Inhalation toxicity studies of cobalt sulfate in F344N rats and B6C3F1 mice

Inhalation toxicity studies of cobalt sulfate in F344N rats and B6C3F1 mice

FUNDAMENTAL AND APPLIED TOXICOLOGY (1990) 15,357-372 Inhalation Toxicity Studies of Cobalt Sulfate in F344/N Rats and 86C3Fl Mice J.R. BUCHER, M.R...

5MB Sizes 2 Downloads 89 Views

FUNDAMENTAL

AND APPLIED TOXICOLOGY

(1990)

15,357-372

Inhalation Toxicity Studies of Cobalt Sulfate in F344/N Rats and 86C3Fl Mice J.R. BUCHER, M.R.

ELWELL, M.B. THOMPSON, B.J. CHOU,* AND H. A. RAGAN*

R. RENNE,*

National Institute ofEnvironmental Health Sciences, National Toxicology Program, P.O. Box 12233, Research Triangle Park, North Carolina 27709; and *BattelleMemorial Institute, Pa& Northwest Laboratories, P.O. Bo-x 999, Richland. Washington 99352

Received December 31, 1989; acceptedApril 26, 1990 Inhalation Toxicity Studies of Cobalt Sulfate in F344/N Rats and B6C3FI Mice. BUCHER, J. R., ELWELL,

M. R., THOMPSON,

M. B., CHOU,

B. J.. RENNE,

R., AND RAGAN,

H. A. (1990).

Fundam. Appl. Toxicol. 15,357-372. Groups of 10 F344/N rats and B6C3Fl mice of each sex were exposed to cobalt sulfate heptahydrate aerosols of 0, 0.3, 1.0, 3.0, 10, or 30 mg/m3, 6 hr per day, 5 days per week, for 13 weeks. All rats and female mice and all but 2/10 male mice exposed at the top concentration survived to the end of the studies. Polycythemia was observed in exposed rats but not in mice. Sperm motility was decreased in mice exposed at 3 mg/m-’ (the lowest concentration evaluated) and at higher concentrations, and increased numbers of abnormal sperm and decreased testis and epididymal weights occurred in mice exposed to 30 mg/m’. Cobalt content in the urine of rats increased with increasing atmospheric cobalt exposure. Primary histopathologic effects were limited to the respiratory tract. Lesions in rats and mice included degeneration ofthe olfactory epithelium, squamous metaplasia ofthe respiratory epithelium, and intlammation in the nose; inflammation, necrosis, squamous metaplasia, ulcers (rats), and inflammatory polyps (rats) of the larynx: metaplasia of the trachea (mice); and fibrosis, histiocytic infiltrates, bronchiolar epithelial regeneration, and epithelial hyperplasia in the alveoli of the lung. The most sensitive tissue was the larynx, with squamous metaplasia observed in rats and mice at the lowest exposure concentration of 0.3 mg/m3. Thus. a no-observed-adverse-effectlevel was not reached in these studies. o 1990 Society ofToxicology.

Cobalt sulfate is a reddish, crystalline, watersoluble powder widely used in the electroplating and electrochemical industries as a coloring agent for ceramics and as a drying agent in inks, paints, varnishes, and linoleum (De Bie and Doyen, 1962). It has been estimated that approximately 1000 workers worldwide are routinely exposed to soluble cobalt salts (Johnston, 1988). There is little information available concerning the toxic effects in humans of inhaled soluble cobalt salts, but pneumoconiosis has been reported in humans exposed in the cobalt-tungsten carbide “hard metal” industry to inhaled cobalt metal dusts or fumes at concentrations between 0.1 and 2 357

mg/m3 (Domingo, 1989). The threshold limit value/time-weighted average for cobalt metal is 0.05 mg/m3 (ACGIH, 1988). Animals have been exposed to aerosols of various forms of cobalt. Results of these studies indicate that lung compliance is decreased and that electrical properties of the heart are affected as in beer-drinkers’ cardiomyopathy, a condition attributed in part to the addition of cobalt sulfate to beer (Mot-in and Daniel, 1967; Kerfoot et al., 1975; Smith, 1980). Inhaled soluble cobalt salts are rapidly cleared from the lung (Menzel et al., 1989). Cobalt is distributed to all tissues (Smith and Carson, 198 1). Tissue retention is not 0272-0590/90 $3.00 Copyright 0 1990 by the Society ofToxicology. All rights of reproduction in any form reserved.

358

BUCHER ET AL.

marked, but higher concentrations have been noted in the liver, kidney, spleen, and heart than in other organs. Excretion is primarily via the urine (Domingo et al., 1984a,b; Llobet et al., 1986). Administration of cobalt to rats inhibits heme synthesis (De Matteis and Gibb, 1977) and induces the enzyme heme oxygenase (Maines and Kappas, 1976). The combined effect of these actions is to deplete the cytochrome P450 in the liver. In contrast to its actions on heme synthesis in the liver, cobalt administration promotes polycythemia. This response is mediated by an increase in circulating erythropoietin (Miller et al., 1974). For a time, cobalt was used in humans to treat various anemias. One of the toxic side effects of this therapy was a depression in thyroid function (Paley et al., 1958). It has been proposed that cobalt interferes with binding of inorganic iodide to tyrosine in the thyroid gland. Effects of cobalt on thyroid function in other species are not as clear (Sederholm et al., 1968). The administration of cobalt salts to animals has also been reported to affect the testis. Sprague-Dawley rats maintained on diets containing 265 ppm cobalt for 98 days showed degenerative changes which were considered secondary to hypoxia (Mollenhaur et al., 1985). Cobalt sulfate was selected for toxicologic evaluation by the National Toxicology Program from a metals class study, as a representative soluble cobalt salt (Helmes et al., 1983). Inhalation was chosen as the route of exposure because of evidence of respiratory toxicity of cobalt in the hard metal industry and because inhalation is a relevant route of worker exposure. In addition to an assessment of anatomic pathology, specific studies of hematologic changes and blood concentrations of thyroid hormones were performed. Studies of sperm morphology and vaginal cytology were also conducted. METHODS

AND

MATERIALS

Chemical. Cobalt sulfate heptahydrate was obtained in one lot from Curtin Matheson Scientific (Kansas City,

MO). Analyses (infrared, ultraviolet, and visible spectroscopy, Karl Fisher water analysis) confirmed the identity; purity was approximately 99%. The impurity with the highest concentration (140 ppm) was nickel. Aerosol generation system. Cobalt sulfate aerosol was generated from an aqueous solution by nebulization using dried compressed air. The aerosol was heated to about 26°C to partially dry the particles and then was passed into a Nalgene settling tank to eliminate large particles and water droplets. Further drying was accomplished by diluting the aerosol with dry air and heating the aerosol to 45°C as it left the tank. The cobalt sulfate/ air stream entered a distribution tube and was injected into each chamber (Hazleton 2000, Lab Products, Inc.) with air multiplier pumps. The aerosol was diluted to the desired concentration within the chamber with air from the chamber air-conditioning system. Aerosol concentration monitoring. Real-time aerosol monitors (Model RAM- 1. GCA Environmental Instruments) were used to determine the concentration of the aerosol in the exposure chambers once every 20 min throughout the exposure period. The monitors were calibrated through the use of filter grab samples. Samples collected on filter paper were analyzed for cobalt by inductively coupled plasma analysis after extraction with dilute nitric acid. Daily mean exposures for 13-week studies are given in Table 1. Atmospheric concentrations are expressed in milligrams of cobalt sulfate heptahydrate. Cascade impactor samples were taken to determine aerosol size distribution. The mass median aerodynamic diameter of the aerosol for all exposures ranged from 0.83 to 1.10 pm. Cobalt sulfate hydration in the aerosol distribution line was determined by ultraviolet/ visible spectroscopy of a dissolved sample from a weighed filter sample. Hydration ratios of 7.66 (water to cobalt mole ratio) and 7.67 were determined for two samples taken during the studies. Chamber characterization. The uniformity of the aerosol concentration in each exposure chamber, with and without animals present, was measured once before the beginning and once during the study with a RAM-l monitor. The between-port variability, expressed as percentage relative deviation, was less than 4.4% for all measurements. The build-up and bleed-off times were similar for all chamber concentrations. The time to reach 90% of the target chamber concentration (T90) as well as the time to bleed off to 10% of the target concentration after the generator was discontinued (TIO) were both approximately 8 min. The stability ofthe chemical in the aerosol distribution line and chamber was evaluated by derivative spectrophotometry in the range 20-800 nm. The similar spectral features of all samples were indicative of the stability of the cobalt sulfate aerosol. Study design. Groups of 10 rats and 10 mice of each sex were exposed to air containing cobalt sulfate heptahy-

TOXICITY

OF COBALT

SULFATE

359

IN RATS AND MICE

TABLE 1 MEANCHAMBER

Target

CONCENTRATIONSOF

Concentration (wlm')

0 0.3

1 3 10 30

(a) (b) (c)

COBALTSULFATEHEFTAHYDRATEINTHE

Determined

0.300

0.990 2.93 9.95 30.0

Concentration Odm')

(a)

Cc) i 0.029

k f 2 f

0.067 0.275 0.579 1.64

Maximum Concentration (wlm')

0.672 1.63 3.79 13.2 36.3

Mean + standard deviation for approximately 880 determinations. Within 10% of target concentration. Less than the detectable value of 0.005 mgtm' for 88% of samples; found but were determined to be due to instrumental baseline drift.

drate at concentrations of0 (chamber controls), 0.3, 1,3, 10, or 30 mg/m3 (calculated on the basis ofthe anhydrous salt) 6 hr (plus T90) per day, 5 days per week for 13 weeks. Male and female F344/N rats and B6C3Fl (C57BL/ 6N. female X C3H/HeN MTV-, male) mice used in these studies were produced under strict barrier conditions at Taconic Farms, Inc. Animals were progeny of defined microflora-associated parents that were transferred from isolators to barrier-maintained rooms. Rats and mice were shipped to the study laboratory at 4 weeks of age, quarantined at the study laboratory for 2 weeks, randomized to study groups, and placed on study at 6 weeks of age. The NIH-07 rat and mouse ration diet (Zeigler Bras., Inc., Gardners, PA) was used for all studies. Clinical examinations, supplemental studies, and pathology. Animals surviving to the end of the studies were humanely killed with carbon dioxide. The brain, heart, right kidney, liver, lung, right testis, and thymus were weighed. Hematologic analyses in rats and mice were performed on samples of anticoagulated blood (EDTA) obtained from the retroorbital sinus. Serum was barvested from samples of blood collected from the same site and allowed to clot. Analyses included leukocyte, lymphocyte, segmented neutrophil, monocyte, basophil, eosinophil, erytbrocyte, reticulocyte, and platelet counts; hemoglobin concentration; mean corpuscular hemoglobin; mean corpuscular hemoglobin concentration; and mean cell volume. All data except those for reticuiocyte and differential counts were obtained by using an Ortho ELT-8 hematology analyzer. Leukocyte differential counts were performed on blood smears stained with Wright-Giemsa. Reticulocyte preparations were made using equal volumes of whole blood and New Methylene Blue. Smears of these preparations were evaluated using a Miller disk. Serum chem-

I3-WEEKINHALATIONSTLJDIES Mi nmum Concentration (w/m')

Percent of Sampies In Range (b)

0.095 0.156 0.22 6.53 21.2

occasionally

81 90 88 93 94

higher

values

were

istry analyses, urinalyses, and thyroid function tests were performed for rats only. Serum chemistry analyses included glucose, cholesterol, and triglyceride concentrations; total creatine kinase activity; and quantitation of the three isoenzymes of creatine kinase. Serum chemistry analytes and enzymes were measured using an Abbott VP analyzer (Abbott Diagnostics, Irving, TX) and standard assaysprovided by the manufacturer. Creatine kinase isoenzymes were separated by electrophoresis (agarose plates) and quantified using reagents (Titan Gel CPK) and equipment (densitometer) obtained from Helena Laboratories (Beaumont, TX). Thyroid function tests were performed by radioimmunoassay methods and included determination of triiodothyronine (T3), thyroxin (T4), and thyrotropin concentrations (Tri-Tab and Tetra-Tab kits, NML Organon Teknika Corp.) with a species-specific antibody and reagents supplied by NIADDK (Bethesda, MD). Serum concentrations of free T4 were determined with a solid-phase component RIA procedure obtained as a kit from Becton-Dickinson (Towson, MD). Urinalyses included volume, appearance, and specific gravity; urinary cobalt content was determined by inductively coupled plasma anaiysis. Urine samples were collected from rats housed individually in metabolism cages for 16 hr. Sperm morphology and vaginal cytology evaluations were performed for rats and mice exposed to 0,3, 10, or 30 ms/m3 according to Morkey et al. (1988). A necropsy was performed on all animals. Organs and tissues were examined for gross lesions. Tissues were preserved in 10% neutral-buffered formalin and routinely processed for preparation of histologic sections for microscopic examination. Lungs were inflated with fixative. Tissues examined microscopically for all control and top concentration exposure groups included adrenal glands, brain, bronchial lymph nodes, cecum, colon, duodenum, epididymis/seminal vesicles/prostate/testes or ova-

360

BUCHER ET AL.

ries/uterus, esophagus, eyes (if grossly abnormal), gallbladder (mice), gross lesions and tissue masses with regional lymph nodes, heart, ileum, jejunum, kidneys, larynx, liver, lungs and mainstem bronchi, mammary gland, mandibular and mesenteric lymph nodes, mediastinal lymph nodes, nasal cavity and turbinates, pancreas, parathyroid glands, pharynx, pituitary gland, preputial or clitoral gland (rats). rectum, salivary glands, skin, spleen, sternabrae including marrow, stomach, thymus. thyroid gland, trachea, and urinary bladder. Tissues examined for all other groups (except 0.3 mg/m3 male mice) included gross lesions, larynx. lungs, and nose; larynx and lungs only examined for the 0.3 mg/m3 male mice. In addition, mediastinal lymph nodes, spleen, testes, thymus, and tracheas were examined for 10 mg/m3 male mice; mediastinal lymph nodes and trachea were examined for 10 mg/m3 female mice; and mediastinal lymph nodes were examined for 3 mg/m3 female mice. A transverse section was made through the larynx of rats caudal to the thyroid cartilage; 4 to 6 step-sections were prepared to ensure that the base of the epiglottis was present for examination. Details of pathology and review procedures have been described by Maronpot and Boorman (1982) and Boorman et al. (1985). Statistical methods. The analysis of organ weight, serum chemistry, hematology, urinalysis, and male reproductive system data was carried out by using the nonparametric multiple comparison procedures of Dunn ( 1964) or Shirley (1977) to assess the significance of pairwise comparisons between dosed and chamber control groups. Jonckheere’s (1954) test was used to evaluate the significance of dose-response trends and to determine whether Dunn’s or Shirley’s test was more appropriate for pairwise comparisons. The proportion of time spent in each stage of the estrous cycle was compared by using the Wilks criterion statistic (Wilks, 1932) ofthe multivariate procedure, which was performed after an arc sine transformation of the data. Methods for other endpoints are identified in the tables.

RESULTS The concentrations selected for use in the 13week studies were chosen based on results of 16&y studies which are reported in detail elsewhere (NTP, 1990). Briefly, these studies demonstrated deaths of rats and mice exposed to concentrations of cobalt sulfate heptahydrate of 50 mg/m3 and higher that resulted from severe inflammation and epithelial necrosis at all levels of the respiratory tract. 13-Week Studies-Rats All rats lived to the end of the studies. Mean body weights of male rats exposed to

225 2001

FIG. 1. Growth curves for male (top) and female (bottom) rats exposed to cobalt sulfate heptahydrate by inhalation for 13 weeks.

30 mg/m3 were lower than those of controls throughout the study (Fig. 1). Mean body weights of exposed female rats were not different than those of controls after Week 1. Compound-related clinical signs included ruffled fur in both sexes and hunched posture in male rats exposed to 30 mg/m’. The absolute lung weights and the lung weight to body weight ratios were significantly increased at the end of the study for rats exposed to 1 mg/ m3 or higher concentrations (Table 2). Polycythemia developed at 10 and 30 mg/ m3 in male and female rats and at 3 mg/m3 in male rats. Significant increases occurred in erythrocytes, in the mean hemoglobin concentration, and in the hematocrit value (Table 3). The reticulocyte count was significantly increased in female rats exposed to 30 mg/m3. The platelet count was significantly decreased in rats exposed to 10 or 30 mg/m3. No changes were found in the leukocyte or differential counts. Mean serum cholesterol concentrations were significantly decreased in males exposed

TOXICITY

OF

COBALT

SULFATE TABLE

LUNG

AND

BODY

WEIGHTS

OF RATS

OF COBALT organ

Control

0.3

RATS

AND

361

MICE

2

IN THE

SULFATE

mglm'

IN

I ~-WEEK

INHALATION

STUDIES

HEPTAHYDRATE’

1 mglm'

10 mgim'

3 mglm'

30 mgtm'

Male Body

weight(grams)

Lung Absolute Relative

331 + 8.0

1,364 4.1

+ 45 f 0.11

330 f

1,451 '4.4

f

8.9

33

* 0.11

325 f

*I.506 '*4.6

6.2

f

53

2 0.11

328 2 10.4

**1,690 l *5.1

f 79 f. 0.12

327 f

l *1,951

8.4

f 78 f 0.11

"'6.0

**282

f 8.8

**2,008 "'7.1

2 75 f 0.11

Female Body

wefght(grams)

Lung Absolute Relative

(a)

188 f

4.7

935 k 28 5.0 f 0.10

Mean f standard error in P values vs. the controls 'P<0.05 l *p
173 k 4.8

904 5 18 5.2 + 0.10

milligrams by Dunn's

181 f

*1.035 l *5.7

(absolute) test (Dunn,

5.9

f 24

+ 0.11

or

**I,282 **6.6

milligrams per or Shfrley's

1964)

to 10 or 30 mg/m3 and in females exposed to 30 mg/m3 (Table 3). There were no consistent dose-related effects on glucose concentrations, activities of total creatine kinase or the relative activities of each of three creatine kinase isoenzymes present (data not shown), or in serum triglyceride concentrations. Concentrations of T3 were significantly lower in female rats exposed to 10 and 30 mg/ m3 and those of thyrotropin were significantly lower in male rats at 30 mg/m3, but the T4 and thyrotropin concentrations for female rats and T3 and total and free T4 concentrations for male rats were not consistently dose-related (Table 3). Granular casts were observed in the urine from many exposed male rats (3-7 animals per group of lo), whereas none were observed in the urine from controls. A dose-related increase was seen in the number of epithelial cells in the urine from males that were exposed to 3 mg/m3 or more. Urine volumes collected over 16 hr were variable and not statistically different from those for controls except in the 30 mg/m3 females, which aver-

196 + 4.8

f 38 + 0.13

191 f 4.2

l *1,344

gram (relative) test (Shirley,

f 40 + 0.16

**7.0

for

groups

175 f 2.8

**1.573 -9.0

of

+ 47 ! : 0.21

10 animals:

1977).

aged approximately two-thirds that of controls. The amount of cobalt excreted in urine over 16 hr was related to exposure to cobalt sulfate (Table 4). No statistically significant effects on sperm motility, sperm counts, or the incidence of abnormal sperm were observed in exposed rats, The average estrous cycle of females exposed to 30 mg/m3 (5.0 4 0.2) was longer (but not significantly) than that of controls (4.6 Ik 0.2). Compound-related lesions were limited to the respiratory tract of rats of each sex exposed to cobalt sulfate. Lesions were concentration-related and generally similar in incidence and severity in males and females (Table 5). In the nose, hyperplasia and squamous metaplasia of the respiratory epithelium were seen primarily at the two highest exposure concentrations. This was most prominent in the tips of the naso- and maxilloturbinates and on the lateral wall of the nasal cavity in the most anterior section of the nose. Degeneration of the olfactory epithelium was characterized by a thinning of the olfactory epi-

362

BUCHER

ET AL.

TABLE 3 SELECTEDHEMATOLOGICANDSERUMCHEMISTRYDATAFORRATSINTHE ~NHALATIONSTUDIESOFCOBALTSULFATEHEFTAHYDRATE" Analysis

Control

0.3 mglm'

13-WEEK

3 mgh?

1 mglm'

10 mglm'

30 mglm'

MALE Hemoglobln(g/dl) Mean corpuscular hemoglobin(pg) Mean corpuscular hemoglobin concentratlon(g/dl) Mean cell volume(f1) Platelets(lO'/~l) Erythrocytes(lO'/~l) Reticulocytas(lOel~1) Hematocrit(percent) Cholerterol(mgld1) Triiodothyronlne(ngld1) Free thyroxin(ngldl) Total thyroxln(pl/dl) Thyrotropin(nglm1)

14.9

f 0.19

14.5

+ 0.18

15.1 f 0.06

16.3

f 0.03

*16.1

f 0.07

16.2

f 0.04

32.4 50.1 543 9.2 0.08 46.0 72.3 66.5 1.8 3.85 187

t f. * f f f t t + f *

+ + f i f f f + ?: f +

32.3 50.4 536 9.3 0.11 46.8 71.3 76.6 1.9 4.28 202

t t * f + f + f f t f

0.09 0.18 a.7 0.12 0.013 0.56 2.48 5.00 0.08 0.115 45.5

32.3 49.7 526 9.0 0.08 44.9 67.4 62.1 1.8 3.72 93

0.10 0.21 11.7 0.09 0.015 0.48 2.60 3.95 0.07 0.135 36.7

l *17.1

+ 0.16

l 15.3 ?r 0.80

“15.9

f 0.04

30.7 49.4 525 l *9.6 0.10 ‘47.8 65.4 67.6 '2.0 4.17 143

‘32.9 '*48.2 **434 "10.8 0.07 “52.1 ‘*61.9 74.4 1.9 4.06 195

f + f f f f f lr f + f

"16.8

f 0.10

14.7

0.09 0.22 10.5 0.04 0.010 0.20 5.65 5.07 0.12 0.285 54.4

f 0.80

f f f f + t f f + + +

1.57 0.31 8.6 0.07 0.0009 0.44 2.64 3.89 0.06 0.129 36.0

0.12 0.13 25.7 0.10 0.011 0.48 2.93 5.49 0.06 0.136 54.7

*‘lg.6

f 0.15

*16.2

+ 0.07

"33.1 **49.2 "382 **12.1 0.08 **59.4 l *51.4 64.3 2.0 4.10 '56

t f + ?: + k f t A + +

0.12 0.33 9.4 0.12 0.013 0.47 2.33 4.40 0.11 0.252 27.0

“19.4

+ 0.11

FEMALE Hemoglobin(gfd1) 15.2 Mean corpuscular hemoglobtn(pg) 17.3 Mean corpuscular hemoglobin concentration(g/dl) 32.9 Mean cell valume(f1) 52.4 Platelets(lO'/pl) 643 Erythrocytes(lO'/pl) 8.8 Reticulacytes(lO'/~l) 0.08 Hematocrit(percent) 46.2 Cholasterol(mg/dl) 123.4 Triiodothyronlne(ng/dl) 96.4 Free thyroxin(ng/dl) 1.2 Total thyrorin(~l/dl) 3.44 Thyrotropln(nglm1) 134

(a)Mean f. standard Shirley's test l Pt0.05 ** P
* 0.20

*15.6

5 0.10

15.1 k 0.22

15.5

2 0.22

+ 0.07

17.2

i 0.06

17.2 + 0.04

17.3

f. 0.03

f f f f + f + + f * *

0.08 0.16 16.3 0.10 0.013 0.59 4.62 5.56 0.10 0.180 17.4

error for groups (Shirley, 1977).

32.6 52.5 619 l 9.1 0.09 '47.9 116.4 80.2 1.2 2.96 195

t + f f. + f * f. f f *

0.10 0.17 8.8 0.07 0.009 0.35 4.43 4.55 0.07 0.169 30.7

of 10 animals;

32.9 52.4 638 a.0 0.08 46.1 '107.5 65.2 1.3 3.31 171

P values

f + i f f + + * f ?r f

are

0.14 0.22 20.3 0.14 0.011 0.75 3.72 4.07 0.09 0.220 19.6

vs.

32.9 52.4 614 9.0 0.08 47.2 114.3 88.9 1.2 3.39 128

the

controls

f f + 5 f 2 f * f * *-

17.2

0.13 0.27 8.4 0.11 0.011 0.78 6.12 4.35 0.08 0.135 17.7

f 0.04

33.2 f 0.21 '51.6 + 0.27 l *569 f 14.4 l *9.e * 0.05 0.09 k 0.010 "50.6 + 0.43 113.9 * 5.08 *79.7 f 5.45 1.1 * 0.08 2.96 f 0.175 143 2 32.9

by Dunn's

test

(Dunn,

17.4 e33.4 52.0 '*512 "11.2 '0.16 “58.1 “991.3 “60.1 1.2 3.01 107

1964)

k 0.05 f f ? f f f + t + f +

or

TABLE 4 COBALTCONTENTINURINEOFRATSINTHE~~-WEEKINHALATIONSTUDIES OFCOBALTSULFATEHEPTAHYDRATE' Control

0.3

mQln~

1 mglm'

3 mglm'

10 mgtn'

30 mglm'

Male

0.22

f

0.03

2.51

+z 0.23

5.21

+ 0.34

33.4

f

5.15

42.6

+ 7.6

Female

0.17

+ 0.05

1.99

2 0.47

2.36

f

18.1

f

1.23

21.4

f

(a)Hicrograms groups vs. lndivfdual

excreted per the controls values.

0.28

16 hours; mean + standard by Dunnett's test (Dunnett,

error for groups 1955) performed

of 10 animals; using a log

1.64

105 66.9

PtO.O1 for transformation

* 11.8 zk

all

4.0

dose of the

0.14 0.21 10.8 0.05 0.028 0.43 5.50 4.27 0.07 0.336 22.8

TOXICITY

OF

COBALT

SULFATE TABLE

IN

RATS

AND

363

MICE

5

NUMBERSOFRATSWITHSELE~EDLESIONSINTHEI~-WEEKINHALATIONSTUDIES OFCOBALTSULFATEHEPTAHYDRATE" SitelLeslon

Control

0.3

mglm'

1 mglrnl

3 mg/m'

10 mglm'

30 mglm'

MALE Nose Inflammation Olfactory epithel ium degeneration Respiratory epithelium hyperplasia Respiratory epithelium squamous metaplasia

0 0 0

0 0 0

0 0 1

2

'5

0

0

0

*5

**9

Larynx(step sections) Mineralization Inflammation Ulcer Necrosis Inflammatory polyp Squamous metaplasia

0 0 0 0 0 0

(bN (b)2 (b)O (b)l (b)O

0 *T! 0 0 0 **lo

**lo +*9 **7 '*lo **lo '*lo

‘*lo

Lung Histlocytic infiltrates Inflammation Fibrosis Bronchlolar eplthelfum Bronchjiolar ectasla Alveolar emphysema Alveolar eptthelium

1 0 0 0 0 0 0

0 0 0 0 0 0 0

**lo

“10 “*lo “‘10

0 0

0 0

0 0

0

0

0

regeneration

hyperplasia

**(b)9

3

0

"'7

0 0 2 **lo

**10

l5 0 0 0 0 3

“10

l +9 ""7 **lo -3 “10

*‘lo 1 0 **a

**7 **lo

1

2

"*6

"6

0 3

l *6 **9

“‘10 -9

1

'"3

"6

1

**a **10

“10 **10

0 2 1 "10

3 *+9 '"10 "10

“10 **9 “‘10

'*lo **9

**lo **10

‘*lo **10

FEMALE Nose Olfactory epithelium Respiratory epithelial Respiratory epithelium metaplasla

degeneration hyperplasla squamous

Larynx(step sections) Mineralization Inflammatfon Ulcer Necrosis Inflammatory polyp Squamous metaplasla Lung Histiocytic infiltrates Inflammation Fibrosis Bronchlolar epithelium regeneration Bronchiolar ectasia Alveolar emphysema Alveolar epithelium hyperplasla (a)Ten rats were examined in each group (b)Nine rats were examined. (c)Eight rats were examined. *Pt0.05 vs controls by Fisher exact test **PtO.Ol vs controls by Fisher exact test

0

(c)2

0 0 0

(c)O (C)O (c)O

1

-7

0

'"10

3 0 0 0 0 0

**10 2 0 0 0 0 0

0 otherwise

thelial cell layer in the dorsal meatus and also on the nasal septum in the ethmoid region (degeneration was slightly more prominent in males).

**10 0 0 0

**(c)7

0 0 0 0 0 0 unless

0

(C)O

1

1 0 2 1 3

'4 0 "8

l *fJ

*5 +5 “10

2 1

"7 1

specified.

At the higher exposure concentrations, inflammatory polyps were seen in the larynx of most rats (Figs. 2 and 3). Polyps were consistently located at the base of the epiglottis and

364

BUCHER ET AL.

FIG. 2. Transverse section through base of the epiglottis of a male rat exposed to 10 mg/m3 cobalt heptahydrate for 13 weeks. Large inflammatory polyp (arrows) arising from the caudal surface of epiglottis extends into lumen(L) oflarynx. (Hematoxylin and eosin, 66X.)

extended into the lumen of the larynx. These polyps had a fibrovascular stroma, which was covered by a well-differentiated squamous epithelium. Focal areas of necrosis and ulceration were frequently present in the epithelium of the polyp. Chronic inflammation and mineralization were prominent in the stroma of the polyp. At the lower concentrations at which polyps did not occur, squamous metaplasia of the laryngeal respiratory epithelium and chronic inflammation in the stroma persisted. At 0.3 mg/m3, the severity of the metaplasia and inflammation was minimal to mild. Regeneration of bronchiolar epithelium with dilatation (ectasia) of bronchioles was observed in the lungs of rats exposed to 30 mg/m3. Regeneration was characterized by the loss of the apical bleb of cytoplasm normally present on Clara cells, and the presence of mitotic figures in some of these cells. At the

higher concentrations, particularly in males, alveoli were lined by a cuboidal Type II epithelium (alveolar hyperplasia). Distension or disruption of alveolar septa (emphysema) was also present; fibrosis was observed around bronchioles and within alveolar septae. Histiocytic infiltration, characterized by intraalveolar accumulation of macrophages and infiltration of alveolar septae with mononuclear inflammatory cells also occurred at this exposure concentration. At lower concentrations, only intraalveolar histiocytic infiltrates and inflammation were present. Lymphoid hyperplasia was present in the mediastinal lymph nodes of exposed rats, but the incidence was not concentration related. Cardiomyopathy was seen in 3/ 10 control and 3/ 10 male rats exposed to 30 mg/m3; the severity was marginally increased in the exposed group (minimal-mild vs minimal). Cardio-

TOXICITY

OF COBALT

SULFATE

IN RATS AND MICE

365

FIG. 3. Detail of another section of the polyp from Fig. 2 shows well-differentiated keratinized squamous epithelium over fibrous stroma that contains inflammatory cells and a focus of mineralization (arrows) (Hematoxylin and eosin. 157X.)

myopathy of minimal severity was seen in I/ 10 female rats exposed to 30 mg/m3.

13-WeekStudies--Mice Two of 10 males exposed to 30 mg/m3 died before the end of the studies, one during the 2nd week and one during the 12th week. Mean body weights of mice exposed to 30 mg/m3 were lower than those of controls throughout the study (Fig. 4). Feed consumption was not measured. No observed clinical signs appeared to be related to cobalt sulfate exposure, with the exception of rapid breathing and skin discoloration in one high exposure concentration male mouse that died during Week 12. The absolute lung weight and the lung weight to body weight ratios were significantly increased in the 10 and 30 mg/m3

exposure groups, and the absolute testis weight and the testis weight to body weight ratios were significantly decreased for males exposed to 30 mg/m3 (Table 6). No consistent or dose-related hematologic effects were observed. The epididymal weight was significantly lower than that of controls for male mice exposed to 30 mg/m3. The number of abnormal sperm in mice exposed to 30 mg/m3 was significantly increased, and sperm motility was significantly reduced in mice exposed to 3, 10, or 30 mg/m3 (Table 7). The estrous cycle was significantly longer in mice exposed to 30 mg/m’. Compound-related lesions were generally limited to the respiratory tract of mice of each sex. Lesions were concentration-related and similar in incidence and severity in males and females (Table 8). In the nose, degeneration

366

BUCHER ET AL.

ent. Chronic inflammation occurred primarily at the highest exposure concentration and was composed of fibrosis around bronchioles and in alveolar septae along with an inflammatory cell infiltrate. At the lower concentration, only a minimal increase in macrophages was seen in the alveoli. Lymphoid hyperplasia was present in the mediastinal lymph nodes of mice at the 30 mg/m3 exposure concentration. At the highest exposure concentration, atrophy of the testis, which consisted of a loss of germinal epithelium in the seminiferous tubules, was observed, more severely affected testes also contained foci of mineralization. DISCUSSION

FIG. 4. Growth curves for male (top) and female (bottom) mice exposed to cobalt sulfate heptahydrate by inhalation for 13 weeks.

of olfactory epithelium, squamous metaplasia of the respiratory epithelium, and an acute inflammatory cell exudate in the nasal cavity were seen primarily at the two highest exposure concentrations. At the highest exposure concentration, necrosis, inflammation, and squamous metaplasia of the laryngeal epithelium were present in most mice (see Fig. 5). Some foci of necrosis in the laryngeal epithelium extended through the basement membrane into the underlying lamina propria. Squamous metaplasia of the respiratory epithelium in the trachea also occurred in mice in this exposure group. At exposure concentrations below 30 and squamous w/m3, only inflammation metaplasia were observed. In the lung of mice exposed to 10 or 30 mg/ m3, there was regeneration of bronchiolar epithelium and hyperplasia of the alveolar epithelium; these changes were similar to those seen in rats. Infiltration of histiocytes (macrophages) into the alveolar spaces was also pres-

The respiratory tract was the primary target of toxicity of inhaled cobalt sulfate in the these studies. The mean aerodynamic diameter of the cobalt sulfate heptahydrate aerosol particles was approximately 1 pm, well within the size range of particles shown to deposit at all levels of the respiratory tract of the rat (Raabe, 1980). Quite similar degenerative, inflammatory, and regenerative changes were present from the nasal cavity to the alveoli in rats and mice of each sex. The difference in the susceptibility of the various components of the respiratory system was consistent in rats and mice. The trachea showed metaplastic changes only in mice exposed to 30 mg/ m3, but squamous metaplasia of the larynx was seen in both rats and mice at concentrations as low as 0.3 mg/m3, the lowest exposure concentration studied. In rats exposed to 3 mg/m3 cobalt sulfate heptahydrate or more, quite remarkable inflammatory polyps were found, typically arising caudal to the base of the epiglottis in the larynx. These exophytic masses occupied up to half the laryngeal lumen and consisted of a hyperplastic squamous epithelium, with abundant vascular stroma. The larynx is a common site for lesions in rodents exposed by inhalation to various chemicals and pharmaceuticals; erosion,

TOXICITY

OF

COBALT

SULFATE TABLE

IN RATS

AND

367

MlCE

6

SELECTEDORGANWEIGHTSOFMICEINTHE 1%WEEKINHALATIONSTUDIES OFCOBALTSULFATEHEFTAHYDRATE' organ

Control

0.3

mglm’

1 mglm’

3 mglm’

10 mglm’

30 mg/m'

Male Body

weight(grams)

Lung Absolute Relative

37.5

37.1

2 1.28

39.9 f 1.28

f

1.54

181 f 4.9 f

4.3 0.13

179 2 9.6

186 + 6.5

187 * 4.2

4.6

4.7

5.2

It 1.9 2 0.11

125 + 2.7 3.4 f 0.07

f 0.18

f

0.08

35.7

f

0.88

zk 0.09

35.8

“‘213 “‘6.0

f 0.98

f 4.5 2 0.15

Testis Absolute Relative

(C)120 (c)3.3

123 + 2.3 3.1 k 0.09

120 + 2.4 3.4 f 0.10

121 f 3.4 r

**(b)32.5

k 0.81

l *(b)321 l *(b)9.9

k 6.7 2 0.32

2.1 0.05

“57 *‘I.7

f 6.8 k 0.19

f 0.59

Female Body

welght(grams)

Lung Absolute Rslatlve

(a)

33.2

+ 1.31

194 f 5.9 f

9.0 0.28

33.8

f

1.25

192 f 5.8 f

4.2 0.26

Mean f standard error In milligrams unless otherwise spedfled: P values (Shirley, 1977). (b) Eight animals were weighed. (c) Nine animals were weighed. *p<0.05 "P
(absolute) vs. the

34.7

* 1.33

187 + 4.7 5.4 t 0.12

33.3

f 0.94

31.6

k 0.74

**26.1

198 + 4.7 6.0 k 0.22

l *232

?r 7.3 f 0.11

**327 **12.6

**7.3

or milligrams per gram (relative) for controls by Dunn's test (Dunn, 1964) or

ulceration, and an inflammatory exudate are frequently observed (Gopinath et al., 1987). What is unusual about the lesion caused by cobalt sulfate inhalation is the apparent organization and vascularization of the inflammatory exudate and the subsequent squamous metaplasia over the surface of these fibrous masses. The pathogenesis of somewhat similar intraluminal fibrotic projections produced in the trachea and bronchioles of mice exposed to methyl isocyanate has been described by Boorman et al., (1987). It is proposed that these types of lesions arise from an area where the respiratory epithelium has been destroyed. A relatively slow regeneration of the epithelium must occur from the margin of the lesion, giving time for migration of fibroblasts into the exudate and further organization of the lesion (Basset et al., 1986). Klonne et al. (1987) observed polypoid protrusions in the larynx of F344 rats

groups Shirley's

of

10

-I 5.8 2 0.40

animals

test

exposed by inhalation to aerosols of an aqueous silane solution. These protrusions arose as part of a granulomatous reaction in response to embedded silane particles and while similar in location, they may differ somewhat in pathogenesis from those observed with cobalt sulfate. At the end of the 13-week studies, rats showed a pronounced polycythemia in males at exposure concentrations as low as 3 mg/ m3 and in females exposed at 10 mg/m3 or at higher concentrations. This appeared to be a simple erythrocytosis, as most other formed elements were within normal ranges. These changes are consistent with the well-characterized cobalt-induced polycythemia, which appears to be due to an increase in circulating erythropoietin (Taylor and Marks, 1978). No consistent significant hematologic effects were seen in mice. Species differences in the polycythemic response to cobalt have pre-

368

BUCHER

ET AL.

TABLE

7

DATAFORMICEINTHE~J-WEEKINHALATIONSTUDIES OFCOBALTSULFATEHEPTAHYDRATE'

REPRODUCTIVESYSTEM

Control

3 mglm'

10 mglm'

30 mglm"

MALE (a) Caudal welght(mg) Right epidldymal welght(mg) Sperm count(X 10') Sperm motility (percent) Abnormal sperm (percent) FEMALE

0.042 1.074 87.0 (c)1.29

f f f f

0.016

0.002 151 0.76 0.164

0.043 1,342 l *76.6 1.36

t 0.001 k k 2 f

0.017

0.001 140 2.44 0.113

0.045 1,136 -75.6 0.98

f 0.001 + f f ?r

0.014

0.001 86 2.25 0.105

f 0.001

**0.034

f 0.001 2 194

(b)776 **(b)46.6 **(b)3.80

+ 7.76 f 0.626

(d)

Estrous stage Proestrus Estrus Metestrus Oiestrus

(percent)

NC(e) Cycle

2 0.001

0.015

length

4.20

(days)

(a)Mean (b)Elght

2 standard animals

(c)Nine

animals

24.3 25.7 21.4 26.6 0.0 0 f .20

error for groups were examined.

were

18.6 25.7 27.1 26.6 0.0 4.11 f 0.11

of 10 animals

27.1 28.6 21.4 22.9 0.0 4.20 + 0.13

unless

otherwise

***5.00

17.1 32.9 22.9 25.7 1.4 f 0.24

Indicated.

examined.

(d)Dosa-related differences occurred in stages of the estrous stages (Wilks, (e) NC = not clear or no cells observed. "Pt0.01 vs. the controls by Dunn's test ***PtO.Ol vs. the controls by Dunnett's

the

relative

1932).

frequency

(Dunn, 1964) test (Dunnett.

viously been reported (Smith and Carson, 1981). There have been many reports of goiter as a side effect of cobalt therapy for anemia in humans (reviewed by Smith and Carson, 198 1), and this effect appears to be due to inhibition of iodine uptake by the thyroid gland. Thyroid function as indicated by serum triiodothyronine, thyroxin, and thyrotropin did not appear to be consistently affected in rats in the current studies. These results support the opinion expressed by Sederholm et al. ( 1968) that effects of cobalt on the thyroid gland have not been clearly demonstrated in studies with rats, mice, or rabbits. Cardiomyopathy appeared slightly more severe in male rats in the higher exposure group compared with the controls, but the incidences were the same. Minimal cardiomy-

of time

spent

in

drfferent

P-0.03. or

Shirley's

test

(Shirley,

1977).

1955).

opathy was seen in one 30 mg/m3 female rat and in no controls. Myocardial injury can also be indirectly assessed by measuring the activity in the serum of specific isozymes of creatine kinase ((X-2 and CK-3), which are released from damaged cardiac muscle cells (Boyd, 1983). In the current study, total serum creatine kinase activity was highly variable in the 30 mg/m3 group of female rats and appeared to be slightly increased in females (but not in males), although not statistically so. The CK-3 form of the enzyme appeared to be slightly increased in the high exposure concentration female rats, but the amount of CK-2 isozyme present was decreased. Thus, the data did not suggest a cardiotoxic effect. The motility of sperm appeared to be lower in exposed mice, and the number of abnormal sperm was increased, especially at the highest exposure concentration, at which

TOXICITY

OF

COBALT

SULFATE TABLE

NUMBERS

IN RATS

Control

0.3

369

MICE

8

OF MICE WITH SELECTED LESIONS IN THE 1~-WEEK OF COBALT SULFATE HEITAHYDRATE’

S1telLesion

AND

INHALATION 3 mglm3

STUDIES

mglm'

I mglm'

10 mglm"

30 mglm'

0 0

__-

0 0

‘*lo **9

*vJ

0

_-

0

**a

**a

Larynx Inflammation Necrosis Squamous metaplasia

0 0 0

0 0

0 0

Trachea Squamous

0

MALE NOW Inflanmlation Olfactory Respiratory metaplasia

Lung Histiocytic Chronic Bronchiolar Alveolar

eptthelium epithelium

degeneration squamous

metaplasia

InfIltrates inflammation eplthelium epithelium

Mediastinal Hyperplasia

lymph

"7

“10

regeneration hyperplasia

2

“10 0 0 0

0 0 0

“10

0

**10

**9

3

-9

l (b)5 __

"*lo 0 0 0

**9

1 0

(b)O (b)O

**10 “10 *‘lo **t3

1 0 3

nodes

__

0

Testis Atrophy Mineralization

l

(c)O

*(b)6

0 0

___

__ --

0 0

0 0

1 0

l4 1

"10 **IO

l *10

0

0

0

1

**9

**9

Larynx Inflammation Necrosl s Squamous metaplasia

0 0 0

0 0

0 0

Trachea Squamous

0

__ -_

0 0

"9 *4

FEMALE Nose Inflammatfon Olfactory Respiratory metaplasia

Lung Hlstiocytic Chronic Bronchiolar Alveolar

apithelium eplthelium

degeneration squamous

metaplasia

infiltrates inflammation epithelium epithelium

l *e

each

0 0 0 0

9roup

0 unless

"6

(b)O (b)O

0 **9

**(b)8

-_

regeneration hyperplasia

Mediastinal lymph nodes Hyperplasia (a)Ten mice were examined in ibjNine mice were examined. (c)Seven mice were examined. (d)Five mice were examined. (e)Slx mice were examined. 'Pt0.05 by Fisher exact test **PtO.Ol by Fisher exact test

**8

__ otherwise

*+9 0 0 0 __ specified:--indicates

**10

0

**10 "5 0

“10 0 0 0

**lo

(d)O tissues

"(b)8 "(b)6 **(b)9

not

(efl examfned.

3

**10 **10 **10 **10

"7

370

BUCHER

ET AL.

FIG. 5. Transverse section through base of the epiglottis of a female mouse exposed to 30 mg/m’ cobalt heptahydrate for 13 weeks. Inflammation and large pale staining areas of epithelial necrosis (arrows) and dark staining foci of mineralized cell debris are present in the laryngeal mucosa. Keratin and cell debris are within laryngeal lumen (L). (Hematoxylin and eosin, 66X.)

clear testicular atrophy occurred. The magnitude and number of these effects indicates that these changes represent a direct toxic effect of cobalt on the reproductive system; the site of action remains to be determined. Rats appeared much less susceptible to testicular toxicity of cobalt than mice. There was no indication microscopically of exposure-related kidney lesions in rats or mice in the 13-week studies; however, granular casts were observed in the urine from a number of male rats in all exposure groups, and a concentration-related increase in the number of epithelial cells sloughed into the urine was seen in exposed male rats. Both observations are suggestive of a minimal nephrotoxic effect. The excretion of absorbed cobalt is primarily via the urine. Urinary cobalt excretion was

measured in male and female rats and found to exhibit a concentration-dependent pattern, although the magnitude of the difference in urinary cobalt from one group to the next was not as large as was the difference in the atmospheric concentrations. Urinary cobalt concentrations have been measured in workers exposed to cobalt in the cobalt-tungsten carbide “hard metal” industry (Ichikawa et al., 1985). Concentrations as high as 0.39 mg/ml have been found in the urine of workers in certain high exposure areas. By comparison, the urine cobalt concentrations in the current rat studies ranged from 0.11 to 7.79 mg/ml in the various groups. The value obtained in male rats exposed to 1.O mg/m3 was 0.39 mg/ml, the same as that cited in the Ichikawa et al. study. Exposure to cobalt in the hard metal industry is probably to the co-

TOXICITY

OF COBALT

SULFATE

balt metal powder, rather than to the sulfate; but in simple terms of cobalt exposure, it is possible that current worker exposure is higher than that shown to cause laryngeal inflammation and squamous metaplasia in rodents. According to the National Institute for Occupational Safety and Health (1977), the symptoms most commonly reported after occupational exposure to cobalt in the cemented tungsten carbide industry include upper respiratory tract irritation and exertional dyspnea. Workers have been shown to develop diffuse interstitial pneumonitis and fibrosis. There are few comparable animal inhalation studies with cobalt in the literature. Kerfoot et al. ( 1975) exposed miniature swine to 0.1 or 1.O mg/m3 cobalt powder for 6 hr per day, 5 days per week for 3 months. They reported decreased lung compliance and microscopic evidence of interstitial fibrosis at both exposure concentrations. Johansson et al. (1984) reported alveolar Type II cell hyperplasia in rabbits after 4- to 6-week exposures to cobalt chloride (0.4-0.6 mg cobalt/ m3, 6 hr per day, 5 days per week). Apparently, no other part of the respiratory system was examined in these studies. In summary, exposure of rats and mice to aerosols of cobalt sulfate heptahydrate resulted primarily in necrotizing injury to the respiratory tract. The larynx appeared to be the most sensitive tissue, showing metaplastic and inflammatory lesions after exposure at concentrations as low as 0.3 mg/m3 cobalt sulfate heptahydrate (equivalent to 0.11 mg cobalt/m’). A no-observed-adverse&& concentration could not be determined from these studies. REFERENCES American Conference of Government Industrial Hygienists (ACGIH) (1988). Threshold Limit Values and Biological Exposure Indices for 1987-88. American Conference of Government Industrial Hygienists, Cincinnati, OH. BASSET, F., FERRANS, V. J., SOLER, P., TAKEMURA, T., FUKUDA, Y., AND CRYSTAL, R. G. (1986). Intralumi-

IN RATS AND MICE

371

nal fibrosis in interstitial lung disorder. Amer. J. Puthol. 122,443-46 I. BOORMAN, G. A., MONTGOMERY, C. A., JR., EUSTIS, S. L., WOLFE, M. J., MCCONNELL, E. E., AND HARDISTY, J. F. (1985). Quality assurance in pathology for rodent carcinogenicity studies. In Handbook of Curcinogen Testing(H. Milman and E. Weisburger, Eds.), pp. 345-357. Noyes Press, Park Ridge, NJ. BOORMAN, G. A., BROWN, R., GUPTA, B. N., URAIH, L. C., AND BUCHER, J. R. (1987). Pathologic changes following acute methyl isocyanate inhalation and recovery in B6C3Fl mice. To.xicol. Appl. Pharmacol. 87, 446-456.

BOYD, J. W. (1983). The mechanisms relating to increases in plasma enzymes and isoenzymes in disease of animals. Vet. Clin. Puthol. 12,9-24. DE BIE, E., AND DOYEN, P. (1962). Cobalt oxides and salts. Cobalt 15,3-13. DE MATTEIS, F., AND GIBB, A. H. (1977). Inhibition of haem synthesis caused by cobalt in rat liver. Biochem. J. 162,213-216. DOMINGO, J. L. (1989). Cobalt in the environment and its toxicological implication. Rev. Environ. Contam. Towicol. 108, 105-123. DOMINGO, J. L., LLOBET, J. M., AND BERNAT, R. (1984a). Nutritional and toxicological study of cobalt administered to rats in their drinking water. Rev. Toxicot 1,43-54. DOMINGO. J. L., LLOBET. J. M., AND BERNAT, R. (1984b). A study of the effects of cobalt administered orally to rats. Arch. Pharmacol. Toxicol. 10, 13-20. DUNN, 0. J. (1964). Multiple comparisons using rank sums. Technometrics6,241-252. DUNNETT, C. W. (1955). A multiple comparison procedure for comparing several treatments with a control. J. Amer. Stat. Assoc. 75,796-800. GOPINATH, C., PRENTICE, D. E., AND LEWIS, D. J. (1987). Atlas qf Experimental Toxicological Pathol0g.v.pp. 22-24. MTP Press, Lancaster, UK. HELMES. C. T., CASEY, S., FUNG, V. A., JOHNSON, 0. H., MCCALEB. K. E., MILLER, A., MILLER, J., PAPA, P., SIGMAN, C., AND STRAUSS, E. (1983). A study of metals for the selection of candidates for carcinogen bioassay. J. Environ. Sri. Health A 18(2), 203295.

ICHIKAWA, Y.. KUSAKA, Y., AND GOTO, S. (1985). Biological monitoring of cobalt exposure based on cobalt concentrations in blood and urine. Int. Arch. Occup. Environ. Health S&269-276. JOHANSSON, A., CURSTEDT. T., ROBERTSON, B., AND CAMNER, P. (1984). Lung morphology and phospholipids after experimental inhalation of soluble cadmium, copper, and cobalt. Environ. Res. 34,295-309. JOHNSTON, J. M. (1988). Cobalt Development Institute, personal communication to J. R. Bucher, NTP, May 5, 1988.

372

BUCHER

JONCKHEERE, A. (1954). A distribution-free k-sample test against ordered alternatives. Biometriku 41, 133145. KERFOOT, E. J., FREDRICK. W. G., AND DOMEIER, E. ( 1975). Cobalt metal inhalation studies on miniature swine. Amer. Ind. Hyg. Assoc. J. 36, 17-25. KLONNE, D. R.. GARMAN, R. H., SNELLINGS, W. M., DODD, D. E.. AND BALLANTYNE. B. (1987). The Larynx as a Potential Target Orgun in Aerosol Inhalation Studies on Rats. Presented at the March 23-27 meeting ofthe ILSI, The Design and Interpretation of Inhalation Studiesand Their Use in Risk Assessment, Hannover. FRG. LLOBET, J. M.. DOMINGO, J. L., AND CORBELLA, J. ( 1986). Comparison of the effectivenessof several chelators after single administration on the toxicity, excretion and distribution of cobalt. Arch. Toxicol. 589, 278-281. MAINES, M. D., AND KAPPAS, A. (1976). Studies on the mechanism of induction of haem oxygenase by cobalt and other metal ions. Biclchern. J. 154, 125- 13 1. MARONPOT, R. R., AND BOORMAN. G. A. (1982). Interpretation of rodent hepatocellular proliferative alterations and hepatocellular tumors in chemical safety assessment. Toxicol. Pathol. 10,7 l-80. MENZEL, D. B., WOLPERT, R. L., FRANCOVITCH, R. J.. SHOAF, C. R.. BOGER, J. R.. AND TAYYEB, M. I. ( 1989). Respiratory tract burdens of cobalt from inhalation of soluble aerosols: Simulation by a two-compartment model. Inhulation 7’o.xicol. 1,49-69. MILLER, M. E., HOWARD, D., STOHLMAN, F.. JR.. AND FLANAGAN, P. (1974). Mechanism of erythropioetic production by cobaltous chloride. Blood 44,339-346. MOLLENHAUR. H. H., CORRIER, D. E.. CLARK, D. E., HARE, M. F., AND ELISSALDE, M. H. ( 1985). Effects of cobalt on testicular structure. Virchows .4rch. 49,24 l248.

MORIN. Y., AND DANIEL, P. (1967). Quebec beer-drinkers’ cardiomyopathy: Etiological considerations. Cunad. Med. Assoc. J. 97,926-928. MORRISSEY, R. E., SCHWETZ. B. A., LAMB, J. C., IV. Ross, M. C.. TEAGUE, J. L., MORRIS, R. W. (1988).

ET AL. Evaluation of rodent sperm, vaginal cytology, and reproductive organ weight data from National Toxicology Program I3-week studies. Fundam. Appl. Toxicol. l&343-358.

National Institute for Occupational Safety and Health (NIOSH) ( 1977). Criteria .for a Recommended Standard Occupational Exposure to Tungsten and Cemented Tungsten Carbide, PB-275 594. National Technical Information Service, Springfield, VA. National Toxicology Program (NTP) (1990). NTP Technical Report on the Toxicity Studies of Cobalt Sulfate Heptahydrate in F344/N Rats and BK3Fl Mice, NTP Tox 5. U.S. Department of Health and Human Services. Public Health Service, National Institutes of Health. Research Triangle Park, NC. PALEY, K. R., SOBEL, E. S., AND YALLOW, R. S. (1958). Effect of oral and intravenous cobaltous chloride on thyroid function. J. Clin. Endoerinol. Metub. 18,850-859. RAABE, 0. G. (1980). Physical properties of aerosols affecting inhalation toxicology. In Pulmonary To.uicology of Respirable Particles (C. L. Sanders, F. Cross, G. E. Dagle, and J. A. Mahaffey, Eds.). U.S. Technical Information Center, U.S. Department of Energy. Oak Ridge, TN. SEDERHOLM, T., KOUVALAINEN, K., AND LAMBERG, B. A. (1968). Cobalt-induced hypothyroidism and polycythemia in lipoid nephrosis. Actu Med. &and. 184,30

l-306.

SHIRLEY. E. ( 1977). A non-parametric equivalent of Williams’ test for contrasting increased dose levels of a treatment. Biometrics 33,386-389. SMITH, I. C., AND CARSON, B. L.. Eds. ( 198 1). Cobalt. Trace Metals in the Environment, Vol. 6. Ann Arbor Science Pub., Ann Arbor, Ml. SMITH, R. P. (1980). Toxic responses of the blood In Cusarett and Doull’s Toxicology, (J. Doull, C. D. Klaassen. and M. Amdur, Eds.), 2nd ed., pp. 3 I l-33 I. Macmillan Co., New York. TAYLOR, A.. AND MARKS, V. (1978). Cobalt: A review. J. Human Nutr. 32. 165- 177. WILKS, S. S. (1932). Certain generalizations in the analysis of variance. Biometrika 24,47 l-494.