Teratogenic evaluation of diglycidyl ether of bisphenol A (DGEBPA) in New Zealand White rabbits following dermal exposure

FUNDAMENTAL

AND

APPLIED

TOXICOLOGY

(1988)

l&736-743

Teratogenic Evaluation of Diglycidyl Ether of Bisphenol A (DGEBPA) in New Zealand White Rabbits following Dermal Exposure’ W. J. BRESLIN,* Mammalian

H. D. KIRK, AND K. A. JOHNSON

and Environmental Toxico1og.v Research Laboratory, Health and Environmental Sciences, The Dow Chemical Company, Midland, Michigan 48674

Received September 21, 1987; accepted December 29, I987 Teratogenic Evaluation of Diglycidyl Ether of Bisphenol A (DGEBPA) in New Zealand White RabbitsFollowing Dermal Exposure. BRESLIN. W. J., KIRK, H. D., AND JOHNSON, K. A. (1988). Fundam. Appl. Toxicol. 10,736-743. Digtycidyl ether ofbisphenol A (DGEBPA) was tested for its potential to cause embryo/fetal toxicity and teratogenicity in pregnant rabbits. DGEBPA was applied daily to the clipped skin of New Zealand White rabbits for approximately 6 hr/day at dose levels of 0 (polyethylene glycol 400. vehicle control), 30, 100, or 300 mg/kg body weight/ day on Days 6 through 18 ofgestation. Fetuses were examined for external, visceral, and skeletal alterations on Day 28 of gestation. Maternal toxicity was observed among pregnant rabbits in the 300 mg/kg/day dose group as evidenced by moderate to severe erythema, fissures. hemorrhage, and slight edema at the exposure site. Similar, but less severe skin lesions were observed in pregnant rabbits in the 100 mg/kg/day exposure group. A slight erythema at the site ofapplication was observed in dams in the 30 mg/kg/day dose group. The erythema in rabbits from the low dose group was indistinguishable from the erythema caused by the occlusive bandages/ jackets used to hold the test material in place and, thus, was not considered toxicologically significant. No evidence of embryo/fetal toxicity or teratogenicity was observed at any dose level. Thus, the embryo/fetal no-observed-effect level for dermally applied DGEBPA was 300 mg/kg body we&t/day. the maximum tolerated dose. 0 1988 Society ofToxicology.

Diglycidyl ether of bisphenol A (DGEBPA) is an epoxy resin used in fiberglass moldings and adhesives. The material is low in acute oral (rat LD50 approx. 11 g/kg) and dermal (rabbit LD50 approx. 20 g/kg) toxicity, has a low eye irritation potential and moderate skin irritation potential in rabbits. DGEBPA is a skin sensitizer in guinea pigs (Hine et al., 1981). Skin contact through the manufacture and use of DGEBPA-based resin products presents the greatest potential exposure problem as most formulations are viscous, sticky liq’ Sponsored by The Dow Chemical Company. CIBAGEIGY Corp. and the Shell Oil Co. under the auspices of the Society of The Plastics Industry. 2 To whom correspondence should be addressed. 0272-0590/88 $3.00 Copyright 0 1988 by the Society ofToxicology. All rights of reprodunmn in any form reserved.

736

uids which are difficult to remove from the skin. The low vapor pressure and low acute toxicity of DGEBPA generally precludes inhalation as a serious hazard. Metabolism studies with 14C-DGEBPA revealed that dermally applied DGEBPA is slowly absorbed through the skin of mice (Climie et al., 198 1a). Recovery of 90% of the radioactivity (97% unchanged resin) from the site of application was obtained 24 hr after exposure. Although DGEBPA is slowly absorbed after dermal exposure, it appears to be rapidly metabolized and excreted (Climie et al., 198 1a.b). The ma.jority of the DGEBPAderived radiolabel was excreted via the feces. Due to the lack of data concerning the teratogenic potential of DGEBPA and the potential for dermal exposure during manufacture

DGEBPA

DERMAL

and use of this material, the study reported herein was conducted in rabbits via the dermal route of exposure. MATERIALS

AND

METHODS

Test material. Diglycidyl ether of bisphenol A was obtained from The Dow Chemical Company, Freeport, Texas. Analysis by differential scanning calorimetry indicated a purity of 99.1%. The test material was further analyzed via capillary gas chromatography/mass spectrometry for residual epichlorohydrin. Epichlorohydrin was not detected in the sample with a detection limit of 0.22 g/g. Test animals. Adult female New Zealand White rab bits3 were allowed to acclimate to the laboratory conditions for a minimum of 3 weeks prior to use. Rabbits were housed individually in cages with wire floors in rooms designed to maintain temperature at approximately 22°C relative humidity at approximately 50%, and a 12-hr light: 12-hr dark photocycle. Rabbits were maintained on 4 oz/day of feed4 prior to the start of the study. Their feed allocations were increased to 8 oz/day during the study to allow for increased nutritional demand during pregnancy. Municipal tap water was available ad libitum at all times. Female rabbits, approximately 3.5-4.5 kg body weight, were injected iv (marginal ear vein) with 50 USP units of chorionic gonadotropin to induce ovulation following artificial insemination (Gibson et al., 1966). The day of insemination was considered Day zero of gestation. Rabbits placed on study were identified by means of uniquely numbered metal ear tags and assigned to treatment groups using a stratified, by weight, randomization procedure. The randomization procedure was intended to increase the probability of uniform mean pretreatment body weights across groups. Experimental design. Groups of 26 inseminated rabbits were treated with 0 (vehicle control), 30, 100, or 300 mg/kg body weight/day of DGEBPA, applied dermaily on Days 6 through 18 of gestation. The dose levels were based on the results of a previous 13-day dermal probe study in which pregnant rabbits administered 300 or 500 mg DGEBPA/kg/day developed significant skin lesions as exhibited by a dose-related increase in erythema, fissuring, and/or crust at the site of application. DGEBPA was prepared as a solution in polyethylene glycol 400 (PEG 400). DGEBPA is a solid at room temperature which necessitates its dissolution in a solvent for dermal application. Control rabbits were treated with PEG 400.

3 Hazelton-Dutchland, Denver, PA. 4 Certified Laboratory Rabbit Chow, Ralston Purina Co.. St. Louis, MO.

TERATOLOGY

737

All rabbits were dosed with a volume of 1 ml/kg body weight/day. The hair from an area approximately 10 X 15 cm on the back of each animal was clipped using electric clippers on the day prior to insemination, and reclipped periodically as required during the course of the study. The test material was uniformly spread over the clipped area using a syringe and blunt-tipped needle. An occlusive bandage of absorbent gauze and nonabsorbent cotton was then placed over the dosing area on the back. The bandage was held in place for a minimum of 6 hr using a lycra/spandex jacket. Following the occlusion period, the bandage and jacket were removed. Residual DGEBPA was not washed from the skin following removal of the bandage and jacket. Preliminary studies conducted in our laboratory indicated that washing the exposure site following removal of the bandage and jacket resulted in additional skin injury. In addition, due to the low water solubility and sticky nature of the DGEBPA dose solutions, washing the exposure site was ineffective in removing residual DGEBPA. The location ofthe exposure site and difficulty in removing residual DGEBPA from the application site also inhibited ingestion of the test material; however, minimal oral exposure may have occurred. Maternal observations. All animals were observed daily throughout the experimental period for indications of toxicity from the test material. The site of application was assessed for erythema/eschar, edema, and exfoliation/fissuring (EPA, 1985; OECD, 1983). Skin lesion scores were recorded daily following the removal of the bandage/jacket. Animals found dead or showing signs of abortion were submitted for gross pathologic examination by a veterinary pathologist. Body weights were recorded on Days 0,6 through 19, and 28 ofgestation. Statistical analyses of body weight and body weight gain were performed using data recorded on Gestation Days 0, 6, 9, 12, 15, 19, and 28. In addition, maternal liver weights and gravid uterine weights were recorded at the time of cesarean section. Fetal observations. Test animals were killed by intravenous injection of T615 on day 28 of gestation. The uterine horns were exteriorized through a midline incision in the abdominal wall and the number of corpora lutea and the number and position of implantations, resorptions, and live or dead fetuses were recorded. The uteri of apparently nonpregnant animals were stained with a 10% solution of sodium sulfide (Kopf et al., 1964) and examined for evidence of implantation sites. This procedure was performed solely to determine the incidence of pregnancy and was not used to evaluate the incidence of resorptions. All fetuseswere weighed, sexed, and examined for external alterations. At least one-half of each litter, selected using a table ofrandom numbers, was examined by dissection under a low power microscope

’ T61 Euthanasia solution, American Hoechst Corp.

738

BRESLIN.

KIRK,

for evidence of soft tissue alterations (Staples. 1974). In the event a fetus not selected for examination of soft tissues exhibited an obvious external anomaly, suggestive of additional soft tissue malformation (e.g., omphalocele), that fetus was also selected for soft tissue examination in addition to those randomly selected. All fetuses were then preserved in alcohol. eviscerated. and subsequently cleared and stained with alizarin red-S (Dawson, 1926), and examined for skeletal alterations, Statistid evaluation. Descriptive statistics (mean and standard deviation) were calculated for skin lesions (erythema, edema, and exfoliation/fissuring). Maternal body weights and body weight gains, fetal body weights, gravid uterine weights, and absolute and relative liver weights were evaluated by Bartlett’s test for equality of variance (Winer, I97 1). Based upon the outcome ofBartlett’s test, a parametric (Steel and Torrie, 1960) or nonparametric (Hollander and Wolfe, 1973) analysis of variance (ANOVA) was performed. If the ANOVA was significant, analysis by Dunnett’s test (Winer, I97 1) or the Wilcoxon Rank-Sum test (Hollander and Wolfe, 1973) with Bonferroni’s correction (Miller, 1966) was performed. Statistical evaluations of the frequency of preimplantation loss. resorptions among litters and the fetal population, and fetal alterations were made by a censored Wilcoxon test (Haseman and Hoel. 1974) with Bonferroni’s correction. Pregnancy rate was analyzed by the Fisher exact probability test (Siegel. 1956). The number of corpora lutea and implantations, and litter size wasanalyzed by the Wilcoxon Rank-Sum test with Bonferroni’s correction. Fetal sex ratios were analyzed by a binomial distribution test (Steel and Torrie, 1960). The nominal CY levels used for statistical evaluation was 0.05. Because numerous measurements were statistically compared in the same group of animals the overall false positive rate (Type I errors) was much greater than the cited 01levels would suggest. Therefore. the final intetpretation of numerical data considered statistical analyses along with other factors such as dose-response relationships and whether the results were significant in the light of other biologic and pathologic findings.

AND

JOHNSON

deaths or abortions occurred in any treatment group during the conduct of this study. The maximum mean daily skin irritation/ damage scores for pregnant rabbits are presented in Table 1. Dermal application of DGEBPA produced a dose-related increase in erythema, exfoliation/fissuring, hemorrhage, and edema at the site of application. The severity of the skin lesions increased as dosing progressed. Most pregnant rabbits in the 300 mg/kg/day treatment group developed skin fissuring and moderate to severeerythema at the site of application. Eleven of these rabbits exhibited multifocal dermal hemorrhages in association with skin fissuring. A barely perceptible to well-defined edema was also observed in approximately half of the rabbits in this dose group. In general, rabbits in the 100 mg/kg/day treatment group developed lesssevereskin lesions;however, some rabbits in this dose level had skin lesions of similar severity to lesions observed in the 300 mg/kg/day dose group. Five pregnant rabbits in the 100 mg/kg/day treatment group exhibited skin fissuring and/or moderate to severe erythema at the exposure site. One of these rabbits also developed dermal hemorrhages. Edema, characterized asbarely perceptible, was observed in three rabbits at this dose level. In the 30 mg/kg treatment group, three pregnant rabbits exhibited a persistent well-defined erythema; however, three pregnant rabbits in the control group also developed skin erythema of similar severity. Erythema in the remaining rabbits from the 0 and 30 mg/kg dosegroups was barely percepRESULTS tible or was not detected. Skin fissuring, edema, and hemorrhage were not observed in Maternal Observations the control or 30 mg/kg treatment group. No other clinical signs of toxicity were observed One pregnant control rabbit died on Gesta- in the test animals. tion Day 2 1. The causeof death of this rabbit No treatment-related effects on body was attributed to a gastric hairball and subse- weights, body weight gains, or liver weights of quent anorexia and metabolic disturbance. pregnant rabbits were observed in this study In addition, two pregnant control rabbits (Table 2). A summary of the reproductive paaborted their litters prior to Gestation Day rameters and fetal observations made at the 28. Gross pathologic examination of these time of cesareansection are presented in Tarabbits indicated no visible lesions. No other ble 3. The pregnancy rate of rabbits in the 30

DGEBPA DERMAL

739

TERATOLOGY

TABLE I DGEBPA DERMAL TERATOLOGY: SKIN REACTION IN PREGNANT RABBITS DGEBPA (mg/kg/day)

Number of dams Edema” Erythema’ Exfoliation/fissuringd Hemorrhage

0

30

100

300

25 0.0 * O.Ob 0.4 f 0.6 0.0 f 0.0 0’

20 0.0 * 0.0

21 0.1 * 0.4

21 0.7 + 0.7

l.OkO.6

1.8-1-0.8 0.8 + 1.3

2.lt 0.6 2.6 t- 1.2

0.0 f 0.0 0

1

11

‘Skin edema: 0, none; 1, very slight (barely perceptible): 2, slight (we11defined); 3, moderate (raised approximately 1 mm); 4, severe (raised > 1 mm). b-C+ SD. ‘Skin erythema: 0, none; 1, very slight (barely perceptible); 2, slight (well defined); 3, moderate to severe; 4, severe erythema to eschar formation. d Exfoliation/fissuring: 0, none; 1, slight exfoliation; 2, moderate to severe exfoliation; 3, slight fissuring; 4, moderate to severe fissuring. ’ Total number of pregnant animals exhibiting hemorrhage.

mg/kg/day dose group was statistically decreased when compared to control rabbits. No other significant differences in reproduc-

tive parameters or fetal observations between treatment groups and the control were observed. However, the distribution of male to

TABLE 2 DGEBPA DERMAL TERATOLOGY: MATERNAL PARAMETERS IN

RABBITS"

DGEBPA (mg/kg/day) Gestation day(s)

0

100

30

300

Maternal body weight (g) 6

4131k278'

4126+268

4086+249

4069k238

Maternal body weight gain (g) 6-9 9-12 12-15 6-19 6-28

-46+75 13k65 29k83

Ilk116 (Total)

93-cl87

-34*53 19k80 19F114 -1Sk160 79*190

-43260 19k61

lo+82 15k124 67k200

-21264 3+83 llk72 13+183 1452208

Maternal liver weight Absolute (g) Relative (g/ 100 g body wt)

99226 2.33 kO.49

99+22 2.35 + 0.42

95k20 2.29 k0.44

107t18 2.56 k 0.35

a Bred New Zealand White rabbits were dermally exposed to DGEBPA on days 6 through I8 of gestation. Values reflect data from pregnant females only. b X+ SD.

740

BRESLIN,

KIRK, AND JOHNSON TABLE 3

DGEBPA DERMALTERATOLOGY:REPRODUCTIVEPARAMETERSINRABBITS DGEBPA (mg/kg/day)

No. deaths/No. females Pregnancies detected by stain % Pregnant No. of litters Corpora lutes/dam Implantations/dam Live fetuses/litter % Implantations resorbed Fetal body weight (g)d Fetal sex ratio, M:F

0

30

100

300

l/26 o/o 100

Of26

O/26

O/26

016

o/5

114

81 21 10*2

23*

101-2

7?3 7+-3 6(8/144) 36.0 k 5.5 45:55

7fi3 6t3 14(20/148) 36.9 t 5.6 49:51

23" 9k2' 7t3 6+3

14(23/164) 36.7 + 5.3 50:50

1-F 20

lo? 1 lk3 7*3 4(6/142) 35.7 f 5.3 40:60*

86

u Two females aborted their litters. ’ Two litters completely resorbed. ‘S+ SD. d Bof litter means t SD. * Statistically different from control or a binomial distribution (a = 0.05).

female fetuses within the 30 mg/kg/day group was statistically different from that of a binomial distribution. Fetal Observations External, visceral, and skeletal observations are presented in Table 4. No statistically significant increases in malformations or variations were observed in any treatment group when compared with the control group. A total of eight fetuses scattered throughout the dose levels exhibited malformations. One control group fetus exhibited omphalocele. In the 30 mg/kg/day dose group, one fetus exhibited multiple cardiac malformations consisting of cardiac inversion with associated malposition of cardiac vessels, persistent truncus arteriosus, and hypoplasia of the ventricles. This fetus also exhibited acephaly with an associated rudimentary cranial ossification, thoracic scoliosis, and a misshapen atlas. A second fetus in this group exhibited omphalocele. One fetus in the 100 mg/kg/day treatment group exhib-

ited multiple cardiac malformations consisting of cardiac hypertrophy and distension with associated hypertrophy of the tricuspid valve and ventricular septal defect. This same fetus also exhibited a fused and hemivertebra, a forked rib, and calloused ribs. A second fetus in this dose group exhibited persistent truncus arteriosus. Three fetuses in the 300 mg/kg/day dose group exhibited malformations; one fetus exhibited hydrocephaly, a second fetus exhibited persistent truncus arteriosus, and a third fetus exhibited a hemivertebra, fused ribs, and a forked rib. All malformations observed within a dose group occurred at low frequencies. Most of these malformations have also been observed at low frequencies in historical control New Zealand White rabbit data generated in this laboratory. DISCUSSION

AND

CONCLUSION

Dermal application of DGEBPA to pregnant rabbits produced maternal toxicity in the high- and mid-dose groups as exhibited

DGEBPA

DERMAL

741

TERATOLOGY

TABLE 4 DGEBPA DERMAL TERATOLOGY: FETAL ALTERATIONS AMONG RABBITS DGEBPA (mg/kg/day) 0

30

100

300

No. Fetuses (No. Litters) Examined External and skeletal examination Visceral examination

141 (22) 19 (22)

136 (20) 80 (20)

136 (21) 80 (21)

128 (21) 73 (21)

No. Fetuses (No. Litters) Examined Acephaly’ Omphalocele” Forelimb flexure Hydrocephaly” Multiple cardiac malformations“ Persistent truncus arteriosus’ Retrocaval ureter Skull Hyoid, delayed ossification Hyoid, crooked Vertebrae Atlas, delayed ossification, misshaped” Atlas, extra site of ossification Axis, dentoid process, delayed ossification Axis, extra site of ossification Spur Fused” Hemi” Scoliosis” Centrum Extra site of ossification Delayed ossification Stemabrae Fused Delayed ossification Ribs Cervical, extra Fused” Forked” Calloused” Asymmetry Delayed ossification Other Flexed digit

0 (0) l(l) 0 (0) 0 (0) 0 (0) 0 (0)

2 (2)

1 (lJb l(l) l(l) 0 (0) 1 (1)b.C 0 (0) 0 (0)

0 (0) 0 (0) 0 (0) 0 (0) 1 (I)“’ l(l) 0 (0)

0 (0) 0 (0) 0 (0) l(l) 0 (0) l(l) l(l)

71(18) 6 (4)

63(18)’ l(l)

71 (16) 4 (3)

55 (17)

0 (0) 0 (0)

cl (0) l(l) 3 (2)

cl (0) 20(13) 0 (0) 0 (0) 0 (0)

1 (lib 0 (0) 3 (2) 0 (0) 26 (14) 0 (0) 0 (0) 1 (lib

0 (0) 0 (0) l(l) 0 (0) 33(16) 0 (0) l(lY 0 (0)

0 (0) 0 (0)

l(l) 0 (0)

0 (0)

2(2)

3 (2) l(1)

l(l) 57 (17)

2 (2) 35 (16)

0 (0) 45 (16)

3 (2) 54(18)

0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

l(l) 0 (0) 1 (lld 1 UJd l(l) l(1)

0 (0) 1 cl)* l(lY 0 (0) 0 (0) 0 (0)

0 (0)

l(l)

0 (0)

0 (0)

2 (2)

2 (2) 23 (13) 1 (ljd 1 (lJd 0 (0)

2 (2)

a Considered a malformation. b,d.gMalformations identified with the same superscript were observed in the same fetus. ’ One fetus exhibited cardiac inversion with associated malposition of cardiac vessels,persistent truncus arteriosus, and hypoplasia of the ventricles. e One fetus exhibited cardiac hypertrophy and distension with associated hypertrophy of the tricuspid value and ventricular septal defect. /The hyoid from one fetus was removed and lost during processing.

742

BRESLIN,

KIRK, AND JOHNSON

by a dose-related increase in erythema, fissuring, hemorrhage, and edema at the site of application, Pregnant rabbits in the 300 mg/ kg/day treatment group exhibited moderate to severe erythema, fissures, hemorrhage, and slight edema. The degree of erythema, fissuring. and hemorrhage observed in high-dose rabbits represented a type of damage to the integument which dramatically altered the integrity of the skin. In general, the dermal lesions observed in rabbits in the 100 mg/kg/ day dose group were less severe than the lesions observed in rabbits in the high-dose group and did not develop to the extent which compromised the integrity of the skin. However. the skin response of individual rabbits in the 100 mg/kg/day dose group showed considerable variation, with some rabbits in this group developing lesions of similar severity to that observed in rabbits from the 300 mg/kg/day dose group. The average dermal response of rabbits in the 30 mg/kg/day dose group was characterized as barely perceptible erythema and was indistinguishable from erythema caused by the occlusive bandages/ jackets used to hold the test material in place. Since most pregnant control rabbits showed similar low levels of skin erythema. the skin response observed in rabbits from the 30 mg/ kg/day dose group was not considered toxicologically significant. No other treatment-related clinical signs or systemic toxicity was observed in this study. Although dermal exposure of pregnant rabbits to 300 mg DGEBPA/kg body weight/ day produced maternal toxicity, no indication of embryo/fetal toxicity or teratogenicity was observed at any dose level tested. The statistically significant decrease in pregnancy rate and deviation from a binomial distribution of the fetal sex ratio observed in the 30 mg/kg/day dose group were considered random events in view of the lack of a dose response for these parameters. A small number of fetuses scattered throughout the dose levels exhibited malformations. The fetal malformation rates (1.5 to 2.3%) in the DGEBPA dose groups were within the historical control

rate for New Zealand White rabbits used in our laboratory (unpublished data) and the rates reported by others (Palmer, 1968; Palmer, 1978; Stadler et ~1.. 1983). In addition, the frequency of specific malformations was low and did not show a dose response. Most of these malformations have also been observed at low frequencies in historical control New Zealand White rabbit data (Palmer, 1978; Stadler et al.. 1983) and thus were not considered due to exposure to DGEBPA. In conclusion. dermal exposure of pregnant rabbits to 0, 30, 100, or 300 mg DGEBPA/kg body weight/day on Gestation Days 6 through 18 produced no evidence of embryo/fetal toxicity or teratogenicity. Thus the embryo/fetal no-observable-effect level (NOEL) was 300 mg/kg body weight/day, the maximum tolerated dose. Exposure to 300 mg DGEBPA/kg body weight/day produced maternal toxicity as evidenced by moderate to severe erythema, fissures, hemorrhages. and slight edema at the site of exposure. Similar, but less severe, skin lesions were observed in pregnant rabbits in the 100 mg/kg/ day exposure group. Exposure to 30 mg DGEBPA/kg body weight/day or the vehicle produced slight erythema (barely perceptible) which was not considered toxicologically significant. REFERENCES CLIMIE. I. J. G.. HUTSON. D. H.. AND STOYDIN, G. (198la). Metabolism of the epoxy resin component 2.2.-bis[C(2,3-epoxypropoxy)phenyl]propane. the diglycidyl ether of bisphenol A (DGEBPA) in the mouse. 1. A comparison of the fate of a single dermal application and of a single oral dose of 14C-DGEBPA. .Yenohiotica II, 39 I-400. CLIMIE, I. J. G., HUTSON, D. H., AND STOYDIN. G. (198 1b). Metabolism of the epoxy resin component 2.2-bis[4-(2.3-epoxypropoxy)phenyl]propane. the diglycidyl ether of bisphenol A (DGEBPA) in the mouse. Il. Identification of metabolites in urine and fecesfollowing a single oral dose of i4C-DGEBPA. Xenobiotica 11,401-424. DAWSON, A. B. (1926). A note on the stainingofthe skeleton of cleared specimens with alizarin red-s. Stain Technol. 1, 123-124.

DGEBPA DERMAL Environmental Protection Agency (EPA) (1985). TSCA Test Guidelines: Final Rules. Federal Register, Sep tember 27,1985, Part II, Vo151, No. 9,39433-39434. GIBSON, J. P., STAPLES, R. E., AND NEWBERNE, J. W. ( 1966). Use of the rabbit in teratogenicity studies. Taxicol. Appl. Pharmacol. 9,398-408. HASEMAN, J. K., AND HOEL, D. G. (1974). Tables of Gehan’s generalized Wilcoxon test with fixed point censoring. J. Statist. Comput. Simulation 3, 117-135. HINE, C., ROWE, V. K., WHITE, E. R., DARMER, K. I., AND YOUNGBLOOD, G. T. (198 1). Epoxy compounds. In Patty’s Industrial Hygiene and Toxicology. (G. D. Clayton and F. E. Clayton, Eds.), Vol. 24. Wiley, New York. HOLLANDER, M., AND WOLFE, D. A. (1973). Nonparametric Statistical Methods. Wiley, New York. KOPF, R., LORENZ, D., AND SALEWSKI, E. ( 1964). Procedure for staining implantation sites of fresh rat uteri. Naunyn-Schmiedebergs Arch. Exp. Pathol. Pharmakol. 241, 121-135. MILLER, R. G., JR. (1966). Simultaneous Statistical Infirence. McGraw-Hill, New York.

TERATOLOGY

743

OECD (1983). OECD Guidelinesfor Testing Chemicals, Section I-Health Effects, No. 404: Acute Dermal Irritation/Corrosion. PALMER, A. K. (I 968). Spontaneous malformations of the New Zealand White rabbit: The background to safety evaluation tests. Lab. Anim. 2, 195-206. PALMER, A. K. (1978). Developmental abnormalities. Rabbits. In Pathology of Laboratory Animals. Vol. 2. (K. Benirschke, F. M. Garner, and T. C. Jones, Eds.), pp. 1848- 1860. Springer-Verlag, New York. SIEGEL, S. (1956). Non-Parametric Statistics for the Behavioral Sciences. McGraw-Hill, New York. STADLER, J., KESSEDJIAN, M.-J., AND PERRAND, J. (1983). Use of the New Zealand White rabbit in teratology: Incidence of spontaneous and drug-induced malformations. Food Chem. Toxicol. 21,631-636. STAPLES, R. E. (1974). Detection of visceral alterations in mammalian fetuses. Teratology 9, A-37. STEEL, R. G. D., AND TORRIE, J. H. (1960). Principles and Procedures of Statistics. McGraw-Hill, New York. WINER, B. J. (197 1). Statistical Principles in Experimental Design. 2nd ed. McGraw-Hill, New York, NY.