http://tpx.sagepub.com/Toxicologic Pathology

http://tpx.sagepub.com/content/26/5/587The online version of this article can be found at:

 DOI: 10.1177/019262339802600501

1998 26: 587Toxicol PatholAnthony B. DeAngelo, Michael H. George, Steve R. Kilburn, Tanya M. Moore and Douglas C. Wolf

F344/N Rats Mice and1Carcinogenicity of Potassium Bromate Administered in the Drinking Water to Male B6C3F

  

Published by:

http://www.sagepublications.com

On behalf of: 

  Society of Toxicologic Pathology

can be found at:Toxicologic PathologyAdditional services and information for    

  http://tpx.sagepub.com/cgi/alertsEmail Alerts:

 

http://tpx.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

http://tpx.sagepub.com/content/26/5/587.refs.htmlCitations:  

What is This? 

- Sep 1, 1998Version of Record >>

by guest on October 7, 2012tpx.sagepub.comDownloaded from

Safety Assessment

TOXICOLOGIC PATHOLOGY, vol. 26, no. 5, pp. 587-594, 1998 Copyright 0 1998 by the Society of Toxicologic Pathologists

Carcinogenicity of Potassium Bromate Administered in the Drinking Water to Male B6C3F, Mice and F344/N Rats*

ANTHONY B. DEANGELO, MICHAEL H. GEORGE, STEVE R. KILBURN, TANYA M. MOORE, AND DOUGLAS C. WOLF

Eiiviroririieiital Carcinogeriesis Division. National Health and Enviroririierital Effects Research Laboratory, US. Enviroitnierital Protection Agency, Research Triangle Park, North Carolina 2771 I

ABSTRACI

Ozone has been proposed for water disinfection because it is more efficient than chlorine for killing microbes and rcsults in much lower levels of carcinogenic trihalomcthanes than does chlorination. Ozone leads to formation of hypobromous hcid in surface waters with high bromine content and forms brominated organic by-products and bromate. The carcinogenicity and chronic toxicity of potassium bromate (KBrO,) was studied in male B6C3F, mice and F344M rats to confirm and extcnd the results of previous work. Mice were treated with 0. 0.08, 0.4, or 0.8 g/L KBrO, in the drinking water for up to 100 wk, and rats werc provided with 0, 0.02, 0.1, 0.2. or 0.4 g/L KBrO,. Animals were euthanatized, necropsied, and subjected to a complete macroscopic examination. Selected tissues and gross lesions were processed by routine methods for light microscopic examination. The present study showed that KBrO, is carcinogenic in the rat kidney, thyroid, and mesothelium and is a renal carcinogen in the male mouse. KBrO, was carcinogenic in rodents at water concentrations as low as 0.02 g/L (20 ppm; 1.5 mglkglday). These data can be used to estimate the human health risk that would be associated with changing from chlorination to ozonation for disinfection of drinking water.

Disinfection by-products; kidney; mesothelioma; mesothelium; nephropathy; renal cell tumor; thyroid; urothelium; water

Keywords.

INTRODUCTION Potassium bromate (KBrO,) is a chemical oxidizing

agent that has been used in analytical chemistry, in flour milling, as an ingredient in fish-paste products in Japan, in beer malting, in cheese making, and as a component of home and salon permanent hair-wave kits (2, 3). Its purpose in the baking industry is to improve the baking properties of flour (2, 3). Potassium bromate is not a nat- urally occurring compound and is synthesized by passing elemental bromine through a solution of potassium hy- droxide (3). In an industrial setting, KBrO, is considered a respiratory irritant, with a time-weighted average work- place exposure of 5 mg/m3 for a standard work week (2).

Most serious human exposures to KBrO, solutions have been through the accidental or purposeful ingestion of permanent hair-wave solutions (3, 11, 30,43). Human patients who ingest KBrO, solutions develop renal dis- ease and neuropathy (3, 11, 30, 43). Nephrotoxicity and ototoxicity occur in human patients within the first few hours after ingestion; frequently the outcome of poison- ing is death (1 1). Partly because of the severity of the acute toxicity in humans following exposure to high oral concentrations, KBrO, solutions have mostly been re-

* Address correspondence to: Dr. Douglas C. Wolf, US. EPA MD- 68, 86 l3V Alexander Drive, Research Triangle Park, North Carolina 277 11; e-mail: wolf.doug@epamail.epa.gov.

placed in permanent hair-wave kits with less acutely toxic solutions (30). Toxicity in human patients has resulted from ingestion of estimated doses as low as 5 mgkg body weight, and death has resulted from ingestion of solutions containing as little as 12 g KBrO, (11, 30).

Chronic KBrO, exposure has been studied in the ham- ster, mouse, and rat (28, 29, 31, 32, 44, 48, 49). Mice and rats fed brominated bread for 2 yr did not have an increased incidence of tumors or treatment-related non- neoplastic lesions (17, 19). Subsequent to these bromi- nated bread studies, it was discovered that the baking process changes KBrO, to KBr; therefore, there is little KBrO, in finished bread products (17, 29). Potassium bromate is stable in aqueous solution and was found to be a carcinogen in the rat after 2 yr of administration in the drinking wmer (28, 44). Male and female F344 rats developed renal cell tumors and thyroid follicular tumors, and the males also had an increased incidence of abdom- inal mesotheliomas (29, 32). Mice and hamsters treated with KBrO, in the drinking water for up to 2 yr did not have an increased tumor incidence but did have increased numbers of dysplastic renal lesions (32, 33, 49).

The Safe Drinking Water Act and its Amendments were enacted to assure that U.S. citizens are provided with clean and safe potable water. In this act, the U.S. Environmental Protection Agency (EPA) is charged with identifying drinking water contaminants and their poten-

587 0 192-6233/98$3.00 + $0.00

by guest on October 7, 2012tpx.sagepub.comDownloaded from

588 DEANGELO ET AL TOXICOLOGIC PATHOLOGY

TABLE I.-Dosing and animal survival in male B6C3F, mice treated with KBrO, in the drinking water.

Targeted concentrations fc/L)

0 0.08 0.4 0.8

0 0.085 2 0.006 0.426 f 0.068 0.836 t 0.056 , hleasured concentrations (EL)” Water consumption (mg/kg/day)h 115 2 3‘ . 114 f 5 106 t 4 96 2 4** hlean daily dose (mg/kg/day)’ 0 9.1 42.4 77.8 Feed consumption (g/kg/day)h 130 t 7 124 2 6 126 2 6 120 2 6 No. animals started on study 49 48 49 51 Unscheduled deaths 12 15 1 1 9

a hlean 2 SD.

‘Time-weighted over 100 wk. ** p 5 0.05 (significant trend with dose).

hlean (ZSEhI) daily consumption measured over the study.

tial adverse health effects and with developing regula- tions and suggested compliance technologies that mini- mize human exposure to potential toxicants and carcin- ogens (5). For almost 20 yr, the public, government, and scientific communities have been discussing alternatives to chlorination of drinking water (1, 9, 16, 35, 45). The rationale for replacing chlorine lay in the demonstration that some disinfection by-products are carcinogenic in laboratory animals (7, 13, 23). One of these alternatives is ozonation.

Ozone has been proposed for water disinfection partly because it is more efficient than chlorine for killing mi- crobes (1, 9, 16, 45). Ozone disinfection also results in much lower levels of trihalomethanes. which are carci- nogenic in rodents and are the major contaminants of chlorination (23, 25). However, ozone also leads to for- mation of hypobromous acid in surface waters with high bromine content. The acid reacts with natural organic ma- terial to form brominated organic by-products and bro- mate (10. 16, 18).

Ozonation of surface waters containing bromide ion (Br-) results in the oxidation of bromide to bromate (18). which can be found in finished drinking water as a by- product. The reaction occurs rapidly at environmental bromide concentrations and is dependent on the ozone, concentration, pH, and contact time (6). Krasner et a1 (25) measured quarterly concentrations of 70-100 pg/L bro- mide in the influent waters of 35 U.S. utilities. Based on that study and upon pilot plant experiments in which bro- mate production was measured in raw water spiked with various bromide concentrations, the waters were ozon- ated. It was calculated that 75% of the water supplies would have 40 pg/L bromate and that levels of up to 100

pg/L bromate would be expected in water containing 2 mg/L bromide (6). The present study was designed to confirm and extend the results of Kurokawa et a1 (28, 29, 32, 44). to examine the previously unreported effects in male mice, and to provide a detailed quantitative exam- ination of target tissues for development of a human health risk assessment by the U.S. EPA.

MATERIALS AND METHODS Preparation of KBrO, Dosirig Soliitioiis. Potassium

bromate (99+%; CAS 7758-01-2; Aldrich Chemical Co., Milwaukee, WI) was dissolved in deionized water to pro- duce target concentrations of 0.02,0.08,0.1,0.2,0.4, and 0.8 g/L. Freshly prepared solutions were administered to the animals in brown glass water bottles fitted with Teflon stoppers and stainless-steel, double-balled sipper tubes. Drinking water solutions were changed every 5-7 days. The purity, targeted KBrO, concentrations, and stability of representative solutions were monitored throughout the study. The solutions were pipetted into 7-ml liquid scin- tillation minivials that were then capped tightly and stored at 5°C. The drinking water was analyzed to deter- mine KBrO, concentration according to EPA Method 300.0 (14) using a Waters high-pressure liquid chroma- tography system equipped with a Dionex CD-20 conduc- tivity detector with a DS-3 conductivity cell and an ASRS-I1 anion self-regenerating suppressor module. The nominal and analyzed KBrO, concentrations are shown in Tables I and 11.

Anirnals arid Ariirnal Husbandry. Only male rats were used because they are the most sensitive species and sex for bromate-induced cancer. Male mice were included be- cause there is no published information on bromate car-

TABLE 11.-Dosing and animal survival of male F344 rats treated with KBrQ, in the drinking water.

Targeted concentration ( g L )

0 0.02 0. I 0.2 0.4

0.11 2 0.017 0.214 2 0.032 0.426 2 0.068 hleasured concentrations (EL)” 0 0.022 t 0.003 Water consumption (ml/kg/day)* 65 2 3 66 Z 3 72 2 3 79 t 3* 88 t 3**

Feed consumption (g/kg/day)h 44 t 2 45 t 2 45 2 2 46 2 2 46 2 2

Unscheduled deaths 19 I5 18 25 30

hlean daily dose (mg/kg/day)’ 0 1.5 7.9 16.9 37.5

No. animals started on study 54 54 54 53 53

“hlean 2 SD. hlean (2SEhl) daily consumption mencure4 o\er the study. Time-weighted o\er 100 \\k. \

* p 5 0.05 (significant \\hen compared with the control). ** p d 0.05 (significant trend with doce).

by guest on October 7, 2012tpx.sagepub.comDownloaded from

Vol. 26, No. 5, 1998 BROMATE CARCINOGENICITY IN RODENTS 589

cinogenicity in male mice. Weanling male B6C3F, mice and F344 rats confirmed free of viral antibodies and bac- terial and parasitic infections were obtained (Charles Riv- er Laboratories, Portage, MI) and held for 1 wk in quar- antine. The animals were randomly assigned to the treat- ment groups and started in the study at 28-30 days of age. The treatment rooms were maintained at 20-22°C and 40-60% humidity on a 12-hr light-dark cycle. Five or 6 mice were housed in each polycarbonate cage and were provided Purina Rodent Laboratory Chow (Purina, St. Louis, MO) and water ad libitimtti. Rats were housed 3/cage on wood chips. Clinical observations were re- corded daily; moribund animals were euthanatized and necropsied. Tissues taken from moribund and dead ani- mals were preserved in 10% neutral buffered formalin (NBF). Body weights and water consumption were mea- sured at the start of the studies, twice during the first month, and then monthly. Time-weighted water con- sumption was calculated over the interim and total dosing period by dividing the amount of water used over a par- ticular time interval by the animal weight/cage (expressed as ml/kg/day). All aspects of these studies were con- ducted in facilities certified by the Association for As- sessment and Accreditation of Laboratory Animal Care in compliance with the guidelines of that association and the National Health and Environmental Effects Research Laboratory Animal Care and Use Committee.

Atiittial Dosiiig. Two hundred 28-30-day-old male B6C3F, mice were randomly assigned to 0, 0.08,0.4, and 0.8 g/L KBrO, treatment groups (50 mice/group), and 250 28-30-day-old male F344 rats were randomly as- signed to 0, 0.02, 0.1, 0.2, and 0.4 g/L KBrO, treatment groups (50 rats/group). Animals received KBrO, in the drinking water for up to 100 wk. All animals that sur- vived to the end of the study were euthanatized and sub- jected to a complete necropsy, with macroscopic exami- nation of all tissues.

Patliology. Tissues to be taken at necropsy were se- lected based on previous studies and included gross le- sions, liver, kidneys, testes, spleen. thyroid gland (rat only), stomach, duodenum, jejunum, ileum, colon, rec- tum, and urinary bladder. The tissues were removed, ex- amined, and fixed in NBE Prior to euthanasia the animals were weighed, and at necropsy organ weights were ob- tained for the liver, kidneys, testes, spleen, and thyroid gland. Fixed tissues were processed by routine methods for paraffin embedment, sectioned at 5 Fm, stained with hematoxylin and eosin, and examined by light micros- copy. Nephropathy scores were graded semiquantitatively based on the percentage of the renal cortex involved: 0 = no nephropathy, 1 = 1-10% of renal cortex involved, 2 = -25% 3 = -50%, 4 = -7596, and 5 = >75% of the renal cortex involved and fibrosis and mineralization were present.

Seri4iti Clietiiistry. Serum samples from mice and rats were analyzed for levels of total protein, albumin, cre- atinine, uric acid, total plasma antioxidant activity (39). and blood urea nitrogen. The serum was also analyzed for .levels of lactate dehydrogenase, alanine transaminase, aspartate transferase, alkaline phosphatase, and sorbitol dehydrogenase. The analyses were done on a Cobas Fara

I1 centrifugal spectrophotometer (Hoffman/La Roche, Branchburg, NJ) using routine methods and standard kits (Sigma Chemical Co., St. Louis, MO).

Statistical Evabatioii. Continuous variables were an- alyzed using a 1-way analysis of variance (ANOVA) (53). Detection of some overall effect of treatment groups was followed by pairwise comparisons with controls us- ing appropriate contrasts. If either the homoscedasticity assumption (Levene’s test) or the assumption of normal distribution (Shapiro-Wilk test) were violated, then a nonparametric analysis (ANOVA on the ranks of the data) was performed, followed by nonparametric (Wil- coxon rank sum) pairwise comparisons with controls. Tests for trend with dose were performed using contrasts in the mean responses. For tumor (or lesion) prevalence, overall differences among treatments and for compari- sons with controls, respectively, the likelihood ratio x2 test (15), and Fisher’s exact test were used. Similar com- parisons involving the counts of tumors (or lesions) per organ were performed using log-rank tests (50). Trends (with dose) of tumor prevalence were evaluated using an extension of the Fisher-Irwin test. Trends (with dose) of tumor counts were evaluated using a log-rank monotone trend test. Tests of differences between tumors and/or le- sion prevalence or multiplicity were 1 sided. Survival curves were determined using the product limit estimates of Kaplan-Meier, and tests for equality of survival curves across strata were performed using the log-rank test. Sig- nificance was defined as p 5 0.05.

RESULTS Male B6C3Ft Mice

KBrO, Concentrations. Simrvival, Water arid Feed Con- suniptioii. Body Weight, arid Dose Deterinitiation. Mea- sured values of 0.085, 0.426, and 0.836 g/L were deter- mined for the 0.08, 0.4, and 0.8 g / L targeted KBrO, drinking water concentrations, respectively (Table I). No significant differences in animal survival, body weight gain, or feed consumption between any treatment group and the control group were observed. There was a trend (p 5 0.05) toward depressed water consumption with in- creasing KBrO, concentration, with the high-dose group significantly lower in water consumption than the control. Time-weighted mean (2SEM) daily water consumptions of 115 2 3, 114 2 5, 106 2 4, and 97 2 4 ml/kg/day and daily feed consumptions of 130 2 7, 124 2 6, 126 ? 6, and 120 ? 6 g/kg/day were calculated for the 0, 0.08,0.4, and 0.8 g/L KBrO, solutions, respectively, over the 100-wk expesure period. Time-weighted mean daily doses of 9.1, 42.4, and 77.8 mgkglday were calculated from the mean daily water consumptions and the mea- sured drinking water KBrO, concentrations (Table I).

Body atid Organ Weights. No significant alterations were seen for the final body weight or the absolute and relative liver, kidney, spleen, and testes weights in any of the KBrO, exposure groups when compared with con- trols.

KBrO,-Itidi(ced Neoplosia. There was a treatment- but not a dose-related increase in the incidence of mouse re- nal tumors after 100 wk of KBrO, in the drinking water.

by guest on October 7, 2012tpx.sagepub.comDownloaded from

590 DEANGELO ET AL TOXICOLOGIC PATHOLOGY

400

300

f ? + E 200

TABLE III.--Incidence' of kidney tumors in KBr0,-treated male B6C3F, mice.

-

-

-8 8

Dose Tumor (a) n Hyperplsia Adenomas Carcinomas incidence

0 40 5 (12.5) 0 0 0 0.08 38 6 (15.8) 2 (5.3) 3 (7.9) 5 (13.1)* 0.4 41 2 (4.9) 2 (4.9) 1 (2.4) 3 (7.3) 0.8 44 7 (15.9) 1 (2.3) 0 1 (2.3)

No. (lo) of affected mice. * p 5 0.05 when compared with conlrols.

The background incidence of renal tumors in male B6C3F, mice is < O S % (41); therefore, the increased in- cidence of renal cell tumors was determined to be bio- logically significant (Table 111). No significantly increased prevalence or multiplicity of hepatocellular adenoma or carcinoma was observed in any of the KBrO, treatment groups (Table IV). There were no other treatment-related increases in benign or malignant neoplasms. The spec- trum and incidence of nonrenal tumors were consistent with expected background for this sex and strain of mouse (41).

Noriiieoplastic Patliology. The kidney was examined for the presence of pathology. The severity of nephrop- athy was not different between treated and control ani- mals and had a nephropathy score of <1 (data not shown). There were no treatment-related increases in nonneoplastic lesions in any tissue examined. There were no significant differences in the incidences of nonneo- plastic lesions between control and treated mice (data not shown).

Senun Chemistry. No KBr0,-induced alterations in the serum enzyme activities, uric acid, total plasma antioxi- dant activity, protein, albumin, or creatinine concentra- tions were present (data not shown).

Male F344N Rats KBrO, Coiicentrations, Sirrvival, Water and Feed Con-

simption, Body Weight, and Dose Determiriation. Values of 0.022, 0.11, 0.214, and 0.426 g/L were measured for the targeted 0.02, 0.1, 0.2, and 0.4 g/L KEirO, drinking water concentrations, respectively (Table 11). There were significant decreases in the survival of animals in the 0.2 g/L and 0.4 g/L KBrO, groups (Table 11). The animals in the high-dose group were euthanatized and necropsied at 94 wk because of the high rate of mortality and morbidity that resulted from a significant depression of the body weight gain (Fig. 1). There was a significant trend toward increased water consumption with increasing KBrO, con- centration. Rats in the 0.2 g / L and 0.4 g/L treatment groups had 22% and 35%, respectively, increased water consumption over controls. Although KBrO, affected wa- ter consumption, there was no effect on feed consumption (Table 11). The final mean body weight of the animals in the high-dose group was significantly decreased (82% of control; Table V). Significant increases in the kidney and thyroid weights and increased relative liver, kidney, thy- roid, and spleen weights were present only in rats given 0.4 g/L KBrO, (Table V).

KBrO,-Iiiduced Neoplasia. A dose-dependent in- creased incidence of mesotheliomas arising on the tunica

TABLE IV.--Incidencea of hepatic adenomas and carcinomas in KBr0,-treated male B6C3F1 mice.

Dose Adenomas + (a) n Altered foci Adenomas Carcinomas carcinomas

0 41 8 119.5) 8 119.5) 13 131.7) 16 139) 0.08 38 13 (34.2j 1 1 i28.9j 16 i42.1j 22 i57:9) 0.4 42 10 (23.8) 8 (19.5) 1 1 (26.2) 19 (45.2) 0.8 44 16 (43.1) 10 (22.7) 16 (36.4) 23 (52.3)

a No. (56) of affected mice.

vaginalis testis was present in KBrO,-treated groups (Ta- ble VI). These vaginal tunic mesotheliomas tended to be bilateral but in some cases were unilateral. In many af- fected rats, mesothelioma was also present on the serosal surface of some or all of the abdominal viscera examined and the mesentery. In no rats was there evidence of meso- thelioma within the pleural cavity. In all rats with meso- theliomas on the abdominal viscera, there were also mes- otheliomas of the vaginal tunic. There were no cases of peritoneal mesothelioma without involvement of the vaginal tunic.

Male F344 rats treated for up to 100 weeks with 0.4 g/L KBrO, had an increase in renal cell tumors (Table VII). Most of the tumors occurred in the cortex, with a few extending into the outer stripe of the outer medulla. The most common cell type was a vacuolated cell with a thin rim or bands of basophilic cytoplasm. This vacu- olated cell type is similar to that reported previously for bromate-induced renal tumors in the rat and is similar to what was seen in the mouse.

Rats treated with 20.2 g / L KBrO, had an increased incidence of thyroid follicular proliferative lesions, in- cluding hyperplasia, adenoma, and carcinoma (Table VIII). There were no differences between treated and control rats in incidence of thyroid C-cell lesions. There were no other treatment-related increases in benign or

0

8 a

0 .

v v 8 e g 0

Water

+ 0.02 g

A 0.1 gll

v 0.2 gll

* 0.4 gll

l o o t 0' I I I I I I

I 1 I I

0 10 20 30 40 50 60 70 80 90 100

weeks of K B r 0 3 exposure

FIG. I.-Body weight gain of male F344 rats treated with KBrO, in their drinking water for up to 100 wk. Significant trend with dose, p 5 0.05; significant when compared with the control, *p 5 0.05.

by guest on October 7, 2012tpx.sagepub.comDownloaded from

Vol. 26, No. 5, 1998 BROMATE CARCINOGENICITY IN RODENTS 59 1

TABLE V.-Body and organ weight' for male F344 rats treated with KBrO, in the drinking water.

Dose (m@g/day)

0 1.5 7.9 16.9 31.5

No. of animals 35 39 36 29 25 Body weight (g) 432 & 4 422 t 15 426 2 0.8 410 t 13 354 2 8* Liver

Weight (g) 17.20 t 0.48 16.65 t 0.44 16.98 2 0.46 17.14 2 0.60 15.89 t 0.48 % Body weight 3.99 t 0.12 4.02 t 0.10 4.01 t 0.12 4.26 2 0.17 4.51 2 0.12*

Weight (g) 3.29 2 0.06 3.24 2 0.06 3.31 2 0.08 0.85 2 0.03 3.80 2 0.17* % Body weight 0.16 t 0.02 0.80 t 0.03 0.80 i: 0.02 1.84 2 0.03 1.09 t 0.05*

Weight (g) 0.031 t 0.001 0.078 t 0.036 0.034 2 0.003 0.031 2 0.002 0.054 2 0.007* % Body weight 0.007 2 0.001 0.021 t 0.011 0.008 2 0.001 0.008 2 0.001 0.015 2 0.002*

Weight (g) 2.29 t 0.56 2.84 2 0.42 2.01 2 0.26 % Body weight 0.53 t 0.13 0.16 2 0.14 0.48 2 0.07 0.99 t 0.28 0.75 2 0.14*

Weight (g) 6.17 t 0.33 5.77 t 0.31 6.09 2 0.31 4.92 2 0.36 6.24 2 0.55 % Body weight 1.43 t 0.08 1.39 ? 0.07 1.44 i: 0.08 1.21 2 0.09 1.76 2 0.15

Kidney

Thyroid

Spleen 3.66 2 0.95 2.70 t 0.55

Testes

* hlean 2 SEhl. * p 5 0.05 when compared with controls.

malignant neoplasms. The spectrum and incidence of oth- er tumors were consistent with expected background for this strain and sex of rat (42).

Notineoplastic Pathology. The kidney was examined for the presence of treatment-related increased seventy of chronic progressive nephropathy, but the severity of ne- phropathy was not different in treated and control ani- mals. Other alterations present in kidneys from bromate- treated rats included foci of mineralization of the renal papilla and eosinophilic droplets in the proximal tubule epithelium. There was no apparent dose-dependent in- crease in the incidence of either of these lesions.

The renal pelvis had a dose-dependent increase in hy- perplasia of the transitional cells lining the renal papilla and pelvis (urothelial hyperplasia, Table IX). Urothelial hyperplasia was characterized by a marked increase in the number of layers of urothelial cells lining the renal papillae and pelvis. These cells frequently extended into the renal space in long fronds or sessile mats. There were no other treatment related increases in nonneoplastic le- sions in any tissue examined. The incidences of nonrenal lesions were not significantly different between control and treated rats (data not shown).

Sennn Chetnistry. All serum chemistry values were within normal limits and were not different from those of control animals, except for an increased total plasma antioxidant activity in the 2 low-dose groups.

TABLE VI.-Incidence of mesotheliomas in KBrO,-treated male F344 rats.

D o s e (Sn, ~

n Incidence

0 47 0 0.02 49 0.1 49 0.2 47 10 (21.3)** 0.4 43 27 (62.8)**.***

4 (8.2)" 5 (10.2)*

* No. (%) of affected rats. * Not statistically significant (p = 0.06). * p 5 0.05 when compared with controls. ** p 5 0.002 when compared with conlrols. * * * p 5 0.002 (significant trend).

DISCUSSION The present study confirmed the observations by Ku-

rokawa et a1 (27-29, 32, 44) that KBrO,, when provided in the drinking water, results in a dose-dependent in- creased incidence of renal cell tumors, thyroid follicular tumors, and mesotheliomas arising on the tunica vaginalis testis in male F344 rats. Unlike previous reports, we also saw a treatment-related increased incidence of renal cell tumors in male B6C3Fl mice.

Rat renal cell tumors developed at concentrations of 20.1 g L (100 ppm) but were significantly increased only at the high dose. The morphologic appearance and dis- tribution of the renal tumors in the rat kidney were sim- ilar to what had been described previously (28). Renal cell tumors in male mice were seen at all treatment con- centrations of 20.08 g / L (80 ppm). Previous reports sug- gested that the mouse kidney may be a target for bro- mate-induced cancer, but this is the first study to confirm that KBrO, does induce male mouse kidney tumors.

Male and female B6C3F1 mice had been treated with KBrO, in the drinking water, but the males had to be removed from the study because of fighting (32). Female mice treated for 78 wk with 500 or 1,000 ppm of KBrO, in the drinking water did not develop renal tumors, and dysplastic foci were not mentioned (32). In a study of male B6C3F1, BDF,, and CDF, mice, an increased inci- dence of renal tumors was found after 88 wk of 750 ppm KBrO, in the drinking water (33). In another study, new- born ICR mice *treated subcutaneously with KBrO, de- veloped dysplastic renal foci but no tumors (38). In these previous studies, the mice may not have been treated long enough for renal tumors to develop or, as with other com- pounds, there may have been a gender difference in re- sponse in the mouse kidney (8, 12, 24, 37). In addition, there could be strain differences. Strain differences were detected for the production of lipid peroxides in male mice treated iv with KBrO, (33). Male CDF, mice had increased lipid peroxide levels and B6C3Fl and BDF, mice had decreased levels after KBrO, treatment (33).

Assessment of nephrotoxicity in the present study did

by guest on October 7, 2012tpx.sagepub.comDownloaded from

592 DEANGELO ET AL TOX~COLOGIC PATHOLOGY

TABLE VII.--Incidence" of renal cell tumors from KBrO,-treated male F344 rats.

Dose n Tumor incidence Total tumors Hyperplash Adenoma incidence Total adenomas Carcinoma incidence (a)

1 0 0.02 43 9 (29.9) 1 (2.3) 1 0 1 (2.3) 0.1 47 13 (27.7) 4 (8.5) 4 2 (4.3) 6 (12.7) 6

45 14 (31.1) 1 (2.2) 1 0 1 (2.2) 1

0.2 39 16 (41) 2 (5.1) 2 1 (2.6) 3 (7.7) 3 0.4 32 17 (53.1)* 9 (28.1)**.**** 10 4 (12.5)*.*** 12 (37.5)**3*** 14

0 No. (70). * p C 0.05 when compared with controls. ** p h 0.002 when compared with controls. * * * p h 0.05 (significant trend). ****p 5 0.002 (significant trend).

not reveal any association between treatment and nephro- sis in rats or mice. This response is different from that reported previously in which a dose-related increase in the severity of chronic progressive nephropathy in rats treated for up to 2 yr with KBrO, was associated with dysplastic foci and tumors (27, 29). We did find a treat- ment-related increase in the presence of eosinophilic droplets within the cytoplasm of proximal tubule epithe- lium. The eosinophilic droplets are thought to be indi- cators of oxidant damage and are possibly lipofuscin granules (33). When male F344 rats were treated with KBrO, and cysteine or reduced glutathione, the occur- rence of eosinophilic bodies was decreased; it was in- creased when rats were pretreated with diethyl maleate, which depletes glutathione (33). The eosinophilic bodies occur as a result of oxidative damage from the KBrO,.

The transitional cells within the renal pelvis were markedly hyperplastic in rats treated with 20.1 g/L (100 ppm) KBrO,, but there was no urothelial hyperplasia in the mouse renal pelvis. This alteration is frequently as- sociated with severe nephropathy and renal papillary ne- crosis (40, 54, 55). Neither of those lesions were preva- lent in this study. Up to 30% of KBr0,is excreted un- changed through the kidney, which could put a consid- erable amount of bromate in direct contact with the pelvis urothelium, causing a proliferative response (30). There was, however, no hyperplastic response in the urinary bladder. Treatment-related hyperplasia of the urinary bladder has not been associated with urothelial hyperpla- sia of the renal pelvis (40).

Mesothelioma arising from the tunica vaginalis testis was induced in the rat in a concentration-dependent man- ner from 20.02 g/L (20 ppm). Testicular mesotheliomas were found at a lower concentration than had been re-

TABLE VIII.-Incidencea of thyroid follicular proliferative lesions from KBr0,-treated male F344 rats.

Dose Adenoma + ( E L ) n Hyperplasia Adenoma Carcinoma carcinoma

0 36 0 0 0 0 0.02 39 I (2.6) 2 (5.1) 2 (5.1) 4 (10.3) 0.1 43 2 (4.7) 1 (2.3) 0 1 (2.3) 0.2 35 2 (5.7) 2 (5.7) 2 (5.7) 4 (11.4)* 0.4 30 2 16.7) 8 126.7)*.*** 6 120)*.*** 14 /46.6)**,***

a No. (9%) of affected rats. p S 0.05 when compared with controls.

** p S 0.002 when compared with controls. *** p S 0.002 (significant trend).

ported previously (27). In a dose-response study for car- cinogenicity of KBrO, in the rat, testicular mesotheliomas were induced at concentrations of 230 pprn in the drink- ing water but were not seen when rats were treated with 15 ppm for up to 104 wk (27). In the present study, mesotheliomas were frequently found scattered through- out the peritoneal cavity on the serosal surfaces of many organs. There did not appear to be a concentration-de- pendence to the frequency of multiple sites for this tumor. This particular mesothelioma is thought to arise on the tunica vaginalis covering the testicle and then spread to other sites by direct implantation or seeding from the primary tumor (20). These additional sites are considered neither additional primary sites nor metastases (20) (Pa- thology Working Group, personal communication).

Thyroid follicular tumors were increased in the rat in a treatment- and concentration-dependent manner. There was an increase in the incidence of thyroid follicular pro- liferative lesions even at the low dose of 0.02 g L (20 ppm), but thyroid follicular tumors were statistically in- creased in rats given 20.2 g/L (200 pprn). This concen- tration of KBrO, is much lower than that previously re- ported to induce thyroid follicular tumors. In a dose-re- sponse study of KBrO, provided in the drinking water to male F344 rats, thyroid follicular tumors were found only in rats treated with 260 ppm, but the increased incidence was significant only at the high dose of 500 ppm (27). Follicular cell proliferation usually follows a progression from hyperplasia to neoplasia (21, 51). The most com- mon causes of thyroid follicular neoplasia are iodine de- ficiency, excess iodine, chemicals that stimulate thyroid stimulating hormone secretion or negative feedback con- trol from the pituitary. Few chemicals induce thyroid tu- mors by a nonhormonal mechanism (21, 51). The mech-

TABLE IX.-Incidence" of renal pclvis urothelial hyperplasia from KBr0,-treated male F344 rats.

Dose (a) n Incidence

0 44 7 (15.9) 0.02 41 6 (13.9) 0.1 47 25 (53.2)** 0.2 39 32 (82.1)** 0.4 32 30 (93.8)**.***

a No. (%) of affected rats. ** p 5 0.002 when compared with controls. * * * p 5 0.002 (significant trend).

by guest on October 7, 2012tpx.sagepub.comDownloaded from

Vol. 26, No. 5, 1998 BROMATE CARCINOGENICITY IN RODENTS 593

anism by which bromate induces thyroid tumors is not known.

KBrO, was found to be mutagenic in the Ames test and to cause chromosomal aberrations in Chinese ham- ster fibroblasts (22). KBrO, had no initiating activity in the rat kidney when given as a single dose followed by a promoter, neither initiated nor promoted skin tumors in a skin paint study, and did not induce renal tumor de- velopment after 1 or 4 weekly subcutaneous injections in neonatal rats and mice (26, 34, 38).

Bromate is absorbed rapidly from the digestive tract and excreted in the urine as bromate (30-35%) and bro- mide (-40%) within 2 hr after administration (4). Bro- mate is thought to produce its toxic response through oxidative damage restulting from increased levels of lipid peroxide (33, 46-48). Bromate forms oxygen radicals, which are known to damage DNA, as evidenced by in- creased 8-hydroxy deoxyguanine levels in response to oxidative damage (36, 46, 52). Bromate must undergo cellular metabolism to cause DNA damage because DNA damage did not occur directly in an iii vitro mixture of DNA and bromate (30). The mechanism of action for bromate-induced carcinogenesis may include lipid per- oxidation, which generates oxygen radicals that induce DNA damage (30). The apparent species differences in the induction of renal cell tumors are correlated with the different levels of lipid peroxidation (36).

The results of the present study agree with those of previous reports that KBrO, is carcinogenic in the rat kidney, thyroid, and mesothelium; the present study also showed that KBrO, is a renal carcinogen in the male mouse. KBrO, was carcinogenic in rodents at concentra- tions in water as low as 0.02 g/L (20 ppm; 1.5 mgkg/ day). The development of tumors from chronic exposure to KBrO, is thought to be due to bromate metabolism with subsequent lipid peroxidation and generation of ox- ygen radicals that cause DNA damage (30). The apparent tissue and species specificity is correlated with the level of lipid peroxidation (33). These data can be used to crit- ically assess the relative risk associated with alternative water disinfection methods as compared with chlorina- tion. Although alternatives to chlorination may decrease the occurrence of some carcinogenic disinfection by- products, they may also increase the occurrence of others.

ACKNOWLEDGhIENTS We express our appreciation to the pathology working

group, Drs. Gary Boorman, Rick Hailey, Ron Herbert, Joel Leininger, Ann Radovsky, and Robert Sills, of the National Toxicology Program/National Institute for En- vironmental Health Sciences and Dr. Jeff Everitt of the Chemical Industry Institute of Toxicology. We are grate- ful to Mr. Lance Brooks for chemical analysis and to Mr. Donald Doerfler, Ms. Judith Schmid, and Dr. R. Wood- row Setzer for statistical analysis. We also thank Dr. Leon King for his assistance with the TnbleCiirve and SignznPlot software programs and Drs. David DeMarini and Kevin Morgan for critical review of the manuscript. Portions of this work were camed out under EPA Con- tract No. 68-D03-0024 to Pathology Associates, Freder- ick, MD. This manuscript has been reviewed and ap-

proved for publication by the Environmental Protection Agency and does not necessarily reflect the views of the agency. Mention of trade names or commercial products does not constitute endorsements or recommendations for use.

1.

2.

3.

4.

5.

6.

7.

8.

9.

REFER E N c E s

Akin EW, Hoff JC. and Lippy EC (1982). Waterborne outbreak control: Which disinfectant? Enltiron. Health Perspecf. 46: 7-12. Anonymous (198 1 ). Workplace environmental exposure level guide: Potassium bromate. Am. Ind. Hyg. Assoc. J. 42: A53-AS5. Anonymous (1986). Potassium bromate. /ARC Monogr. Eval. Car- cinog. Risk Cliem. H i m . 40: 207-220. Anonymous (1991). Final report on the safety assessment of sodium bromate and potassium bromate. J. Am. Coll. Toxicol. 13: 400- 414. Anonymous (1996). Safe Drinking Water Act Amendments of 1996. Public Law 10-1-182. Bull RJ and Kopfler FC (1991). Health Effects of Disitlfectarits and DisinJiectiorz By-products. American Water Works Association Re- search Foundation, Dcnver. Bull RJ, Sanchez IM, Nelson hlA, Larson JL, and Lansing AJ (1990). Liver tumor induction in B6C3F, mice by dichloroacetate and trichloroacetatc. Toxicology 63: 341-359. Carlton WW and Englehardt JA (1986). Chloroform nephrosis, male mouse. In: Urinaty Sjsrem TC Jones. U Mohr. and RD Hunt (eds). Springer-Verlag. Berlin, pp. 225-228. Carmichael NG, Winder C, Borgcs SH, Backhouse BL, and Lewis PD (1982). The health implications of water treatment with ozone. Life Sci. 30: 117-129.

10. Cavanagh JE, Weinberg HS, Gold A, Sangalah R. Marbury D, Glase WH, Collette T\V, Richardson SD, and Thruston AD (1992). Ozon- ation byproducts: Identification of bromohydrins from the ozona- tion of natural waters with enhanced bromide levels. Environ. Sci.

11. De Vriese A, Vanholder R, and Lameire N (1997). Severe acute renal failure due to bromate intoxication: Report of a case and discussion of management guidelines based on a review of the literature. Nephrol. Dial. Trarisplnrzr. 12: 201-209.

12. Dunn TB (1918). Sex differences in thc alkaline phosphatase dis- ribution in the kidney of the mouse. Am. J. Pathol. 24: 719-720.

13. Dunnick JK, Eustis SL, and Lilja HS (1987). Bromodichloro- methane, a trihalomethane that produces neoplasms in rodents. Cancer Res. 47: 5 189-5 193.

14. Environmental Protection Agency (1996). Reprints of €PA Mefhods for Cliernicnl Analyses tinder the Irlfonnation Collection Rule. Technical Report No. 8 14-B-96-006, Office of Groundwater and Drinking Water's Technical Support Division, Cincinnati, Ohio.

15. Fienberg S W (1980). The Analysis of Cross-Classified Data. MIT Press, Cambridge, Massachusetts.

16. Fiessinger E Rook JJ, and Duguet JP (1985). Alternative methods for chlorination. Sci. Total Enltiron. 47: 299-3 15.

17. Fisher N. Hutchinson JB, Berry R. Hardy J, Ginocchio AV. and \Vaitc V (1979). Long-term toxicity and carcinogenicity studies of

. the bread improver potassium bromate. 1. Studies in rats. Food Costtier. To.ric%l. 17: 33-39.

18. Glaze WH (1986). Reaction products of ozone: A review. Environ. Health Perspecf. 69: 151-157

19. Ginocchio AV. LVaite V. Hardy J. Fisher N. Hutchinson JB. and Berry R (1979). Long-term toxicity and carcinogenicity studies of the bread improver potassium brornate. 2. Studies in mice. Food Costnet. To.vicol. 17: 4 1-47.

20. Hall WC (1990). Peritoneum, retroperitoneum. mesentery, and ab- dominal cavity. In: Pathology of the Fischer Rar. GA Boorman, SL Eustis, hlR Eltvell, CA hlontgomery, and WF MacKenzie (eds). Academic Press, San Diego, pp. 63-69.

21. Hardisty JF and Boorman GA (1990). Thyroid gland. In: Pathology of the Fisclier Rar. GA Boorman, SL Eustis, h1R Elwell, CA Mont-

Teclitd. 26: 1658-1662.

by guest on October 7, 2012tpx.sagepub.comDownloaded from

594 DEANGELO ET AL TOXICOLOGIC PATHOLOGY

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

eomerv. and WF hlacKenzie (eds). Academic Press, San Diego, v - - pp. 519-534. Ishidate hl. Sofuni T. Yoshikawa K. Hayashi hi, Nohmi T. Sawada M. and Matsuoka A (1984). Primary mutagenicity screening of food additives currently used in Japan. Food Clietii. To.rico1. 22:

Jorgenson TA, hleierhenry EE Rushbrook CJ, Bull RJ, and Rob- inson M (1985). Carcinogenicity of chloroform in drinking water to male Osborne-hlendel rats and female B6C3F, mice. Firtidavi. Appl. Toxicol. 5 : 760-769. Koenig H, Goldstone A, Blume G. and Lu C (1980). Testosterone- mediated sexual dimorphism of mitochondria and lysosomes in mouse kidney proximal tubules. Science 209: 1023-1026 Krasner SW. McGuire LlJ, Jacagnelo JG, Patania NL. Reagan Khl. and Aieta EM (1989). The occurrence of disinfection by-products in U.S. drinking water. J. A m \Varer Works Assoc. 81: 41-43. Kunta Y. Diwan BA. and Ward JM (1992). Lack of renal tumour- initiating activity of a single dose of postassium bromate, a geno- toxic renal carcinogen in male F344MCr rats. Food Cliern. Toxicol.

Kurokawa Y, Aoki S , hlatsushima Y, Takamura N, Imazawa T. and Hayashi Y (1986). Dose-response studies on the carcinogenicity of potassium bromate in F344 rats after long-term oral administration. J. Nail. Cancer Inst. 77: 977-982. Kurokawa Y. Hayashi Y, Maekawa A. Takahashi hl, and Kokubo T (1982). Induction of renal cell tuniors in F-344 rats by oral ad- ministration of potassium bromate. a food additive. Garirr 73: 335- 338. Kurokawa Y, Hayashi Y, Maekawa A, Takahashi M. Kokubo T, and Odashima S (1983). Carcinogenicity of potassium bromate admin- istered orally to F344 rats. J. Natl. Cancer Inst. 71: 965-972. Kurokawa Y, Maekawa A, Takahashi M, and Hayashi Y (1990). Toxicity and carcinogenicity of potassium bromate-A new renal carcinogen. Envirori. Healrli Perspect. 87: 309-335. Kurokawa Y, Matsushima Y, Takamure N, Imazawa T, and Hayashi Y (1987). Relationship between the duration of treatment and the incidence of renal cell tumors in male F344 rats administered po- tassium bromate. Jpn. J. Cancer Res. 78: 358-364. Kurokawa Y, Takayama S , Konishi Y, Hiasa Y, Asahina S , Taka- hashi M, Maekawa A, and Hayashi Y (1986). Long-term in vivo carcinogenicity tests of potassium bromate, sodium hypochlorite. and sodium chlorite conducted in Japan. Erivirott. Healrh ferspect.

Kurokawa Y, Takamura N, Matsuoka C, Imazawa T, Matsushima Y, Onodera H, and Hayashi Y (1987). Comparative studies on lipid peroxidation in the kidney of rats. mice, and hamsters and on the effect of cysteine, glutathione, and diethyl maleate treatment on mortality and nephrotoxicity after administration of potassium bro- mate. J. Ani. CON. To.ricol. 6 489-501. Kurokawa Y, Takamura N, Matsushima Y, Imazawa T, and Hayashi Y (1984). Studies on the promoting and complete carcinogenic ac- tivites of some oxidizing chemicals in skin carcinogenesis. Cancer

Lawrence J and Cappelli FF' (1977). Ozone in drinking water treat- ment: A review. Sci. Total Emvirori. 7: 99-108. Lee YS, Choi JY, Park hlK, Choi EM, Kasai H, and Chung MH (1996). Induction of OHSGua glycosylase in rat kidneys by potas- sium bromate (KBrO,), a renal oxidative carcinogen. bfittat. Res.

Liebelt AG.( 1986). Unique features of anatomy, histology and ul-

623-636.

30: 251-259.

69: 221-235.

Lett. 24: 299-304.

364: 227-233.

trastructure, kidney, mouse. In: Urinary System, TC Jones, U Mohr, and RD Hunt (eds). Springer-Verlag. Berlin, pp. 24-44.

38. Matsushima Y. Takamura N. Imazawa T. Kurokawa Y. and Hayashi Y (1986). Lack of carcinogenicity of potassium bromate after sub- cutaneous injection of newborn mice and newborn rats. Sci. Rep. Res. Insf. Tohokir Univ. 33: 22-26.

39. hliller NJ, Rice-Evans C, Davies hlJ. Gopinathan V, and Milner A (1993). A novel mcthod for measuring antioxidant capacity and its application to monitoring the antioxidant status in premature neo- nates. Clin. Sci. 84: 407-412.

40. Montgomery CA and Seely JC (1990). Kidney. In: Potl1010gy of rAe Fischer Rat. GA Boorman, SL Eustis, hlR Elwell. CA Mont- gomery, and WF hlacKenzie (eds). Academic Press, San Diego,

41. National Institute for Environmental Health Sciences (1996). Tionor Incidence in Control Anirmls 6y Roicte mid Vehicle of Adininisrm- tiorr: B6C3F1 Mice. Analytical Science. Durham, North Carolina.

42. National Institute for Environmental Health Sciences (1996). Tianor Iiiciderice iri Control Animals by Roiite arid Vehicle of Adtniiiistra- tion: F343/N Rots. Analytical Science, Durham, North Carolina.

43. Niwa T, Ito T, and hlatsui E (1974). Serial renal biopsy in potassium bromate intoxication. Jpn. Circ. J. 38: 387-392

44. Ohno Y, Onodera H, Takamura N, Imazawa T, hlaekawa A, and Kurokawa Y (1982). Carcinogenicity of potassium bromate in F- 344 rats. Eisci Sliikenjo Hokoku 100: 93-100.

45. Robeck GC (1981). Chlorine, is that 3 better alternative? Sci. Total Enviroii. 18: 235-243.

46. Sai K. Takagi A, Umemura T, Hasegawa R, and Kurokawa Y (1991). Relation of 8-hydroxydeoxyguanosine formation in rat kid- ney to lipid peroxidation, glutathione level and relative organ weight after a single administration of potassium bromate. Jpn. J. Cancer Res. 82: 165-169.

47. Sai K, Uchiyama S. Ohno Y, Hasegawa R, and Kurokawa Y (1992). Generation of active oxygen species in vitro by the interaction of potassium bromate with rat kidney cell. Carciriogenesis 13: 333- 339.

48. Sai K, Umemura T, Takagi A. Hasegawa R, and Kurokawa Y (1992). The protective role of glutathione, cysteine and vitamin C against oxidative DNA damage induced in rat kidney by potassium bromate. Jpn. J. Cancer Res. 83: 45-51.

49. Takamura N, Kurokawa Y, Matsushima Y. Imazawa T. Onodera H. and Hayashi Y (1985). Long-term oral administration of potassium bromate in male Syrian golden hamsters. Sci. Rep. Res. Inst. To- 1iok1i Unit. Ser. C hfed. 32: 43-46.

50. Tarone RE and Ware J (1977). On distribution-free tests for equality of survival distributions. Biometrika 64: 156-169.

51. Thomas GA and Williams D (1994). Changes in structure and func- tion of the thyroid follicular cell. In: fatholobiology of the Aging Rat. Vol. 2, U Mohr, DL Dungworth, and CC Capen (eds). ILSI Press, Washington, D.C., pp. 277-280.

52. Umemura T, Sai K, Takagi A, Hassegawa R, and Kurokawa Y (1995). A possible role for oxidative stress in potassium bromate (KBrO,) carcinogenesis. Carcinogeriesis 16: 593-597.

53. Wner B (197 1). Statistical Principles in Expeririierital Design. Mc- Graw-Hill, New York, pp. 337-347.

54. Wolf DC and Carlton WW (1990). Acute cortical tubular necrosis in the Swiss ICR mouse induced by 2-bromoethylamine hydro- bromide. Food Clirn~, To.rico1. 28: 35 1-360.

55. Wolf DC, Carlton \VW, and Turek JJ (1992). Experimental renal papillary necrosis in the hlongolian gerbil. To.ricol. fothol. 20:

pp. 140-142.

150-172.

by guest on October 7, 2012tpx.sagepub.comDownloaded from