conducted in a blinded fashion. The committee shares that concern. It is important to distinguish between comparative-morphology studies and morphometric studies. Comparative morphology yields qualitative assessments of the appearance of tissues; specifically, tissues of control animals are compared with those of exposed animals to determine whether specific pathologic changes are present in the exposed group. Morphometric studies yield quantitative measures of linear thickness of tissues. In Argus (2001), linear thickness of various brain regions was measured. Although it may some- times be appropriate to collect data in a nonblinded fashion in a com- parative-morphology study, there is no defensible reason for collecting data in a nonblinded fashion in a morphometric study. To address the concern about differences in the plane of section across animals, EPA contracted with Consultants in Veterinary Pathology, Inc. in 2003 to resection and remeasure brains from the postnatal day 22 pups from Argus (2001). The new sections were matched to anatomic atlas plates to ensure consistency in the plane of section among animals. Only the striatum and anterior and posterior corpus callosum were measured in the follow-up study. Because there was no controversy over plane of section for the cerebellum, measurements of the cerebellum from the original Argus (2001) dataset were included in the statistical analysis. The data were subjected to a profile analysis similar to that performed on the original 2001 dataset. The analysis was limited to a subset of the original 16 animals per group for which tissue was available for the resectioning in 2003. Measurements from 15 control animals, nine 0.01-mg/kg animals, four 0.1-mg/kg animals, 10 1.0-mg/kg animals, and 11 30.0-mg/kg animals were available. The 0.1-
Animal Studies 133 mg/kg group was excluded from the statistical analyses because fewer than the minimum of six animals required by the EPA Developmental Neuro- toxicity Testing Guidelines were available (EPA 1998a). As in the original dataset, the profile analysis was significant and indicated that the pattern differed among treatment groups. All three of the brain regionsâcerebellum, striatum, and corpus callosumâwere thicker in the treated groups than in the control group (Geller 2003, Figure 3). When the older data from the cerebellum were excluded from the analysis, the treatment effect remained significant. That corpus callosum thickness was increased at multiple doses and at several stages of development (postnatal days 10-12, 22, and 82-85) in two studies is suggestive of a relationship between perchlorate exposure and altered neurodevelopment. EPAâs analysis of additional tissue from Argus (2001) animals helps to dispel some of the concerns that have been raised about systematic differences in the plane of section among treatment groups. However, it also has contributed to the inconsistencies in the dataset because the thickness of the striatum was decreased in the 2001 brain sections and increased in the 2003 brain sections from the same pups. An important issue that several critiques of the Argus studies (e.g., TERA 2001) have raised is that linear measurement of the thickness of brain areas is not the most sensitive method of detecting alterations in neural structure. Measurement of volume or area would be more sensitive and more accurate, particularly for structures, such as the corpus callosum and hippocampus, that change shape across serial coronal sections. For those structures, linear thickness will depend on the location of the section, and small changes in the plane of section could have a large effect on the results. Several experts agree that a more appropriate method of assessing changes in the size of the corpus callosum would be to measure the area of the structure in a midsagittal section through the brain (TERA 2001; Wahlsten 2002; Elberger 2003). However, it is important to note that all measurements of size, including measurements of volume or area, are no more than a surrogate for changes in the cellular structure of a brain region, and it is ultimately the underlying changes that are important to understand. Laboratory rats have been used extensively as a model to study the effects of thyroid hormone deficiency or excess on brain development, and there is a large literature on the neuroanatomic effects of neonatal hypothy- roidism in rats that should be used to inform the design of future studies of perchlorate. In particular, the laminar organization of layered struc- turesâincluding the cerebellum, hippocampus, and cortexâis disrupted by neonatal hypothyroidism (see, for example, Alvarez-Dolado et al. 1999). Neonatal hypothyroidism appears to alter the development of those struc-
134 Health Implications of Perchlorate Ingestion tures by reducing the expression of reelin, a protein that is involved in the radial migration of neurons (Alvarez-Dolado et al. 1999). That results in aberrant migration and blurring of otherwise clearly defined layers in those structures. Another classic sign of neonatal hypothyroidism is stunting of the dendritic trees of pyramidal cells in the hippocampus (Rami et al. 1986) and of Purkinje cells in the cerebellum (Legrand 1986). The changes can be dramatic and are easily visualized with Golgi stain (Rami et al. 1986). Finally, disappearance of the external granular layer of the cerebellum is delayed in neonatal hypothyroidism (Lauder 1977). Analysis of some of those classic signs of neonatal hypothyroidism in perchlorate-exposed animals will be important in establishing whether a pattern of neuromor- phologic effects consistent with neonatal hypothyroidism exists. Alternative approaches could also be useful. A number of thyroid hormone-responsive genes have been identified in the brain (Zoeller et al. 2002). They are differentially expressed in different brain regions at differ- ent stages of development. For example, neurotrophin and brain-derived neurotrophic factor gene expression are altered by thyroid status (Koibuchi et al. 2001). Analyses of the effects of perchlorate exposure on the expres- sion of those genes could provide additional information about whether the developing brain responds to perchlorate exposure in a way that is consis- tent with the response to other manipulations that result in low thyroid hormone concentrations. In its draft risk assessment (EPA 2002a), EPA summarizes all the morphologic data from the Argus studies but focuses on the effect of perchlorate on the corpus callosum. Several reviewers stated that perchlo- rate is believed to exert its effects by reducing the production of thyroid hormones, and the corpus callosum is not known to be a target of thyroid hormone. However, although the corpus callosum is not widely recognized as sensitive to thyroid hormones, several studies indicate that maturation of the corpus callosum is under the influence of thyroid hormones. The studies suggest that hypothyroidism may arrest callosal axons at an imma- ture stage of development (Gravel and Hawkes 1990; Berbel et al. 1993). In other words, callosal axons that normally withdraw during development appear to be retained after neonatal hypothyroidism, which could explain an increased callosal thickness. More recent research suggests that the effects may be caused by the inappropriate expression of proteins TAG-1 and L1 (Alvarez-Dolado et al. 2000, 2001). Those proteins are members of the immunoglobulin super family of cell adhesion molecules and are thought to play an important role in axon growth and guidance during development. Alvarez-Dolado et al. (2000, 2001) found that both TAG-1 and L1 are overexpressed in major fiber tracts, including the anterior
Animal Studies 135 commissure and the corpus callosum, of hypothyroid rats during develop- ment. Therefore, although not widely recognized as a classic marker of neonatal hypothyroidism, increased thickness of the corpus callosum appears to be a biologically plausible effect. Another important issue related to the biologic plausibility of effects on brain structure is whether perchlorate exposure reduces maternal or pup thyroid hormone concentrations enough to alter neurodevelopment. Most studies of neonatal hypothyroidism in animals have used excessive doses of potent goitrogens that reduce serum T4 concentrations to near the detec- tion limit of the assay. The potential for modest reductions in serum T4 to alter gross morphologic end points, such as the thickness of the corpus callosum or frontal cortex, is unknown and may need to be investigated with the use of positive control groups exposed to moderate doses of model goitrogens. In summary, two studies (Argus 1998; Argus 2001) were conducted in Sprague-Dawley rats to investigate the effect of maternal perchlorate exposure on offspring brain development. Dams were exposed to various doses of perchlorate in drinking water throughout gestation and lactation. Pups were sacrificed at several postnatal ages, and their brains were fixed and sectioned for histologic evaluation. The thickness of various brain regions was measured. Statistical analyses revealed a number of significant effects, most notably an increase in the thickness of the posterior corpus callosum. However, questions and concerns about the studies have been raised, including (1) apparent systematic differences in the plane of section among treatment groups, (2) lack of a clear and consistent dose-response relationship, (3) doubts about the biologic plausibility of the changes that were observed, and (4) concerns that the measures that were used were relatively insensitive and would be unlikely to pick up subtle differences in neurodevelopment. On the basis of its review of the data, the committee concludes that the evidence in Argus (1998) and (2001) is inadequate to determine whether or not a causal relationship exists between maternal perchlorate exposure and pup neurodevelopmental abnormalities. NEUROBEHAVIORAL STUDIES The 1998 Argus study included a battery of behavioral tests that in- cluded tests of passive avoidance, water-maze learning, motor activity, and auditory startle. The behavioral tests were not repeated in the 2001 Argus study. Auditory-startle habituation amplitude was evaluated on postnatal days 23 and 61. It was an appropriate screening test for the effects of
136 Health Implications of Perchlorate Ingestion perchlorate because previous studies had shown that auditory-startle ampli- tudes are altered after exposure to propylthiouracil (PTU), a model goitro- gen (Goldey et al. 1995b), or polychlorinated biphenyls (PCBs), a thyroxine-suppressing environmental contaminant (Goldey et al. 1995a). Neonatal exposure to PTU and PCBs resulted in suppressed startle ampli- tudes at 24 days of age and increased amplitudes in adulthood. In the 1998 Argus study, no significant effects of perchlorate on startle amplitude were noted at either age. Two learning tests, passive avoidance learning and a water-filled M maze, also were administered. Learning deficits have been reported in children and in animals that had hypothyroidism (Brosvic et al. 2002; Rovet 2002), so inclusion of the tests was appropriate. Passive-avoidance testing conducted on days 23-25 consisted of placing the rats in the bright side of a two-compartment chamber. The rats were allowed to explore until they entered the dark compartment, where a 1-mA electric shock was delivered. The rats were then removed from the dark compartment and placed back in the bright compartment. Testing continued until the rats remained in the bright side of the chamber for 60 seconds in two consecutive trials. No significant effects of perchlorate were noted. Water-maze testing conducted on days 59-63 consisted of a series of trials in which the rats were placed in a standard start position and had to swim to one of two goals. Rats were trained to a criterion of five consecutive errorless swims. Again, no signifi- cant effects of perchlorate were noted. As discussed above, neonatal hypothyroidism results in abnormal development of various brain regions. The hippocampus and cerebellum are two of the most severely affected regions, so such learning tests as the radial-arm maze, which is sensitive to hippocampal damage (Becker et al. 1980), and eye-blink conditioning, which is sensitive to cerebellar damage (Steinmetz 1996), could potentially pick up subtle effects of altered thyroid function on cognition that might be missed by the tests that were adminis- tered. The tests that were used would not be the most appropriate ones for identifying functional effects of alterations in the structure of the corpus callosum, the most striking (although controversial) neuromorphologic findings reported in Argus (1998, 2001). In humans, a thicker corpus callosum has been linked to deficits in executive function (Bookstein et al. 2002), so tests of functions, such as working memory, attention, cognitive flexibility, and inhibitory control, might be useful. Examples would be a delayed-response task, a serial-reversal learning task, and an operant sched- ule that reinforced animals for low rates of responding. Those tests and the ones mentioned above require considerable time and effort and would be
Animal Studies 137 difficult to incorporate into large-scale projects like the Argus studies, which were designed to assess multiple heath outcomes. Furthermore, they are not required in EPAâs Developmental Neurotoxicity Testing Guidelines (EPA 1998a). However, without additional cognitive assessments that use tests that focus on the specific cognitive domains most likely to be affected by subtle to moderate reductions in thyroid hormone concentrations, it is impossible to rule out an effect of perchlorate on cognition. Another useful approach would be a transgenic mouse model with a phenotype characterized by enlarged corpus callosum axons (Seeger et al. 2003). Characterization of the behavioral profile in such animals would provide valuable information that would help to correlate behavioral changes with morphologic effects of perchlorate exposure. Locomotor-activity testing was conducted at the ages of 14, 18, 22, and 61 days in the 1998 Argus study, and at the ages of 14, 18, and 22 days by Bekkedal et al. (2000). The design of the Bekkedal et al. (2000) study was similar to that of the Argus study. Female Sprague-Dawley rats were treated with ammonium perchlorate at 0, 0.1, 1.0, 3.0, or 10.0 mg/kg per day in the drinking water beginning 2 weeks before gestation and continuing until postnatal day 10. The rats in both studies were tested in an open-field activity monitor transected by infrared photobeams. The screening tool was appropriate because treatment with PTU or PCBs has been shown to sup- press locomotor activity during preweaning development and increase locomotor activity at later ages (Goldey et al. 1995a, 1995b). However, locomotor activity can be variable because it is influenced by many parame- ters. Therefore, it is not a very sensitive measure of exposure-related effects. No significant effects were noted in the Bekkedal et al. (2000) study. The only significant effect noted in the Argus study was an increase in activity on postnatal day 14 in male pups in the highest-exposure per- chlorate group (10.0 mg/kg). The effect was manifested as a failure to habituate over the 1.5-hour test session. Argusâs statistical analyses indi- cated a significant effect when male and female pups were evaluated sepa- rately, but a reanalysis in which male and female data were combined found no significant effects. In November 2001, David Dunson, of the National Institute of Environ- mental Health Sciences (NIEHS), was asked to provide an analysis of Argus (1998) and of Bekkedal et al. (2000) because the original contractor was unable to find statistically significant effects despite large increases in several outcome variables. EPA neurotoxicologists believed that the increases were of concern from a biologic perspective. Because of the high correlation between different types of motor-activity measurements, Dunson focused on the number of ambulatory movements, which was viewed as
138 Health Implications of Perchlorate Ingestion most reflective of overall motor activity. He used a Bayesian hierarchic model to assess the dose-response trend in motor activity. In a Bayesian statistical approach, expert judgment and available data are modeled in the prior-probability distribution. When data and expertise are scarce, the prior-probability distribution is diffuse (spread over a wide range of plausi- ble values, each with low probability). When the prior distribution is developed, Bayesâs theorem is used to update it with information from the current data to arrive at the posterior-probability distribution. In many cases, implementation of Bayesâs theorem involves complex integrals. In those complex modeling situations, Markov Chain Monte Carlo methods can be used to make calculation of Bayesâs theorem simple by using se- quential sampling algorithms. Dunson (2001) included dose, sex, age, habituation time, and a habituation-time-by-dose-interaction term in the model. The correlation due to the repeated-measure structure of the data was accounted for by using an animal-specific intercept. Plausible but diffuse prior distributions were used for all parameters. The prior distribu- tion of most interest, that for the dose-response relationship, was centered on a value corresponding to no effect of perchlorate exposure. The commit- tee observed that the Bayesian model was carefully designed and that the analysis was properly conducted according to standard procedures. Dunsonâs reanalysis of the Bekkedal data yielded a posterior probability of 98% that motor activity increased with perchlorate exposure. The probability increased from 79% in the first habituation interval to over 99% in the final interval. The dose estimated to increase the mean number of ambulatory movements at the final habituation time by 10% (effective dose 10%, or ED10) was 1.62 mg/kg per day with a 95% credible interval1 of 0.90-7.87 mg/kg per day. Dunsonâs reanalysis of the Argus data yielded a high posterior probability that the motor activity increased with perchlorate exposure. The probability increased from 58% in the first habituation interval to 94% in the final interval. The ED10 was 4.60 mg/kg per day with a 95% credible interval of 2.18 mg/kg per day to infinity. To combine the two studies, the model was modified to include distinct baseline parame- ters for the error variances, intercept, age effects, and habituation-time effects, but the slopes for the dose-response relationship were assumed to be common across studies. In the combined analysis, there was a posterior probability of 99% that motor activity increased with perchlorate exposure in the final habituation interval. The ED10 was 3.33 mg/kg per day with a 95% credible interval of 1.91-12.78 mg/kg per day. 1 A Bayesian credible interval is an interval such that the probability that the parameter lies in the interval is at least the given percentage.
Animal Studies 139 Two concerns arise with Dunsonâs analysis. The first is derived from the fact that the point estimate of the ED10 from the Bekkedal et al. (2000) study was 1.62 mg/kg per day and does not fall into the posterior credible interval of the combined analysis. Although Dunson was using a âstan- dardâ method for the day, it is likely that he underestimated the uncertainty in the ED10 by his choice of distributions for the combined analysis. Underestimating uncertainty is an example of problems with the âstandardâ that statisticians have come to recognize (Berry 2000). The second concern is derived from using the lower bound of the credible interval to come up with conservative estimates. Indeed, whenever the sample is small, inter- vals will be wide, and lower bounds will be small. Although Dunson presents credible intervals, presenting only conservative estimates based on lower bounds is misleading. Dunson claims that an ED10 of 1-2 mg/kg per day represents a conservative estimate, but it is over-conservative. On the basis of his own analysis, the posterior probability that the ED10 is less than or equal to 2 mg/kg per day would be very small; therefore, the posterior probability that the ED10 would be greater than 2 mg/kg per day would be large. As with the other measures that were used to assess behavioral function, general motor activity is not necessarily the most relevant or most sensitive aspect of motor function to assess if neonatal hypothyroidism is the sus- pected mechanism of action. As discussed above, the cerebellumâa brain region that is critically important for controlling balance and coordina- tionâis severely affected by neonatal hypothyroidism. Such tests as the rotating rod and rope climb, which were devised to assess the functional consequences of cerebellar damage (Altman and Bayer 1997), are relatively easy to conduct and may be more sensitive to effects of perchlorate expo- sure. Those tests have been used to assess motor function in rats exposed to alcohol, which is known to target the cerebellum (Klintsova et al. 2000), or PCBs, which are known to decrease serum T4 concentrations (Roegge et al. 2004). Another functional end point that may be important to assess in perchlorate-exposed pups is hearing. Thyroid hormones are known to play an important role in the development of the cochlea, and neonatal hypothy- roidism has been linked to hearing loss in animals (Knipper et al. 2000) and humans (Wasniewska et al. 2002). PCBs also cause hearing loss (Goldey et al. 1995b), which can be at least partially ameliorated by concurrent treatment with T4 (Goldey and Crofton 1998). The deficit appears to result from damage to the outer hair cells of the cochlea (Crofton et al. 2000). Recent studies suggest that a moderate decrease in T4 (about 50%) is suffi- cient to cause significant low-frequency hearing loss. Thus, the moderate
140 Health Implications of Perchlorate Ingestion decreases in T4 observed by Argus (1998, 2001) may be large enough to cause subtle deficits in auditory function. Hearing can be assessed in rodents with a number of methods, including reflex-modification audiometry (e.g., Goldey et al. 1995a,b) and the mea- surement of distortion-product otoacoustic emissions (Lasky et al. 2002). The latter requires fairly sophisticated equipment but has several advan- tages. First, a detailed assessment of hearing across a broad range of fre- quencies can be obtained from an anesthetized animal in a single 1-hour test session. Second, the method directly assesses the functional integrity of the cochlea, a primary site of damage in hypothyroid-induced hearing loss (Ng et al. 2004). Third, the method can be applied in animals and human infants, facilitating extrapolation of findings from animals to sensitive human populations. A recent study with the method found hearing loss across a range of low and middle frequencies in PCB-exposed animals (Widholm et al. 2003). Future rodent studies should focus on the specific functional end points described above that are likely to be adversely affected by moderate reduc- tions in thyroid hormone. Those include balance and coordination and motor learning (Altman and Bayer 1997), spatial working memory (Olton et al. 1979), and auditory function (Lasky et al. 2002). The behavioral and sensory tests should be accompanied by studies of brain morphology that go beyond simply measuring the thickness, area, or volume of particular brain areas and look for neuromorphologic changes in the cerebellum, hippocampus, and cochlea that are consistent with neonatal hypothyroidism. The studies should make use of specialized staining techniquesâsuch as the Golgi, Timms, and Nissl stainsâwhich can be used to visualize changes in cerebellar Purkinje and hippocampal pyramidal cell morphology (Gould et al. 1990; Golgi stain), hippocampal mossy fiber pathways associated with spatial learning and memory (Schwegler et al. 1990; Timms stain), and general neuronal organization and neuronal cell counts in various brain regions (Ribak, 1986; Nissl stain). Outer hair cells of the cochlea should also be assessed because they are damaged when insufficient thyroid hormone is present during development (Crofton et al. 2000). For those approaches to be successful, it is imperative that the brains be perfusion- fixed before sectioning and staining and that the assessments be conducted in a blinded fashion. Relationships between corpus callosum alterations and behavioral sequelae have been described (Magara et al. 2000), so the biologic signifi- cance of changes in the corpus callosum should not be dismissed. As described above, a transgenic mouse model with a phenotype characterized by enlarged corpus callosum axons (Seeger et al. 2003) may be of use in
Animal Studies 141 determining behavioral deficits that may be associated with enlargement of the corpus callosum. The p21H-Ras gene is associated with that transgenic phenotype; looking at changes in expression of this gene after treatment with perchlorate may provide additional means of correlating the corpus callosum changes with behavioral changes and may help validate that correlation. It would also be useful to measure the extent to which perchlorate exposure during brain development alters the expression of thyroid hormone-responsive genes and gene products in the brain. That approach has been useful in understanding the actions of other agents that affect the thyroid system (Zoeller et al. 2002; Gauger et al. 2004). In summary, a number of behavioral measures were assessed in the 1998 Argus study. Motor activity, auditory startle, and learning and mem- ory were appropriate functions to assess, given the suspected mode of action of perchlorate. However, the tests used in the Argus study were screening measures and would be unlikely to pick up subtle alterations in motor or cognitive function associated with moderate reductions in serum T4 concentrations. Some important end points, such as auditory function and balance or coordination, were not assessed, so it is not surprising that no significant effects were observed in any of the behavioral measures except an increase in motor activity in male pups on one day of testing. Given the lack of sensitivity of the tests that were used, the committee concludes that the results are inadequate to determine whether or not gestational or lactational exposure to perchlorate affects behavioral function. THYROID GLAND TUMORS Few long-term cancer bioassays involving perchlorate exposure have been conducted.2 In its draft risk assessment (EPA 2002a), EPA reviewed two studies that evaluated the occurrence of thyroid tumors in rodents exposed to perchlorate. First, Kessler and Kruskemper (1966), as cited in EPA (2002a), administered potassium perchlorate to male Wistar rats in drinking water at 0 or 1% for 24 months. EPA estimated the daily dose at 1,339 mg/kg. Histologic changes were similar to those produced by expo- sure to antithyroid agents. Four of the 11 treated rats developed benign 2 Perchlorate has been evaluated in standard in vitro and in vivo assays used to assess genotoxicity (EPA 2002a). The results of those assays were negative.
142 Health Implications of Perchlorate Ingestion tumors of the thyroid gland; no tumors were observed in the 20 control rats. Second, Pajer and Kalisnik (1991) administered 0 or 1.2% sodium perchlorate in drinking water to female BALB/c mice for 46 weeks. There were three groups of controls and three groups of mice treated with per- chlorate (12 mice per group). The estimated perchlorate dose was 2,147 mg/kg per day (EPA 2002a). One control group and one group of perchlorate-exposed mice were subjected to whole-body irradiation at 8 or 32 weeks with 0.8 Gy on 5 consecutive days at a dose rate of 1.45 Gy per minute for a total radiation dose of 4 Gy. Surviving animals were sacrificed at 46 weeks for histologic examination of the thyroid and pituitary glands. Thyroid follicular-cell carcinomas were observed in five of the six non- irradiated perchlorate-treated mice and in all 14 irradiated perchlorate- treated mice. No thyroid follicular-cell carcinomas were observed in the 22 control animals. The committee found additional information in three other studies. First, Hiasa et al. (1987) found that ammonium perchlorate administered in the diet at 1,000 ppm was a thyroid gland tumor promoter in Wistar rats after tumor initiation with N-bis(2-hydroxypropyl)-nitrosamine (DHPN). After 19 weeks of treatment with perchlorate after DHPN administration, all 20 rats had thyroid adenomas, compared with one of 20 rats treated with DHPN alone. Rats treated with ammonium perchlorate alone had no thyroid adenomas. Second, Fernandez Rodriguez et al. (1991) reported that female Wistar rats administered 1% potassium perchlorate in the drinking water for 1-12 months developed a progressive increase in thyroid gland weight and a diffuse hypertrophy and hyperplasia of follicular cells with increased vascularity and decreased lumenal colloid. After 6 months of treatment, multiple (often bilateral) follicular-cell nodules of complex morphology appeared in the diffusely enlarged thyroid glands with a follicular, papillary, or trabecular histologic pattern. The follicular cells comprising the nodules often had a more basophilic cytoplasm and dense nuclear chromatin than normal thyroid cells. The authors suggested that the thyroid nodules were probably due to overstimulation by TSH. Third, FernÃ¡ndez-Santos et al. (2004) reported Ki-ras mutational analysis of thyroid follicular-cell lesions induced by the administration of radioactive iodine (50 microcuries 131I) and potassium perchlorate (1% in drinking water to female Wistar rats) for 6, 12, and 18 months. No muta- tions were found in the amplified gene segment of any of the induced thyroid tumors. The results of the study suggested that Ki-ras activation via mutations at codons 12 and 13 is neither a constant nor an early event in the development of thyroid follicular-cell carcinoma in rats.
Animal Studies 143 In a two-generation study (Argus 1999), thyroid follicular-cell adenomas were observed in three male rats. The following discussion focuses only on the male rats. On arrival, Charles River CR/CD rats were individually housed and assigned to groups (30 rats of each sex per group) at ammonium perchlorate doses of 0 (control), 0.3, 3.0, and 30 mg/kg per day in drinking water provided ad libitum. Male rats (P1 generation) were treated for at least 70 days before mating, through the mating period, and until sacrifice at the age of about 24-25 weeks. The male offspring from that mating (F1 generation) were treated in a similar manner from at least 70 days before their mating until sacrifice at the ages of about 21-22 weeks. The duration of dosing in the drinking water was about the same for the P1 and F1 groups of animals except that there was additional exposure of the F1 pups during the gestation and lactation periods. After sacrifice, many tissues, including the thyroid, of the P1 and F1 rats were examined histo- logically. Slides from the Argus study were initially read by the Argus patholo- gist, and the findings are tabulated in the final report of the two-generation study (Argus 1999). Because of variability in the histologic criteria and examination among the various perchlorate studies, the slides from all the studies, including the two-generation study (thyroid gland only), were reread by an EPA pathologist and were reviewed by a pathology working group (Wolf 2000). The findings of the re-evaluation are listed in the tables prepared by EPA (Wolf 200l, Tables 14 through 21) and in the text of EPAâs draft risk assessment (EPA 2002a). EPAâs report of the re-evalua- tion states that two male rats (7094 and 7117) in the F1 generation treated at the high dose of perchlorate had follicular-cell adenomas and that one of them (7117) had two adenomas, for a total of three adenomas. The original study pathology report indicates that there was also a follicular-cell adenoma in a control male rat in the P1 generation (3617). In the F1 generation, rat 7094 (F1 high-dose male) had a follicular-cell adenoma. For rat 7117 (F1 high-dose male), the original pathology report indicates that there was a focal proliferative lesion and that it was consid- ered by the pathologist as hyperplasia, not neoplasia. In view of the differences between the original Argus study pathology report and the EPA text and tabular information, the committee requested photomicrographs of the three rats. The committee confirmed that the control male rat in the P1 generation had a follicular-cell adenoma. The committee agreed that rat 7094 had a follicular-cell adenoma and that the lesion originally diagnosed by the Argus pathologist as focal hyperplasia (severe) in rat 7117 was a follicular-cell adenoma, which was also the conclusion of the review of the pathology working group. Thus, in the two-
144 Health Implications of Perchlorate Ingestion generation study, there were follicular-cell adenomas in one control male rat (3617) in the P1 generation and two high-dose male rats (7094 and 7117) in the F1 generation. There were no adenomas at lower doses or in the controls in the F1 generation. The age at which the P1 and F1 generations were sacrificed and deter- mined to have thyroid follicular-cell adenomas are comparable (about 5-6 months). No tumors were observed in the high-dose male rats in the P1 generation. The duration of treatment of the P1 generation was similar to that of the F1 generation, but the F1 generation may have had additional exposure during the gestation and lactation periods. In its draft risk assessment, EPA considered the tumors in the F1 generation to be treatment-related and noted that they were particularly remarkable because they occurred at 19 weeks (Wolf 2000). EPA requested that the National Institute of Environmental Health Sciences perform statistical analyses of the tumor data in the F1 males, which involved an extensive Bayesian analysis that incorporated historical control data from F344 and Sprague-Dawley rats. It was concluded that thyroid adenomas were statistically increased in the high-dose (30-mg/kg) group of F1 ani- mals sacrificed as adults (P2 generation) at 19 weeks and that the latency and incidence of the tumors were remarkable relative to the entirety of the National Toxicology Program database for this type of tumor in this strain of rat (Dunson 2001). With respect to the observation of thyroid follicular-cell adenomas in the two-generation study, these would be expected in high-dose male rats in the presence of a markedly goitrogenic regimen, which existed under the conditions of the study. Although the tumor response observed in the high- dose F1 males is not statistically significant by conventional statistical analyses (Fisherâs exact test), the tumors observed in the male high-dose group are most likely related to treatment. Spontaneous thyroid follicular- cell adenomas can occasionally be observed in control rats of this strain and age, so follicular-cell adenomas in rats treated with a markedly goitrogenic regimen would not be an unexpected or unusual finding. The International Agency for Research on Cancer (IARC) held a working group on thyrotropic agents and issued a monograph in 2001 (IARC 2001). The monograph provided the following statements: âAgents that lead to the development of thyroid neoplasia through an adaptive physiological mechanism belong to a different category from those that lead to neoplasia through genotoxic mechanisms or through mechanisms involv- ing pathological responses with necrosis and repair. Agents that induce thyroid follicular cell tumors in rodents by interfering with thyroid hormone homeostasis, can with some exceptions, notability the sulfonamides, also
Animal Studies 145 interfere with thyroid hormone homeostasis in humans if given at a suffi- cient dose for a sufficient time. These agents can be assumed not to be carcinogenic in humans at concentrations that do not lead to alterations in thyroid hormone homeostasis.â In addition, EPAâs science policy document on the assessment of thyroid follicular-cell tumors notes that although there may be some qualita- tive similarities, there is evidence that âhumans may not be as sensitive quantitatively to thyroid cancer development of thyroid-pituitary disruption as are rodentsâ (EPA 1998b). The increased sensitivity may be due to marked species differences in the physiology of the thyroid gland (EPA 1998a; Hill et al. 1989). The EPA and IARC documents provide guidance for the evaluation of thyroid follicular-cell tumors based on mode of action (for example, tumors secondary to hormone imbalance). Thus, the committee reached the following two conclusions: â¢ In the case of perchlorate, follicular-cell tumors in rats are not an unexpected finding at doses that are goitrogenic. â¢ It is unlikely that perchlorate poses a risk of thyroid cancer in humans. PBPK MODELING Physiologically based pharmacokinetic (PBPK) modeling is one of the methods of choice for determining human equivalent exposures (HEEs) and adjusting default uncertainty factors associated with the derivation of reference doses and reference concentrations for lifetime human exposures from animal studies (EPA 2002a). Thus, EPA relied on a series of PBPK models developed by the Department of Defense after conducting a peer review to facilitate interspecies extrapolations in its draft perchlorate risk assessment (EPA 2002b, 2003). The PBPK models were initially devel- oped to describe the disposition (absorption, distribution, metabolism, and elimination) of perchlorate in adult rats (Fisher 2000). As data became available and the mode of action for perchlorate-induced effects were shown to be mediated through interactions with iodide at the thyroid so- dium (Na+)/iodide (I2) symporter (NIS), the initial model was expanded to include the disposition of iodide and the inhibition of iodide uptake at the NIS in pregnant rats and fetuses, lactating rats and neonates, and adult humans to address dose-response issues associated with potentially sensitive populations (Clewell et al. 2001, 2003a,b; Merrill 2001, Merrill et al. 2003). It is currently impractical, and in many cases unethical, to validate
146 Health Implications of Perchlorate Ingestion PBPK simulations of the kinetic or dynamic responses to chemical chal- lenges in the human fetus or neonate. Therefore, the developers of the PBPK models proposedâand EPA concurredâto use a parallelogram approach to constrain predictions of equivalent exposures for human fetuses and infants corresponding to the no-observed-adverse-effect level (NOAEL) in animal studies (see Figure 4-4). According to this approach, an internal dose that is relevant to toxicity in the fetus or neonate (for example, the concentration of perchlorate in blood or serum or the inhibition of iodide uptake by the thyroid in the dams) is first determined for the NOAEL in the critical animal toxicity study. By applying factors for life-stage differences (such as pregnant or lactating female rat vs adult male rat) and for species differences (such as adult rat vs adult human or pregnant or lactating rat vs pregnant or lactating human), one can constrain PBPK simulations by using known physiologic measures and biochemical constants to estimate HEEs in potentially sensitive populations that may not be suitable for experimen- tal validation. The committee agrees with EPA that PBPK modeling constitutes the best available approach to determining HEEs and adjustments of default uncertainty factors when reference doses are based on animal data. The PBPK models developed by DOD for the adult rat, adult human, pregnant rat and fetus, and lactating rat and neonate represents the current state of the science for integrating available animal and human data on the disposition of perchlorate and iodide and the interactions between these anions at the level of the thyroid NIS. Although many of the PBPK model parameters had to be estimated on the basis of a small set of in vivo pharmacokinetic studies, enough studies were available for validation of model simulations over a broad range of doses of both perchlorate and iodide to lend confi- dence to the applicability of the models for extrapolating from animal-study doses to human exposures. Further details on the PBPK models developed for EPAâs risk assessment can be found in Appendix E. OTHER EFFECTS OF PERCHLORATE Effects on Iodide Metabolism in Nonthyroid Tissues in Animals The NIS is present in substantial amounts not only in the thyroid gland but also in several other tissues, including the salivary glands, mammary glands, stomach, choroid plexus of the brain, and ciliary body of the eye (see Chapter 2). Iodide that is transported into those tissues returns rapidly
Animal Studies 147 Laboratory Animal Exposure (mg/kg per day) Internal Dose Adult Rat Factor for Male Rat Gestation Model and Life-Stage Lactation Differences Models (KL) Factor for Species KS Differences (KS) Adult Human Human Gestation and Model Lactation KL Models External Dose Human Equivalent Exposure (mg/kg per day) FIGURE 4-4 Parallelogram approach for using adult human, adult male rat, and female rat gestation and lactation models to estimate human equivalent exposures for human pregnancy and lactation models. Bold arrows indicate presence of validated PBPK models for estimating relationships between effective internal doses associated with key events and perchlorate exposure; dashed arrows and shaded box indicate theoretical extrapolations guided by physiologic and biochemi- cal constraints in models and parallelogram approach. to the circulation or is secreted into the saliva or breast milk. With molecu- lar methods, very small amounts of NIS have also been detected in other tissues, including the heart, kidneys, lungs, and placenta, but there is little
148 Health Implications of Perchlorate Ingestion evidence that iodide is transported into them (Dohan et al. 2003). Three sitesâmammary gland, placenta, and kidneyâare briefly considered here because inhibition of NIS by perchlorate at these sites might contribute to in vivo effects of perchlorate. The NIS of the mammary gland is important for two reasons: it can provide a means of transferring perchlorate to newborn infants, and per- chlorate inhibition of the mammary gland NIS could decrease the iodide content of breast milk. Those issues were at least partially addressed by pharmacokinetic modeling of iodide and perchlorate metabolism in rats (Clewell et al. 2003b). Experiments conducted by Mahle et al. (2003) established that perchlorate is transferred from nursing rats to their pups. Expression of NIS in the mammary gland differs from that in the thyroid gland. Mammary gland NIS is not affected by TSH, and the content of NIS in the mammary gland is low in virgin rats and high in lactating rats (Spitzweg et al. 1998; Tazebay et al. 2000). High NIS concentrations during lactation appear to be under the control of oxytocin and prolactin (Cho et al. 2000). Iodide availability to the developing fetus is probably influenced by a variety of factors, including maternal iodide intake, placental and uterine deiodinating activity, and, potentially, placental NIS. Unlike the transport of iodide into thyroid and mammary tissue against a concentration gradient, placental transfer of iodide from mother to fetus is down a concentration gradient, so the importance of placental NIS in the process is not known. Nonetheless, inhibition of NIS in the placenta might reduce the transfer of iodide to the fetus, although it is not known to do so. In humans, expression of the NIS gene is higher in cultured placental cells from first-trimester pregnancies than in placental cells from third-trimester pregnancies, but NIS gene expression in crude placental extracts is similar at both times (Bidart et al. 2000). However, gene expression might not correlate with functional iodide transport, because NIS that is produced might not reach the cell membrane, where iodide transport occurs. In pregnant rats, NIS gene expression in the placenta was higher in response to iodide deficiency and potassium perchlorate administration on gestation day 21 than in control rats (Schroder-van der Elst et al. 2001). Very small amounts of NIS are present in the cells that form the renal tubules. In the proximal tubules, it is primarily in the basolateral membrane of the cells (the side that is not exposed to urine) whereas it is more diffuse in the cells of the distal tubules (Spitzweg et al. 2001; Dohan et al. 2003). Little iodide is absorbed from the urine into the renal tubular cells, so it is unlikely that the small amounts of NIS in renal tubules reduce urinary iodide excretion. In a study in which some pregnant rats were fed perchlo-
Animal Studies 149 rate, their urinary excretion of iodide was higher than that of pregnant rats fed a normal diet, but the increase was due to decreased iodide uptake by the thyroid (SchrÃ¶der-van der Elst et al. 2001). General Toxicity Despite its clinical use, few studies have focused on the general toxicity of perchlorate. Perchlorate has been widely used in biologic systems to study anion channels, anion-transport systems, and anion-constrained conformations of macromolecules. The literature was searched to identify potential interactions of perchlorate with non-NIS molecular targets. Such targets consisted of a variety of enzymes, such as cytochrome c and rabbit muscle enolase (Andersson et al. 1980; Robinson et al. 1983; Arai et al. 1984; Kornblatt et al. 1996). The concentrations of perchlorate required to bind to or inhibit those enzymes were up to several orders of magnitude greater than the concentrations needed to inhibit NIS. Likewise, perchlorate was found to affect the function of isolated cell systems, such as pancreatic islet cells and skeletal muscle fibers, but at much higher concentrations than those causing inhibition at the thyroid NIS (Gomolla et al. 1983; Sehlin 1987; Csernoch et al. 1987; Frankel and Sehlin 1994; Larsson-Nyren 1996; Larsson-Nyren et al. 2001). Thus, the high concentrations required for the interactions of perchlorate suggest that none of those other targets are important toxicologically in comparison with its more specific interaction with the NIS. The primary publications that address a broad spectrum of toxicologic end points are reports of a 90-day study (Siglin et al. 2000) and of two reproductive and developmental studies (York et al. 2001a,b). The papers were important to the committee in determining whether any adverse effects, other than those associated with inhibition of iodide uptake by the thyroid, might arise from ingestion of perchlorate at low doses. Siglin et al. (2000) administered ammonium perchlorate in drinking water to groups of Sprague-Dawley rats (10 rats of each sex per group) at 0, 0.01, 0.05, 0.2, 1.0, and 10 mg/kg per day for 14 or 90 days. Groups of rats also were evaluated 30 days after termination of the 90-day exposure at 0, 0.05, 1.0, and 10 mg/kg per day. In addition to thyroid function and thyroid histologic evaluations, standard toxicologic end points were evalu- ated, including clinical signs, body weights, food and water consumption, and routine hematologic and clinical-chemistry measures. An extensive list of tissues was examined microscopically in the control and high-dose rats; the liver, kidneys, lungs, and thyroids were examined in all animals.
150 Health Implications of Perchlorate Ingestion Estrous cycles were monitored, and sperm were evaluated for count, motil- ity, and structure. Bone samples also were evaluated for bone marrow micronucleus formation. Changes in thyroid hormones and TSH were observed as low as 0.01 mg/kg per day during exposure periods. However, changes in thyroid gland weight and thyroid histopathology were observed only at the highest dose (10 mg/kg per day). Although a few values were statistically significantly different from control values, no treatment-related effects were observed in any of the nonthyroid toxicologic measures. The authors concluded that the study provided further evidence that the thyroid is the primary target of perchlorate exposure in the rats. York et al. (2001a) conducted a reproductive and developmental study (two-generation) in rats given ammonium perchlorate at daily doses of 0, 0.3, 3.0, and 30 mg/kg in drinking water. In addition to measures of thyroid pathology and function, organ weights and pathology of nonthyroid tissues were examined in the parental generation (P1) and two generations of pups (F1 and F2). Reproductive measures were examined in the P1 and F1 generations. There was some evidence of reduced sperm density, spermatid count, spermatid concentration, and spermatid density at the high dose administered, but the differences were not statistically significant. There were no other remarkable changes in any nonthyroid measure. York et al. (2001b) conducted a developmental toxicity test in rabbits treated with ammonium perchlorate in drinking water from gestational days 6 through 28. Doses were adjusted to give targets of 0, 0.1, 1.0, 10, 30, and 100 mg/kg per day. No developmental anomalies were observed. Evidence of morphologic changes of thyroid follicular cells was observed at doses at 10 mg/kg per day or higher. On the basis of the data reviewed, the committee concludes that per- chlorate is very unlikely to have toxicologic effects at doses lower than those which would affect thyroid function. Immunologic Effects Perchlorate could theoretically produce several types of adverse immu- nologic reactions, including suppression of the function of a cellular compo- nent of the innate or adaptive immune systems, suppression of antibody- or cell-mediated responses, upregulation of immune-cell function or the full immune response, and induction of an immediate hypersensitivity (allergic) or delayed-type hypersensitivity reaction. The effects could possibly be mediated by a direct effect on the cells involved or indirectly by modulation of thyroid hormone-immune system homeostasis (Blalock 1994; Fabris et al. 1995; Klecha et al, 2000).
Animal Studies 151 The potential of perchlorate to cause any of those adverse immunologic reactions has been studied in animals. In a 1993 study in which female rats were fed a low-iodide diet and perchlorate to induce iodide deficiency, the rats developed some of the pathologic changes of autoimmune thyroiditis and an increase in production of antithyroid antibodies (Mooij et al. 1993). The tremendous expansion in recent years of knowledge of the compo- nents and functions of the human immune system has depended heavily on experiments in mice. Thus, mice are clearly the animals of choice to explore possible immunotoxic effects of perchlorate. However, it must be kept in mind that there are major species differences in various aspects of the immune system that preclude direct extrapolation of results from mice to humans (Mestas and Hughes 2004). Among mouse strains, B6C3F1 female mice have become the test strain of choice for immunotoxicity studies and have been used in extensive studies of perchlorate by Keil et al. (1998, 1999). In those studies, mice were exposed to perchlorate in drinking water to achieve doses of 0.1, 1, 3, and 30 mg/kg per day. Mice were evaluated after 14 or 90 days of treat- ment and 30 days after 90 days of treatment. Serum T4 concentrations were decreased significantly after 14 days in the 3- and 30-mg/kg groups and after 90 days in the 1-, 3-, and 30-mg/kg groups, but serum T3 and TSH concentrations were normal. The mice were evaluated with standard assays (Luster et al. 1988) for assessment of chemical-induced immunotoxicity. There were no significant or consistent differences between the control and perchlorate-exposed groups in any of the following outcomes: blood counts except reticulocytes only at the 90-day 3-mg/kg dose; weight or cellularity of the thymus, spleen, and bone marrow; CD4+/CD8+ lymphocyte counts; cytotoxic T- lymphocyte activity; IgG and IgM antibody responses to injections of sheep red blood cells (considered one of the most sensitive indicators of whether a chemical has immunosuppressive activity); macrophage nitric oxide production; resistance to a tumor challenge; and antinuclear antibody pro- duction (Keil et al. 1999). There was a persistent increase in NK cell activ- ity only in the 30-mg/kg group. The one somewhat consistent finding was decreased phagocytosis of Listeria monocytogenes by peritoneal macro- phages from perchlorate-exposed mice (all dose groups without a dose- response effect). The assay consisted of counting apparently internalized bacteria in stained smears. The biologic importance of the reduced phago- cytosis was brought into question by in vivo experiments in which the perchlorate-exposed mice had normal resistance to infection by the same listeria. In later studies (BRT-Burleson Research Technologies 2000 a,b,c), the antibody responses, as measured by the appearance of plaque-forming cells,
152 Health Implications of Perchlorate Ingestion to injections of sheep red blood cells were similar in normal and perchlo- rate-exposed mice. The local lymph node response to skin exposure to the sensitizing agent dinitrochlorobenzene was accentuated in the perchlorate groups, but the effect was not consistent, and there was no clear dose- response relationship. Perchlorate was not tested as a skin-sensitizing agent. Those studies have been analyzed extensively by outside reviewers (RTI 1999; EPA 2002a,b; TERA 2002). The studies were generally judged to have been broad in scope and performed carefully with standard assays. The majority of reviewers concluded that the studies as a group had not demonstrated a causal relationship in mice between perchlorate ingestion in drinking water and any biologically meaningful stimulatory or inhibitory effect on the immune system. The committee agrees with previous review- ers and concludes that the evidence favors rejection of a causal relationship between ingestion of perchlorate and an immunotoxic effect in animals. SUMMARY The committee found that the animal studies of potential adverse effects of perchlorate provided qualitative information, but the usefulness of the studies for quantitatively estimating the risk of adverse effects in humans is small. The major conclusions from the animal data are summarized below. â¢ Perchlorate has an antithyroid effect on rats at high doses (30 mg/kg of ammonium perchlorate per day). That effect is characterized by de- creases in serum thyroid hormone and increases in serum TSH with morphologic changes in the thyroid gland. â¢ The data are inadequate to determine whether or not a causal relationship exists between perchlorate exposure of pregnant rats and neurodevelopmental abnormalities in their pups, given the flaws in experi- mental design and methods in the studies conducted to evaluate that end point. â¢ The data are inadequate to determine whether or not perchlorate exposure during gestation and lactation in rats has effects on behavior, given the lack of sensitivity of the tests conducted to evaluate that end point. â¢ Exposure to perchlorate can increase the incidence of thyroid tumors in rats when the doses are high enough to decrease thyroid hormone production and increase TSH secretion. â¢ The data favor rejection of a causal relationship between perchlo- rate exposure and immunotoxicity.
Animal Studies 153 â¢ There are no data to suggest that perchlorate has effects that are not mediated through inhibition of iodide transport in the thyroid gland. â¢ It is not possible to extrapolate data quantitatively from rodents to humans for purposes of human health risk assessment. Most experimental studies in animals designed to characterize the effects of perchlorate expo- sure have been done in rats. However, rats are much more sensitive to agents that disturb thyroid function than are humans, so the relevance of rat studies in quantitative terms to humans is limited. REFERENCES Altman, J., and S.A. Bayer. 1997. Epilogue: Behavioral consequences of experi- mental interference with cerebellar development. Pp. 726-751 in Development of the Cerebellar System: In Relation to Its Evolution, Structure, and Func- tions. Boca Raton: CRC Press. Alvarez-Dolado, M., M. Ruiz, J.A. Del Rio, S. Alcantara, F. Burgaya, M. Sheldon, K. Nakajima, J. Bernal, B.W. Howell, T. Curran, E. Soriano, and A. Munoz. 1999. Thyroid hormone regulates reelin and dab1 expression during brain development. J. Neurosci. 19(16):6979-6993. Alvarez-Dolado, M., A. Cuadrado, C. Navarro-Yubero, P. Sonderegger, A.J. Furley, J. Bernal, and A. Munoz, A. 2000. Regulation of the L1 cell adhesion molecule by thyroid hormone in developing brain. Mol. Cell. Neurosci. 16(4):499-514. Alvarez-Dolado, M., A. Figueroa, S. Kozlov, P. Sonderegger, A.J. Furley, and A. Munoz. 2001. Thyroid hormone regulates TAG-1 expression in the develop- ing rat brain. Eur. J. Neurosci. 14(8):1209-1218. Andersson, T., J. Angstrom, K.E. Falk, and S. Forsen. 1980. Perchlorate finding to cytochrome c: A magnetic and optical study. Eur. J. Biochem. 110(2):363- 369. Arai, Y., L. Oreland, and H. Kinemuchi. 1984. Effect of perchlorate treatment on mitochondrial MAO-A and -B activities. Med. Biol. 62(4):245-249. Argus Research Laboratories, Inc. 1998. A Neurobehavioral Developmental Study of Ammonium Perchlorate Administered Orally in Drinking Water to Rats. ARGUS 1613-002. Argus Research Laboratories, Inc., Horsham, PA. [See also York, R.G., J. Barnett, W.R. Brown, R.H. Garman, D.R. Mattie, and D. Dodd. 2004. A rat neurodevelopmental evaluation of offspring, including evaluation of adult and neonatal thyroid, from mothers treated with ammonium perchlorate in drinking water. Int. J. Toxicol. 23(3):191-214.] Argus Research Laboratories, Inc. 1999. Oral (Drinking Water) Two-Generation (One Litter per Generation) Reproduction Study of Ammonium Perchlorate in Rats. ARGUS 1416-001. Argus Research Laboratories, Inc., Horsham, PA. Argus Research Laboratories, Inc. 2001. Hormone, Thyroid and Neurohisto- logical Effects of Oral (Drinking Water) Exposure to Ammonium Perchlorate in Pregnant and Lactating Rats and in Fetuses and Nursing Pups Exposed to
154 Health Implications of Perchlorate Ingestion Ammonium Perchlorate During Gestation or Via Maternal Milk. ARGUS 1416-003. Argus Research Laboratories, Inc., Horsham, PA. Becker, J.T., J.A. Walker, and D.S. Olton. 1980. Neuroanatomical bases of spatial memory. Brain Res. 200(2):307-320. Bekkedal, M.Y.V., T. Carpenter, J. Smith, C. Ademujohn, D. Maken, and D.R. Mattie. 2000. A Neurodevelopmental Study of Oral Ammonium Perchlorate Exposure on the Motor Activity of Pre-Weanling Rat Pup. Report No. TOXDET-00-03. Neurobehavioral Effects Laboratory, Naval Health Research Center Detachment (Toxicology), Wright-Patterson Air Force Base, OH. Berbel, P., A. Guadano-Ferraz, M. Martinez, J.A. Quiles, R. Balboa, and G.M. Innocenti. 1993. Organization of auditory callosal connections in hypothyroid adult rats. Eur. J. Neurosci. 5(11):1465-1478. Berry, S.M. 2000. Meta-analysis versus Large Trials: Resolving the controversy. Pp. 65-81 in Meta-Analysis in Medicine and Health Policy, D. Stangl, and D.A. Berry, eds. New York: Marcel Dekker. Bianco, A.C., D. Salvatore, B. Gereben, M.J. Berry, and P.R. Larsen. 2002. Biochemistry, cellular and molecular biology, and physiological roles of the iodothyronine selenodeiodinases. Endocr. Rev. 23(1):38-89. Bidart, J.M., L. Lacroix, D. Evain-Brion, B. Caillou, V. Lazar, R. Frydman, D. Bellet, S. Filetti, and M. Schlumberger. 2000. Expression of Na+/I- symporter and Pendred syndrome genes in trophoblast cells. J. Clin. Endocrinol. Metab. 85(11):4367-4372. Blalock, J.E. 1994. The syntax of immune-neuroendocrine communication. Immunol. Today 15(11):504-511. Bookstein, F.L., A.P. Streissguth, P.D. Sampson, P.D. Connor, and H.M. Barr. 2002. Corpus callosum shape and neuropsychological deficits in adult males with heavy fetal alcohol exposure. Neuroimage 15(1):233-251. Brosvic, G.M., J.N. Taylor, and R.E. Dihoff. 2002. Influences of early thyroid hormone manipulations: Delays in pup motor and exploratory behavior are evident in adult operant performance. Physiol. Behav. 75(5):697-715. BRT-Burleson Research Technologies, Inc. 2000a. Ammonium Perchlorate: Effect on Immune Function. Quality Assurance Audit: Study No. BRT 19990524 -- Plaque-Forming Cell (PFC) Assay; Study No. BRT 19990525 - Local Lymph Node Assay (LLNA) in Mice. BRT-Burleson Research Technol- ogies, Inc., Raleigh, NC. June 30, 2000. BRT-Burleson Research Technologies, Inc. 2000b. Addendum to Study Report: Ammonium Perchlorate: Effect on Immune Function. BRT 19990524 Study Protocol Plaque-Forming Cell (PFC) Assay; BRT 19990525 Study Protocol Local Lymph Node Assay (LLNA) in Mice. BRT-Burleson Research Technol- ogies, Inc., Raleigh, NC. August 31, 2000. BRT-Burleson Research Technologies, Inc. 2000c. Ammonium Perchlorate: Effect on Immune Function. Study Report. BRT 19990524 Study Protocol Plaque-Forming Cell (PFC) Assay; BRT 19990525 Study Protocol Local Lymph Node Assay (LLNA) in Mice. BRT-Burleson Research Technologies, Inc., Raleigh, NC.
Animal Studies 155 Calvo, R., M.J. Obregon, C. Ruiz de Ona, B. Ferreiro, E. Escobar del Rey, and G. Morreale de Escobar. 1990. Thyroid hormone economy in pregnant rats near term: A âphysiologicalâ animal model of nonthyroidal illness? Endocrinology 127(1):10-16. Cho, J.Y., R. Leveille, R. Kao, B. Rousset, A.F. Parlow, W.E. Burak Jr., E.L. Mazzaferri, and S.M. Jhiang. 2000. Hormonal regulation of radioiodide uptake activity and Na+/I- symporter expression in mammary glands. J. Clin. Endocrinol. Metab. 85(8):2936-2943. Clewell, R.A., E.A. Merrill, and P.J. Robinson. 2001. The use of physiologically based models to integrate diverse data sets and reduce uncertainty in the prediction of perchlorate and iodide kinetics across life stages and species. Toxicol. Ind. Health 17(5-10):210-222. Clewell, R.A., E.A. Merrill, K.O. Yu, D.A. Mahle, T.R. Sterner, D.R. Mattie, P.J. Robinson, J.W. Fisher, and J.M. Gearhart. 2003a. Predicting fetal perchlor- ate dose and inhibition of iodide kinetics during gestation: A physiologically- based pharmacokinetic analysis of perchlorate and iodide kinetics in the rat. Toxicol. Sci. 73(2):235-255. Clewell, R.A., E.A. Merrill, K.O. Yu, D.A. Mahle, T.R. Sterner, J.W. Fisher, and J.M. Gearhart. 2003b. Predicting neonatal perchlorate dose and inhibition of iodide uptake in the rat during lactation using physiologically-based pharmaco- kinetic modeling. Toxicol. Sci. 74(2):416-436. Crofton, K.M., D. Ding, R. Padich, M. Taylor, and D. Henderson. 2000. Hearing loss following exposure during development to polychlorinated biphenyls: A cochlear site of action. Hear. Res. 144(1-2):196-204. Csernoch, L., L. Kovacs, and G. Szucs. 1987. Perchlorate and the relationship between charge movement and contractile activation in frog skeletal muscle fibers. J. Physiol. 390:213-227. Dohan, O., A. De La Vieja, V. Paroder, C. Riedel, M. Artani, M. Reed, C.S. Ginter, and N. Carrasco. 2003. The sodium/iodide Symporter (NIS): Character- ization, regulation, and medical significance. Endocr. Rev. 24(1):48-77. Dohler, K.D., C.C. Wong, and A. von zur Muhlen. 1979. The rat as model for the study of drug effects on thyroid function: Consideration of methodological problems. Pharmacol. Ther. 5(1-3):305-318. Dunson, D.B. 2001. Statistical Analysis of the Effects of Perchlorate on Neuro- behavior (Motor Activity) in SD Rats. Memorandum to Annie M. Jarabek, National Center for Environmental Assessment (MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC, from David B. Dunson, Biostatistics Branch (MD A3-03), National Institute of Environmental Health Sciences, Research Triangle Park, NC. November 11, 2001. [Online]. Avail- able: http://www.epa.gov/ncea/perchlorate/references2/ documents/100510.pdf [accessed July 19, 2004]. Elberger, A.J. 2003. Omaha Perchlorate Meeting â Module A. Analysis and Interpretation of Neurodevelopmental Rat Brain Morphometry Studies. Presentation at the Second Meeting on Assess the Health Implications of Perchlorate Ingestion, December 13, 2003, Irvine, CA.
156 Health Implications of Perchlorate Ingestion EPA (U.S. Environmental Protection Agency). 1998a. Health Effects Test Guide- lines OPPTS 870.6300 Developmental Neurotoxicity Study. EPA 712-C-98- 239. Office of Prevention, Pesticides and Toxic Substances, U.S. Environmen- tal Protection Agency, Washington, DC. [Online]. Available: http://www. epa.gov/opptsfrs/OPPTS_Harmonized/ 870_Health_Effects_Test_Guidelines/ Series/870-6300.pdf [accessed August 23, 2004]. EPA (U.S. Environmental Protection Agency). 1998b. Assessment of Thyroid - Follicular Cell Tumors. EPA/630/R-97/002. Risk Assessment Forum, Office of Research and Development, U.S. Environmental Protection Agency, Wash- ington, DC. [Online]. Available: http://cfpub.epa.gov/ncea/raf/recordisplay. cfm?deid=13102 [accessed November 22, 2004]. EPA (U.S. Environmental Protection Agency). 2002a. Perchlorate Environmental Contamination: Toxicological Review and Risk Characterization. External Review Draft. NCEA-1-0503. National Center for Environmental Assess- ment, Office of Research and Development, U.S. Environmental Protection Agency, Washington, DC. [Online]. Available: http://cfpub1.epa.gov/ncea/ cfm/recordisplay.cfm?deid=24002 [accessed August 23, 2004] EPA (U.S. Environmental Protection Agency). 2002b. Report on the Peer Review of the U.S. Environmental Protection Agencyâs Draft External Review Docu- ment âPerchlorate Environmental Contamination: Toxicological Review and Risk Characterizationâ. EPA/635/R-02/003. National Center for Environmen- tal Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Washington, DC. [Online]. Available: http://www.epa. gov/ncea/pdfs/perchlorate/final_rpt.pdf [accessed August 23, 2004]. EPA (U.S. Environmental Protection Agency). 2003. Disposition of Comments and Recommendations for Revisions to "Perchlorate Environmental Contami- nation: Toxicological Review and Risk Characterization External Review Draft (January 16, 2002). Office of Research and Development, U.S. Environmental Protection Agency, Washington, DC. [Online]. Available: http://cfpub2.epa. gov/ncea/cfm/recordisplay.cfm?deid=72117 [accessed August 25, 2004]. Fabris, N., E. Mocchegiani, and M. Provinciali. 1995. Pituitary-thyroid axis and immune system: A reciprocal neuroendocrine-immune interaction. Horm. Res. 43(1-3):29-38. Fernandez Rodriguez, A., H. Galera Davidson, M. Salguero Villadiego, A. Moreno Fernandez, I. Martin Lacave, and J. Fernandez Sanz. 1991. Induction of thyroid proliferative changes in rats treated with antithyroid compound. Anat. Histol. Embryol. 20(4):289-298. Fernandez-Santos, J.M., M. De-Miguel, R. Gonzalez-Campora, M. Salguero- Villadiego, J.J. Cabrera, and H. Galera-Davidson. 2004. Ki-ras mutational analysis in rat follicular-cell proliferative lesions of the thyroid gland induced by radioactive iodine and potassium perchlorate. J. Endocrinol. Invest. 27(1):12-17. Fisher, J.W. 2000. Consultative Letter, AFRL-HE-WP-CL-2000-0035, Physiolog- ical Model for Inhibition of Thyroidal Uptake of Iodide by Perchlorate in the Rat. Memorandum to Annie M. Jarabek, National Center for Environmental
Animal Studies 157 Assessment (MD-52), U.S. Protection Agency, Research Triangle Park, NC, from Jeffrey W. Fisher, Air Force Research Laboratory, Wright-Patterson Air Force Base, OH. June 28, 2000. Frankel, B.J., and J. Sehlin. 1994. Effect of perchlorate on glucoseâstimulated insulin release and 45CaÂ²+ uptake in pancreatic islets from diabetic Chinese hamsters. Pancreas 9(5):550-557. Fukuda, H., K. Ohshima, M. Mori, I. Kobayashi, and M.A. Greer. 1980. Sequen- tial changes in the pituitary-thyroid axis during pregnancy and lactation in the rat. Endocrinology 107(6):1711-1716. Gauger, K.J., Y. Kato, K. Haraguchi, H.J. Lehmler, L.W. Robertson, R. Bansal, and R. Zoeller. 2004. Polychlorinated biphenyls (PCBs) exert thyroid hormone- like effects in the fetal rat brain but do not bind to thyroid hormone receptors. Environ. Health Perspect. 112(5):516-523. Geller, A.M. 2003. Revised Brain Morphometry Analysis Incorporating Consul- tant in Veterinary Pathology (2003) Review of Morphometry Data from Argus 1416-003. Memorandum to Annie M. Jarabek, National Center for Environ- mental Assessment, U.S. Environmental Protection Agency, Research Triangle Park, NC, from A.M. Geller, National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC. September 19, 2003. Glinoer, D. 1997. The regulation of thyroid function in pregnancy: Pathways of endocrine adaptation from physiology to pathology. Endocr. Rev. 18(3):404- 433. Goldey, E.S., and K.M. Crofton. 1998. Thyroxine replacement attenuates hypothyroxinemia, hearing loss, and motor deficits following developmental exposure to Aroclor 1254 in rats. Toxicol. Sci. 45(1):94-105. Goldey, E.S., L.S. Kehn, C. Lau, G.L. Rehnberg, and K.M. Crofton. 1995a. Developmental exposure to polychlorinated biphenyls (Aroclor 1254) reduces circulating thyroid hormone concentrations and causes hearing deficits in rats. Toxicol. Appl. Pharmacol. 135(1):77-88. Goldey, E.S., L.S. Kehn, G.L. Rehnberg, and K.M. Crofton. 1995b. Effects of developmental hypothyroidism on auditory and motor function in the rat. Toxicol. Appl. Pharmacol. 135(1):67-76. Gomolla, M., G. Gottschalk, and H.C. Luttgau. 1983. Perchlorate-induced alter- ations in electrical and mechanical parameters of frog muscle fibres. J. Physiol. 343:197-214. Gould, E., A. Westlind-Danielsson, M. Frankfurt, and B.S. McEwen. 1990. Sex differences and thyroid hormone sensitivity of hippocampal pyramidal cells. J. Neurosci. 10(3):996-1003. Gravel, C., and R. Hawkes. 1990. Maturation of the corpus callosum of the rat: I. Influence of thyroid hormones on the topography of callosal projections. J. Comp. Neurol. 291(1):128-146. Greer, M.A., G. Goodman, R.C. Pleus, and S.E. Greer. 2002. Health effects assessment for environmental perchlorate contamination: The dose response
158 Health Implications of Perchlorate Ingestion for inhibition of thyroidal radioiodine uptake in humans. Environ. Health Perspect. 110(9):927-937. Hard, G.C. 1998. Recent developments in the investigation of thyroid regulation and thyroid carcinogenesis. Environ. Health Perspect. 106(8):427-436. Harry, J. 2001. Re: Comments on Original Experimental Design, Study Perfor- mance, and Brain Morphometry Results of Argus Research Laboratories, Inc. 14 March 2001 Study (Protocol Number 1416-003) and Supplemental Materi- als Provided by Dr. Robert Garman, Consultants in Veterinary Pathology, Inc. Letter to Annie M. Jarabek, National Center for Environmental Assessment, U.S. Environmental Protection Agency, from Jean Harry, Acting Chief, Laboratory of Toxicology Neurotoxicology Group Leader, National Institute of Environmental Health Sciences, Research Triangle Park, NC. October 11, 2001. Hiasa, Y., Y. Kitahori, Y. Kato, M. Ohshima, N. Konishi, T. Shimoyama, Y. Sakaguchi, H. Hashimoto, S. Minami, and Y. Murata. 1987. Potassium perchlorate, potassium iodide, and propylthiouracil: Promoting effect on the development of thyroid tumors in rats treated with N-bis(2-hydroxypropyl)- nitrosamine. Jpn. J. Cancer Res. 78(12):1335-1340. Hill, R.N., L.S. Erdreich, O.E. Paynter, P.A. Roberts, S.L. Rosenthal, and C.F. Wilkinson. 1989. Thyroid follicular cell carcinogenesis. Fundam. Appl. Toxicol. 12(4):629-697. IARC (International Agency for Research on Cancer). 2001. Pp. 40-41 in Some Thyrotropic Agents, IARC Monographs on the Evaluation of Carcinogenic Risk to Humans, Vol. 79. Lyon, France: International Agency for Research on Cancer. Keil, D., A. Warren, M. Jenny, J. EuDaly, and R. Dillard. 1998. Effects of Ammo- nium Perchlorate on Immunotoxicological, Hematological, and Thyroid Parameters in B6C3F1 Female Mice. DSWA01-97-0008. Medical University of South Carolina, Charleston, SC. September 30, 1998. [Online]. Available: http://www.epa.gov/ncea/perchlorate/references2/documents/44966.pdf [accessed August 25, 2004.] Keil, D., A. Warren, M. Jenny, J. EuDaly, and R. Dillard. 1999. Effects of Ammo- nium Perchlorate on Immunotoxicological, Hematological, and Thyroid Parameters in B6C3F1 Female Mice, Final Report. DSWA01-97-0008. Medical University of South Carolina, Charleston, SC. June 19, 1999. [On- line]. Available: http://www.epa.gov/ncea/perchlorate/references2/documents/ 99555.pdf [accessed July 16, 2004]. Kessler, M.E., and H.L. Kruskemper. 1966. Experimental thyroid tumors caused by long-term potassium perchlorate administration. [in German]. Klin Wochenschr. 44(19):1154-1156. Klecha, A.J., A.M. Genaro, A.E. Lysionek, R.A. Caro, A.G. Coluccia, and G.A. Cremaschi. 2000. Experimental evidence pointing to the bidirectional interac- tion between the immune system and the thyroid axis. Int. J. Immuno- pharmacol. 22(7):491-500.
Animal Studies 159 Klintsova, A.Y., C.R. Goodlett, and W.T. Greenough. 2000. Therapeutic motor training ameliorates cerebellar effects of postnatal binge alcohol. Neuro- toxicol. Teratol. 22(1):125-132. Knipper, M., C. Zinn, H. Maier, M. Praetorius, K. Rohbock, I. Kopschall, and U. Zimmermann. 2000. Thyroid hormone deficiency before the onset of hearing causes irreversible damage to peripheral and central auditory systems. J. Neurophysiol. 83(5):3101-3112. Koibuchi, N., S. Yamaoka, and W.W. Chin. 2001. Effect of altered thyroid status on neurotrophin gene expression during postnatal development of the mouse cerebellum. Thyroid 11(3):205-210. Kornblatt, M.J., A. Al-Ghanim, and J.A. Kornblatt. 1996. The effects of sodium perchlorate on rabbit muscle enolase--Spectral characterization of the mono- mer. Eur. J. Biochem. 236(1):78-84. Larsson-Nyren, G. 1996. Perchlorate is hypoglycemic by amplifying glucose- stimulated insulin secretion in mice. Acta Physiol. Scand. 158(1):71-76. Larsson-Nyren, G., J. Sehlin, P. Rorsman, and E. Renstrom. 2001. Perchlorate stimulates insulin secretion by shifting the gating of L-type CaÂ²+ currents in mouse pancreatic B-cells towards negative potentials. Pflugers Arch. 441(5):587-595. Lasky, R.E., J.J. Widholm, K.M. Crofton, and S.L. Schantz. 2002. Perinatal exposure to Aroclor 1254 impairs distortion product otoacoustic emissions (DPOAEs) in rats. Toxicol. Sci. 68(2):458-464. Lauder, J.M. 1977. The effects of early hypo- and hyperthyroidism on the devel- opment of rat cerebellar cortex. III. Kinetics of cell proliferation in the external granular layer. Brain Res. 126(1):31-51. Lawrence, J.E., S.H. Lamm, S. Pino, K. Richman, and L.E. Braverman. 2000. The effect of short-term low-dose perchlorate on various aspects of thyroid func- tion. Thyroid 10(8):659-663. Legrand, J. 1986. Thyroid hormone effects on growth and development. Pp. 503- 534 in Thyroid Hormone Metabolism, G. Hennemann, ed. New York: Marcel Dekker. Luster, M.I., A.E. Munson, P.T. Thomas, M.P. Holsapple, J.D. Fenters, K.L. White, Jr., L.D. Lauer, D.R. Germolec, G.J. Rosenthal, and J.H. Dean. 1988. Devel- opment of a testing battery to assess chemical-induced immunotoxicity: National Toxicology Programâs guidelines for immunotoxicity evaluation in mice. Fundam. Appl. Toxicol. 10(1):2-19. Magara, F., L. Ricceri, D.P. Wolfer, and H.P. Lipp. 2000. The acallosal mouse strain I/LnJ: A putative model of ADHD? Neurosci. Biobehav. Rev. 24(1):45- 50. Mahle, D.A., K.O. Yu, L. Narayanan, D.R. Mattie, and J.W. Fisher. 2003. Changes in cross-fostered Sprague-Dawley rat litters exposed to perchlorate. Int. J. Toxicol. 22(2):87-94. McClain, R.M. 1995. Mechanistic considerations for the relevance of animal data on thyroid neoplasia to human risk assessment. Mutat. Res. 333(1-2):131-142.
160 Health Implications of Perchlorate Ingestion Merrill, E.A. 2001. Consultative Letter AFRL-HE-WP-CL-2001-0008. PBPK Model for Perchlorate-induced Inhibition of Radioiodide Uptake in Humans. Memorandum with attachments to A. Jarabek, National Center for Environ- mental Assessment, Washington, DC, from E.A. Merrill, Air Force Research Laboratory, Wright-Patterson AF, OH. June 5, 2001. Merrill, E.A., R.A. Clewell, J.M. Gearhart, P.J. Robinson, T.R. Sterner, K.O. Yu, D.R. Mattie, and J.W. Fisher. 2003. PBPK predictions of perchlorate distribu- tion and its effect on thyroid uptake of radioiodide in the male rat. Toxicol. Sci. 73(2):256-269. Mestas, J., and C.C. Hughes. 2004. Of mice and not men: Differences between mouse and human immunology. J. Immunol. 172(5):2731-2738. Mooij, P., H.J. de Wit, A.M. Bloot, M.M. Wilders-Truschnig, and H.A. Drexhage. 1993. Iodine deficiency induces thyroid autoimmune reactivity in Wistar rats. Endrocrinology 133(3):1197-1204. Ng, L., R.J. Goodyear, C.A. Woods, M.J. Schneider, E. Diamond, G.P. Richardson, M.W. Kelley, D.L. Germain, V.A. Galton, and D. Forrest. 2004. Hearing loss and retarded cochlear development in mice lacking type 2 iodothyronine deiodinase. Proc. Natl. Acad. Sci. U.S.A. 101(10):3474-3479. Obregon, M.J., C. Ruiz de Ona, R. Calvo, F. Escobar del Rey, and G. Morreale de Escobar. 1991. Outer ring iodothyronine deiodinases and thyroid hormone economy: Responses to iodine deficiency in the rat fetus and neonate. Endo- crinology 129(5):2663-2673. Olton, D.S., J.T. Becker, and G.E. Handelmann. 1979. Hippocampus, space and memory. Behav. Brain Sci. 2:313-365. Pajer, Z., and M. Kalisnik. 1991. The effect of sodium perchlorate and ionizing irradiation on the thyroid parenchymal and pituitary thyrotropin cells. Oncol- ogy 48(4):317-320. Rami, A., A.J. Patel, and A. Rabie. 1986. Thyroid hormone and development of the rat hippocampus: Morphological alterations in granule and pyramidal cells. Neuroscience 19(4):1217-1226. Ribak, C.E. 1986. Contemporary methods in neurocytology and their application to the study of epilepsy. Adv. Neurol. 44:739-764. Robinson Jr., J.B., J.M. Strottmann, and E. Stellwagen. 1983. A globular high spin form of ferricytochrome c. J. Biol. Chem. 258(11):6772-6776. Roegge, C.S., V.C. Wang, B.E. Powers, A.Y. Klintsova, S. Villareal, W.T. Green- ough, and S.L. Schantz. 2004. Motor impairment in rats exposed to PCBs and methylmercury during early development. Toxicol. Sci. 77(2):315-324. Rovet, J.F. 2002. Congenital hypothyroidism: An analysis of persisting deficits and associated factors. Neuropsychol. Dev. Cogn. Sect. C. Child Neuropsy- chol. 8(3):150-162. RTI (Research Triangle Institute). 1999. Pp. 3-28 to 3-31 in Perchlorate Peer Review Workshop Report. EPA Contract No. 68-W98-085. RTI No. 7200- 019. Prepared for Office of Solid Waste, U.S. Environmental Protection Agency, Washington, DC, by Center for Environmental Analysis, Research
Animal Studies 161 Triangle Institute, Research Triangle Park, NC. May 13, 1999. [Online]. Available: http://www.epa.gov/ncea/perchlorate/references2/ documents/98020.pdf [accessed August 25, 2004.] SchrÃ¶der-van der Elst, J.P., D. van der Heide, J. Kastelijn, B. Rousset, and M.J. ObregÃ³n. 2001. The expression of the sodium/iodide symporter is up-regu- lated in the thyroid of fetuses of iodine-deficient rats. Endocrinology 142(9):3736-3741. Schwegler, H., W.E. Crusio, and I. Brust. 1990. Hippocampal mossy fibers and radial-maze learning in the mouse: A correlation with spatial working memory but not with non-spatial reference memory. Neuroscience 34(2):293-298. Seeger, G., U. Gartner, M. Holzer, and T. Arendt. 2003. Constitutive expression of p21H-Ras(Val12) in neurons induces increased axonal size and dendritic microtubule density in vivo. J. Neurosci. Res. 74(6):868-874. Sehlin, J. 1987. Effect of perchlorate on calcium uptake and insulin secretion in mouse pancreatic islets. Biochem. J. 248(1):109-115. Siglin, J.C., D.R. Mattie, D.E. Dodd, P.K. Hildebrandt, and W.H. Baker. 2000. A 90-day drinking water toxicity study in rats of the environmental contaminant ammonium perchlorate. Toxicol. Sci. 57(1):61-74. Spitzweg, C., W. Joba, W. Eisenmenger, and A.E. Heufelder. 1998. Analysis of human sodium iodide symporter gene expression in extrathyroidal tissues and cloning of its complementary deoxyribonucleic acids from salivary gland, mammary gland, and gastric mucosa. J. Clin. Endocrinol. Metab. 83(5):1746- 1751. Spitzweg, C., C.M. Dutton, M.R. Castro, E.R. Bergert, J.R. Goellner, A.E. Heufelder, and J.C. Morris. 2001. Expression of the sodium iodide symporter in human kidney. Kidney Int. 59(3):1013-1023. Steinmetz, J.E. 1996. The brain substrates of classical eyeblink conditioning in rats. Pp. 89-114 in The Acquisition of Motor Behavior in Vertebrates, J.R. Bloedel, T.J. Ebner, and S.P. Wise, eds. Cambridge, MA: MIT Press. Tazebay, U.H., I.L. Wapnir, O. Levy, O. Dohan, L.S. Zuckier, Q.H. Zhao, H.F. Deng, P.S. Amenta, S. Fineberg, R.G. Pestell, and N. Carrasco. 2000. The mammary gland iodide transporter is expressed during lactation and in breast cancer. Nat. Med. 6(8):871-878. TERA(Toxicology Excellence for Risk Assessment). 2001. Report on Five Expert Reviews of the Primedica 2001 Study Report (Hormone, Thyroid and Neurohistological Effects of Oral (Drinking Water) Exposure to Ammonium Perchlorate in Pregnant and Lactating Rats and in Fetuses and Nursing Pups Exposed to Ammonium Perchlorate During Gestation or Via Maternal Milk, March 2001). Prepared for Perchlorate Study Group. May 18, 2001. TERA (Toxicology Excellence for Risk Assessment). 2002. Pp. 6-7 in Quantita- tive Evaluation of Perchlorate Risk Assessment, February 2002. Toxicology Excellence for Risk Assessment, Cincinnati, OH. [Online]. Available: http://www.tera.org/Perchlorate/complete/quantitative%20eval%20perchlor ate.pdf [accessed July 16, 2004.].
162 Health Implications of Perchlorate Ingestion Versloot, P.M., J. Gerritsen, L. Boogerd, J.P. SchrÃ¶der-van der Elst, and D. van der Heide. 1994. Thyroxine and 3,5,3'-triiodothyronine production, metabolism, and distribution in pregnant rat near term. Am. J. Physiol. 267(6 Pt 1):E860- E867. Vranckx, R., M. Rouaze-Romet, L. Savu, P. Mechighel, M. Maya, and E.A. Nunez. 1994. Regulation of rat thyroxine-binding globulin and transthyretin: Studies in thyroidectomized and hypophysectomized rats given tri-iodothyro- nine or/and growth hormone. J. Endocrinol. 142(1):77-84. Wahlsten, D. 2002. Perchlorate Effects on Rat Brain Morphometry: A Critical Evaluation. Submitted to Eastern Research Group, Inc. for the U.S. EPA /ORD Peer Review Workshop-Perchlorate Environmental Contamination: Toxicological Review and Risk Characterization, March 5-6, Sacramento, CA. February 19. [Online]. Available: http://www.perchloratesymposium.com/ docs/Wahlsten2002.pdf [accessed August 23, 2004]. Wasniewska, M., F. De Luca, S. Siclari, G. Salzano, M.F. Messina, F. Lombardo, M. Valenzise, C. Ruggeri, and T. Arrigo. 2002. Hearing loss in congenital hypothalamic hypothyroidism: A wide therapeutic window. Hear Res. 172(1- 2):87-91. Widholm, J.J., B.W. Seo, B.J. Strupp, R.F. Seegal, and S.L. Schantz. 2003. Effects of perinatal exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin on spatial and visual reversal learning in rats. Neurotoxicol. Teratol. 25(4):459-471. Wolf, D.C. 2000. Report of the Peer Review of the Thyroid Histopathology From Rodents and Rabbits Exposed to Ammonium Perchlorate in the Drinking Water. Memorandum to Annie Jarabek and William Farland, National Center for Environmental Assessment, U.S. Environmental Protection Agency, Research Triangle Park, NC, from Douglas C. Wolf, Environmental Carcino- genesis Division, National Health and Environmental Effects Research Labora- tory, Research Triangle Park, NC. May 5. 2000. [Online]. Available: http://www.epa.gov/ncea/pdf/perchlorate/ea121000.pdf [accessed August 26, 2004]. Wolf, D.C. 2001. Erratum to the Report of the Peer Review of the Thyroid Histopathology From Rodents and Rabbits Exposed to Ammonium Perchlorate in Drinking Water. Memorandum to Annie Jarabek and William Farland, National Center for Environmental Assessment, U.S. Environmental Protection Agency, Research Triangle Park, NC, from Douglas C. Wolf, Environmental Carcinogenesis Division, National Health and Environmental Effects Research Laboratory, Research Triangle Park, NC. October 26, 2001. York, R.G., W.R. Brown, M.F. Girard, and J.S. Dollarhide. 2001a. Two-genera- tion reproduction study of ammonium perchlorate in drinking water in rats evaluates thyroid toxicity. Int. J. Toxicol. 20(4):183-197. York, R.G., W.R. Brown, M.R. Girard, and J.S. Dollarhide. 2001b. Oral (drinking water) developmental toxicity study of ammonium perchlorate in New Zealand white rabbits. Int. J. Toxicol. 20(4):199-205. Yu, K.O., L. Narayanan, D.R. Mattie, R.J. Godfrey, P.N. Todd, T.R. Sterner, D.A.
Animal Studies 163 Mahle, M.H. Lumpkin, and J.W. Fisher. 2002. The pharmacokinetics of perchlorate and its effect on the hypothalamus-pituitary-thyroid axis in the male rat. Toxicol. Appl. Pharmacol. 182(2):148-159. Zoeller, T.R., A.L. Dowling, C.T. Herzig, E.A. Iannacone, K.J. Gauger, and R. Bansal. 2002. Thyroid hormone, brain development, and the environment. Environ. Health Perspect. 110(Suppl. 3):355-361.