FUNDAMENTAL

AND

APPLIED

19, 176-185 (1992)

TOXICOLOGY

The Developmental Toxicity of 2-Ethylhexanol Applied Dermally to Pregnant Fischer 344 Rats’ R . W . TYL,*” L. C. FISHER,* M. F. KUBENA, * M. A. VRBANIC,* R. GINGELL,j’ D. GUEST,? J. R. HoDGsON,t S. R. MURPHY,-/’ T. R. TYLER,? AND B. D. ASTILL? *Bushy

Run Research

Center,

E.yx)rt, Chemical

Pennsylvania Marmfactwers

15632: and f’To.xicology Research Task Group, Association, Washington. D.C. 20037

Oxo-Process

Panel.

Received October 7, 199 I : accepted March 6. 1992

The Developmental Toxicity of 2-Ethylhexanol Applied Dermally to Pregnant Fischer 344 Rats. TYL, R. W., FISHER,L. C., KUBENA, M. F., VRBANIC, M. A., GINGELL, R., GUEST, D., HODGSON, J. R., MURPHY, S. R., TYLER, T. R., AND ASTILL, B. D. (1992). Fundum. A&. Toxicol. 19, 176-185.

Undiluted 2-ethylhexanol (2-EH) was administered by occluded dermal application for 6 hr per day on Gestation Days 6 through I5 to pregnant Fischer 344 rats, in range-finding (R) and main (M) studies.The dermal route is consideredto be the most relevant for human exposure. Treatment levels were (R) 0.0,0.5, 1.0, 2.0, and 3.0 ml/kg/day (equivalent to 0, 420, 840, 1680, and 2520 mg/kg/day) and (M) 0.0, 0.3, 1.0, and 3.0 ml/ kg/day (equivalentto 0,252,840, and2520 mg/kg/day). Controls (0.0 ml/kg/day, shamcontrols) received deionized water at 3.0 ml/kg/day. Dermal-positive control groups received undiluted 2-methoxyethanol(2-ME) at(R) 0.5 and 1.5 ml/kg/day and (M) 1.0 ml/kg/day as a reference compound in a similar regimen. An oral reference compound, valproic acid, was administered by gavagein the range-findingstudy on Gestation Days 6 through 15 at 400 mglkglday. The range-finding study employed an untreated (naive) control group. Numbersof plug-positive females per group were (R) 8 and (M) 25. Maternal weight gain was reducedfor 2-EH at 1680 (R) and 2520 (R and M studies)mg/ kg/day. Exfoliation andencrustationwereseenat the application site in both studies at 840, 1680, and 2520 mg/kg. Maternal liver, kidney, thymus, spleen,adrenal, and uterine weights, and gestational and fetal parameterswere unaffected by treatment with 2-EH. There were no treatment-related increasesin the incidenceof individual or pooled external, visceral, and skeletal malformations or variations following the application of 2-EH. The NOAELs for the maternal toxicity of 2-EH were 252 mg/ kg/day basedon skin irritation and 840 mglkgfday basedon systemic toxicity. The developmental toxicity NOAEL was at least 2520 mg/kg/day, with no teratogenicity. Administration of 2-ME at 840 mg/kg/day resulted in reducedmaternal weight ’ Presented at the 29th Annual Meeting of the Teratology Society. Fisher 989a). ’ Present address: Research Triangle Institute. Research Triangle Park, NC 27709. 3 Reprint requests should be addressed to: Betsy A. Recap. Chemical Manufacturers Association, 2501 M. St. N.W. Washington, D.C. 20037. et a/. (I

0272-0590/92 $5.00 Copyright CD 1992 by the Society of Toxicology. All rights of reproduction in any form reserved.

gain and food consumption, increasedpostimplantation loss, reducednumbersof live fetusesper litter, and reducedfetal body weights per litter. The incidence of fetal malformations and variations wasincreased.Oral administration of VPA produced maternal toxicity, developmental toxicity, and teratogenicity. The Fischer 344 rat is thus susceptibleto known rodent teratogensby both the dermal and oral routes. It is concluded that 2EH is not developmentally toxic by the dermal route in the Fischer 344 rat at and below treatment levels which produce maternal toxicity. 0 1992 Society 0f Toxicology.

2-Ethylhexanol (2-EH) is a widely used synthetic branched-chain primary alcohol with an annual world production in excess of 2 million tons a year. Its principal use is in the preparation of plasticizers such as bis(2-ethylhexyl) phthalate (DEHP) and bis(2-ethylhexyl) adipate. 2-EH has a low order of acute oral toxicity in rats and mice and low acute dermal toxicity in rabbits and guinea pigs (Rowe and McCollister, 1982). Studies in rats and rabbits, mostly at high doses, show that 2-EH is a skin, eye, lung, and GI tract irritant and a CNS and respiratory depressant. It also produces hepatomegaly, hepatic peroxisome proliferation, and mixed function oxidase induction at high doses in rodents (Hodgson, 1987). Oral gavage doses of up to 300 mg/kg are readily absorbed from the gastrointestinal tract of rats and are almost completely eliminated in the urine within 28 hr; the major metabolite is 2-ethylhexanoic acid (2-EHA) (Albro, 1975). Concern has arisen over the potential developmental toxicity of 2-EH since it is a major metabolite of DEHP in the rat (Albro et al.. 1973). DEHP is reported to be teratogenic in the mouse and fetally toxic in the rat (Tyl et al., 1988). In addition, its major metabolite 2-EHA is isomeric with 2propylpentanoic acid (valproic acid, VPA). VPA is a wellestablished animal and likely human teratogen (Lindhout and Meinardi, 1984; Vorhees, 1987). These concerns have led to studies on the developmental toxicity of 2-EH, principally by the oral route. Thus Hardin et al. (1987) tested 2EH in a proposed teratology screening procedure. Pregnant

176

TERATOGENICITY

OF DERMALLY

female CD- 1 mice (49 in number) received oral gavage doses of 1525 mg of 2-EH/kg/day on Gestation Days (gds) 6 through 13. There was a 30% mortality rate in the dams, and maternal body weights, numbers of viable litters, and litter sizes were reduced at delivery. Ritter et al. (1987) administered a single gavage dose of undiluted 2-EH to seven pregnant female Wistar rats on gd 12, in comparing the teratogenicities of 2-EH, 2-EHA, DEHP, and VPA. Dose levels were 8 12 and 1624 mg of 2-EH/kg. Decreases in body weight and increased incidences of hydronephrosis and tail, limb, and cardiovascular defects occurred in fetuses at 1624 mg/ kg. These effects were absent or minimal at 825 mg/kg. No data on maternal toxicity were reported. The effects of repeated oral doses of 2-EH on pregnant female Wistar rats have also been studied (US EPA, 1990). Gavage doses of 130,650, and 1300 mg of 2-EH/kg/day in an aqueous emulsion were given on gds 6 through 15 to groups of 10 animals per dose level. Marked maternal toxicity was seen at the highest dose together with increased numbers of fetal resorptions, marked post-implantation loss, and reduced fetal body weight. An increased incidence of skeletal malformations and variations and dilated renal pelvis and hydroureter were also seen at the top dose. There were only marginal indications of these effects at the intermediate dose level, and there were no adverse effects at the lowest dose. Price et ul. (199 1) reported the absence of developmental toxicity following the oral administration of 2-EH to timed pregnant female Swiss mice, in a comparative study with 2-EH, DEHP, and mono(2-ethylhexyl) phthalate. Doses of 17,60, and 190 mg of 2-EH per kg/day were given on gds 0 through 17 to 25-27 animals per test group by microencapsulation and incorporation in the diet. These studies suggest that there are developmental effects due to 2-EH by the oral route of exposure associated with relatively high intake levels and maternal toxicity. The only information on the potential for developmental effects due to 2-EH by a non-oral route is derived from an inhalation study by Nelson et nl. (1989). A group of 15 pregnant female Sprague-Dawley rats was exposed to 850 mg/m3 of 2-EH (estimated by the authors to be 24 mg/kg) daily for 7 hr throughout gestation. There were no effects on various gestational parameters, no fetotoxic effects, and no fetal malformations; adverse effects were limited to reduced feed intake by dams. A principal potential route of exposure to 2-EH is via the skin (Chemical Manufacturers Association, 1987). The data reviewed above are not adequate to estimate the developmental toxicity of 2-EH by the dermal route, and in consequence the study described in this report was undertaken. 2-EH was repeatedly applied dermally to pregnant female Fischer 344 (F344) rats during the period of major organogenesis. The F344 rat was selected as the experimental strain because of concurrent pharmacokinetic studies on 2-EH and recently reported developmental toxicity studies on 2-EHA

APPLIED

2-EH

177

TABLE 1 Study Design for 2-Ethylhexanol Dermal Developmental Toxicity GKWp” Range-finding study Naive control Sham control Test

Reference

Material

Distilled water 2-Ethylhexanol”

2-Methoxyethanol Valproic

Main study Sham control Test

Reference

applied

By volume (mVk/dw)

By weight @w/kg/day)

3 0.5 1 2 3 0.5 I.5

acid’

Distilled water 2-Ethylhexanol

2-Methoxyethanol

3 0.3 1 3 1

420 840 1680 2520 420 1260 400

252 840 2520 840

Number of animals

8 8 8 8 8 8 8 8 8 25 25 25 2s 25

a Plug-positive females treated on Gestation Days 6 through 15. * 2-Ethylhexanol and 2-methoxyethanol were undiluted and applied dermally. ’ Valproic acid was given as a 20% solution in corn oil by oral gavage.

by the oral gavage route (Fisher et al., 1989b). This dermal study included a range-finding test for dermal and systemic toxicities and a main study. Data from both studies are reported. Reference controls for the F344 rat were provided by a gavage treatment with VPA in the range-finding study and dermal applications of 2-ME in both range-finding and main studies. 2-ME is an established animal teratogen by the inhalation (Doe et al., 1983), gavage (Ritter et al., 1989) and dermal (Feuston et al., 1990) routes of exposure. MATERIALS

AND METHODS

Chemicals. 2-Ethylhexano14 (I-hexanol, 2-ethyl) had gas chromatographic purity of 99.72% as verified by the contract laboratory. 2Methoxyethanol’ and valproic acid’ had purities of 99.9 and 98.0%. respectively. Animals and dosages. Animals were virgin F344 rats [CDF (R) F344 Cr 1./Br] supplied by Charles River Breeding Laboratories (Kingston, NY). On arrival, males weighed 175-200 g and were 70 days old, and females weighed 130-150 g and were 63 days old. Animals were deemed suitable for use after fecal sampling, histological testing of selected organs, and serum viral antibody testing examination. After a 2-week quarantine period animals were mated I: I. Gestational Day 0 was dated from the appearance of a copulatory plug. Pregnant females were housed singly in stainless steel wiremesh cages and kept on a 12hr photoperiod at 68-73°F and at 42-65% humidity. Animals were supplied with feed6 and water ad lib. The study design is given in Table 1. Studies were performed according to current US EPA Health Effects Guidelines and Good Laboratory Practice Regulations. In the range-finding study four treatment levels of 2-EH. two

4 Provided by the Shell Developmental Company, Houston, TX. 5 Aldrich Chemical Company, Milwaukee, WI. 6 ProLab Certified Ground Rodent Chow, Agway Inc., St. Mary’s, OH.

178

TYL ET AL.

Maternal Distribution

TABLE 2 and Fate of Pregnant F344 Rats Following Dermal Administration or Gavage Dosing with Valproic Acid

of 2-Ethanol or 2-Methoxyethanol Main study

Range-finding study

2-EH

2-EH Treatment group: w/kg/day: Females in study”: Delivered Necropsied’: nonpregnant Pregnant With only nonviable implants With viable fetuses % Pregnant

Naive 0

Sham 0

8b

8

420 8

840 8

1680 8

2520

VPA

2-ME

2-ME

400 8

420 8

1260 8

8

0

0

0

0

0

0

0

0

4 4

2 6

2 6

2

0

0

0

1 I 0

3 5 4

4 50

6 15

6 15

16 15

1 I 0 I 87.5

62.5

6

0 15

25

2-ME 252

25

840 25

2520 25

840 25

0

0 5 20 8 20 80

0

2

2

0

2

2

4 19

5

6 6

3 20 2 18 88

2

6

23

0

I

Sham 0

6 75

0 15

0

0

20 0

19 84

23 92

20 92

y No females died, aborted, or were removed from the study. b Values are numbers of animals. ’ At scheduled necropsy. levels of the dermal reference compound 2-ME and a sham control ofdeionized water were applied dermally. A gavage reference compound (VPA) and a naive (untreated) control were also used. In the main study there were three treatment levels of 2-EH, one of 2-ME, and a deionized water control, all applied dermally. Eight plug-positive females were used per treatment level in the range-finding study, and 25 per level in the main study. Dermal test and reference compounds were applied undiluted; VPA was given at a concentration of 200 mg/ml in corn oil. Treatment volumes were based on the animal body weight on gd 6 and were not adjusted for subsequent weight changes. Treatment days were gds 6 through 15. For dermal treatment the appropriate volume of test or reference compound was dispensed from a l.O-cc syringe on to the clipped and shaved dorsal skin (ca. I .5 in.‘) between the scapulae, under a 2-in.’ gauze square. The application site was occluded with a Lycra-Spandex jacket with Velcro closures. A I .5 X 2.5-in. polyethylene patch was attached at the application site under the jacket. After a 6-hr exposure period the gauze and jacket were removed, and the application site was wiped gently with moist gauze and blotted dry. For oral gavage of VPA in corn oil a dose volume of 2.0 ml was administered with an 1I-gauge ball-end 3-in. needle attached to a 1.Occ syringe. In-life observations. Body weights were recorded on gds 0,6. 9, 12, 15. and 2 1. Food consumption was measured for each 3-day interval from gds 0 through 2 I. Observations were made at least once daily for clinical signs and skin irritation. Skin irritation was measured before and after each 6-hr treatment period. Irritation was scored according to the FHSA standard (1985). Maternal and fetal observations at necropsy. Females which delivered early were terminated, examined grossly, and removed from the study. Surviving females in both studies were euthanized by CO2 asphyxiation on Day 2 1. Maternal body cavities were opened by midline thoracolaporotomy. Maternal uterine and liver weights (both studies) and spleen, adrenals, kidneys, and thymus weights (main study) were recorded. Corpora lutea and uterine implantation sites were counted, and ovaries, cervices, vaginas, and abdominal and thoracic cavities were examined grossly.Uteri were examined externally, removed, and dissected longitudinally to expose the contents. All live and dead fetusesand resorption sites were noted; uteri from nongravid females were tested for early resorptions with ammonium sulfide solution (Salewski, 1964). Malformations and variations. All live fetuseswere sexed, weighed, and examined for external malformations and for variations. After external ex-

amination approximately 50% of the live fetuses per litter from the main study were examined for visceral (thoracic and abdominal: Staples, 1914, modified) and craniofacial abnormalities (Wilson, 1965, 1973, modified; Van Hulsingha and Bennett, 1911). The remainder were examined for skeletal malformations and variations after evisceration, fixation in ethanol, and staining with alizarin red S (Crary, 1962; Pletzer and Schardein, 1966). Analysis of data. The units of comparison were the pregnant rat or the litter. Quantitative continuous variables such as maternal body and organ weights were compared between 2-EH and sham control groups, between dermal reference and sham control groups, and between gavage reference and naive control groups. Levene’s test for equal variances (Levene, 1960) ANOVA, and t tests with Bonferroni probabilities for pairwise comparisons were used. The pooled t test was used when Levene’s test indicated homogeneous variances and ANOVA was significant. When Levene’s test indicated heterogeneous variances, all groups were compared by an ANOVA for unequal variances (Brown and Forsythe, 1914) followed when necessary by the separate variance t test. Nonparametric data following laparohysterectomy were evaluated using the Kruskal-Wahis test followed by the Mann-Whitney test when appropriate (Sokal and Rohif, 1969). Incidence data were compared using Fisher’s exact test (Sokal and Rohif, 1969). The fiducial limit of0.05 (two-tailed) was used as the criterion for significance.

RESULTS Test and Reference Substance Analyses The test substance 2-EH had purity of 99.7% at the outset and at the end of an 84-day period by gas chromatographic analysis. It is thus stable over the treatment period. The dosing solution of the gavage reference compound (valproic acid) was stable and homogeneous at the 200 mg/ml level over a 16-day period by gas chromatographic analysis. Maternal

Fate Data

The distribution and fate of all plug-positive females from both studies is given in Table 2. No females died, aborted, or were removed from either study in any control or treated

TERATOGENICITY

OF DERMALLY

APPLIED

179

2-EH

TABLE 3 Body Weight Changes of Pregnant F344 Rats in Grams Following Dermal Administration 2-Methoxyethanol or Gavage Dosing with Valproic Acid Range-finding

of 2-Ethylhexanol or

study

Main study

2-EH Treatment group: mg/kg/day: Pregnant females: Gestation days" O-6 (means) SD (+) 6-9 (means) SD (k) 6- 15 (means) SD (+) 15-2 I (means) SD (+) O-2 1 (means) SD (L)

Naive 0

Sham 0

420

1680

840

4

6

6

7

14.1 4.1

12.5

13.4 2.7

3.3 0.75 22.4 2.3 42.9 16.8 19.3

0.15 2.8 18.9 6.4 45.6

11.9 3 0.96 1.3 15.2 3.8 30.4 11.7 57.5 14.7

21.2

3.1

10 16.8 15.6

1.1 1.9 16.1 4.6 41.9 15.7 72 18.1

2-EH 2520

6

I

13.9 1.9 -2.2 1.5

14.7 1.4 -4.2b 6.6

lO.lb,d

10.7”J

7.1 32.5 17.4 56.5 24.2

4.8 34.5 11.4 56.5 24.2

VPA

2-ME

2-ME

Sham

400

420

1260

0 20

252 19

840 23

2520 20

840 20

13.8 2.9 3.3 1.21 19.2 8.4 21.9 16.3 60.9 25.1

15.1 2.1 2.3 2 18.5 4.5 32 15.5 65.6 19.7

14.4 5.1 1.7 2.3 18.7 5.6 38.9 13.4 71.9 17.3

15.9

16 2.3 0.99' 2.6 20 4.7

5

6

13.8 4.4 -3.5b 4.6 5.76 4.1 8.1b 4.9 28.1b 9

12.2 1.6 2.8 1.2 19.8 3 42 14.5 74 16.7

6

12.8 3.5 -2.2 2.6 15.5 3.6

- 1.Od 1.6 27.2d 9

2-ME

3 0.14c 2.4 14.9 4.9 36 12.1 66.8 15.8

16.4 18.3 52.3 20.9

y Means and standard deviations calculated from individual differences. b p < 0.0 1 compared with naive control. ‘p < 0.01 compared with main sham control. dp < 0.05 compared with range-finding sham control.

group. In the main study two females in the sham control and two in the low dose group delivered early, and their data are omitted. Pregnancy rates at laparotomy for sham controls and for 2-EH- and 2-ME-treated animals were 75-92%. The pregnancy rate for VPA was 62.5%. All females in 2-EHtreated groups had a 100% viability. Females treated with 2ME showed a treatment-related incidence of pregnant animals lacking viable implants, amounting to 40% of pregnant animals at 840 and 100% at 1260 mg/kg/day. There was a corresponding decrease in the number of 2-ME-treated animals with live fetuses at necropsy, amounting to 60% at 840 and 0% at 1260 mg/kg/day (Table 2). Only one of the five

females which were pregnant at necropsy, and had received VPA, had live fetuses. Maternal

In-Life Observations

Data in Tables 3 through 7 are based on females pregnant at necropsy. In summary, clinical findings for 2-EH-treated animals were limited to body weight changes, skin irritation, and nasal and ocular effects. Clinical findings with 2-ME were limited to body weight changes, food intake changes, and nasal and ocular effects, and with VPA to body weight and food intake changes.

Exfoliation

o-5

6-15

Encrustation

16-21

21

O-5

6-15

16-21

21

Gestation Days 0252

q 840

El2520

mglkglday

FIG. 1. Exfoliation and encrustation at the application site following dermal application of 2-ethylhexanol to pregnant F344 rats (main study). Incidence as percentage of treated animals. Sham control and 2-methoxyethanol-treated animals were unaffected.

TYL ET AL.

180

Range-finding

study

1.0 0.9 0.7 0.6 0.4 0.3 0.1 0.0 4

5 6a

b 78

b 6a

b 9a

b10abllabl2abl3abl4abl5ab

161718192021

Gestation days 0420

0640

42520

~1660

mglkglday

Main study

1.0 0.9 0.7 0.6 0.4 0.3 0.1 0.0 4

5 6a

b 7a

b 6a

b 9a

b10abllab12ab13ab14abl5ab

161716192021

Gestation days 0252

0640

+2520mglkglday

FIG. 2. Erythema at the application site following dermal application of 2-ethylhexanol to pregnant F344 rats. Values are mean Drake scores read before (a) and after (b) treatment. Sham control and 2-methoxyethanol-treated animals were unaffected.

Body weight changes. Maternal body weight changes for controls and treated animals are given in Table 3. Gestational weight changes (gds 0 through 2 1) for 2-EH-treated animals in both studies were not significantly different from sham controls. Reductions in weight gain with 2-EH treatment were seen in the range-finding study for gds 6 through 15 at 1680 and 2520 mg/kg/day. Weight gain was significantly reduced in the main study with 2-EH for gds 6 through 9 at 2520 mg/kg/day compared with the sham control: there was also a nonsignificant reduction at 840 mg/kg/day. Females treated with 1260 mg of 2-ME/kg/day in the range-finding study showed an overall reduction in weight gain from gds 0 through 2 1. This was seen as a marked reduction in weight immediately following the onset of treatment and a continued post-treatment reduction. In the main study, 2-ME at 840 mg/kg/day produced reduced weight gain on gds 6 through 9 compared with sham controls. 2-ME produced no effects on weight at 420 mg/kg/day. Maternal weight gain

was reduced in rats receiving VPA on gds 6 through 9 and 9 through I2 in comparison with the naive control. Weight gain then returned to normal in the post-treatment period. Food consumption. There were no significant changes in food consumption at any treatment level of 2-EH in either study throughout gestation (data not given). Animals treated with 2-ME at 840 and 1260 mg/kg/day showed a 12% reduction in food intake on gds 6 through 9 compared with the respective sham controls. Food consumption by animals receiving VPA was reduced by 32% on gds 6 through 9 and by 23% on gds 9 through 12 in comparison with naive controls and returned to normal in the post-treatment period. Clinical observations. Treatment-related effects attributable to 2-EH at the application site were exfoliation, encrustation, and erythema. There was no edema. Exfoliation and encrustation occurred in both range-finding and main studies at all treatment levels of 2-EH: Fig. 1 presents data from the main study. There was no erythema or edema in

TERATOGENICITY

OF DERMALLY

APPLIED

2-EH

181

TABLE 4 Mean Body and Gravid Uterine Weights of Pregnant F344 Rats at Termination Following Dermal Administration of 2-Methoxyethanol or Gavage Dosing with Valproic Acid Range-finding study

Main study Main study

Treatment group: mg/kg/day: Pregnant females: Body weight (g) SD (?) Gravid uterine weight (g) SD (+) Corrected body weightd (g) SD(*)

Naive 0 4

Sham 0 6

VPA 400 5

2-ME 420 6

2-ME 1260 6

Sham 0 20

2-ME 840 20

240 25 51.6 21.4 188 5.4

238 15.7 53.6 15.9 185 7.6

190” 9.2 4.7” 4.2 186 5.9

236 16.4 49.4 21.4 186 8.8

186’ 13.9 3.14b 1.05 183 12.9

220 28.6 39.4 23.8 186 10.1

213 21.3 23.4’ 20.5 190 8.5

ap < 0.0 1 compared with naive control. bp < 0.01 compared with range-finding control. ’ Forty-one percent depression compared with main control: not statistically significant, d Body weight minus gravine uterine weight.

sham controls. Erythema occurred during treatment with 2EH at levels of 840 mg/kg/day and above. Draize scores are presented in Fig. 2; irritation was essentially mild. Maximum mean treatment scores occurred on gd 10 at 1680 mg/kg/ day (0.4, range-finding study), on gd 11 at 2520 mg/kg/day ( 1.1, range-finding study), and on gd 14 at 1680 mg/kg/day (0.3, main study). There was no exacerbation by continued treatment. Erythema subsided immediately after the cessation of treatment. There were no exfoliation, encrustation, erythema, or edema at the application site with 2-ME. Nasal encrustation and ocular encrustation and discharge were seen mostly in the main study with 2-EH and 2-ME and in sham controls. Since these effects occurred in controls and mostly disappeared after treatment ceased they are attributed to handling stress. Necropsy Findings The only treatment-related findings for 2-EH-dosed animals at necropsy were residual exfoliation and crusting at the application site at mid and high treatment levels. There were no differences from controls for any treatment level of 2-EH, in maternal body weights, gravid uterine or corrected body weights, or in relative or absolute liver, thymus, spleen, adrenal and kidney weights. Treatment-related decreases in body and gravid uterine weights were found at necropsy for 2-ME and VPA-treated animals (Table 4). Thus at 1260 mg of 2-ME/kg/day, mean body and gravid uterine weights were significantly less than those of sham controls and were reduced nonsignificantly at the next lower treatment level. Similarly, mean body and gravid uterine weights were reduced for VPA-dosed animals in comparison with naive controls. Since corrected body

weights for VPA are similar to controls, the gross body weight reductions are considered to reflect post-implantation loss. Necropsy findings for 2-ME-treated animals were not otherwise different from controls. There were hepatic and spleen color changes in animals treated with VPA and a reduction of about 19% in absolute and relative liver weights compared with naive controls. Gestational

Parameters

Gestational data are presented in Table 5. In both studies 2-EH was without adverse effect at any treatment level, compared with controls, on total and nonviable i*mplants, early or late resorptions, live or dead fetuses, fetal sex ratio, and mean fetal body weights per litter. In contrast, 2-ME treatment caused a treatment-related increase in the numbers of nonviable implants. These were significantly increased at 840 mg/kg/day (main study) and at 1260 mg/kg/day (range-finding study) all implants were nonviable. There was also an increase in the numbers of early and late resorptions and dead fetuses at 840 mg of 2-ME/kg/day; thus the percentage of live fetuses at necropsy at this dose level was reduced by over 50% compared with the sham control. In addition, the mean body weight per litter of fetuses at this dose level was reduced by 38% compared with controls, although the number of fetuses was also reduced. Only one dam in the VPAtreated group had viable fetuses; these were two in number with body weights 40% less than controls. All other fetuses in VPA-treated animals were fully resorbed. Fetal Ma&wmations

and Variations

The incidence of statistically significant individual and total external, visceral, and skeletal malformations in fetuses

182

TYL ET AL.

TABLE 5 Gestational Parameters for Pregnant F344 Rats, Following Dermal Administration or Gavage Dosing with Valproic Acid Range-finding

of 2-Ethylhexanol or 2-Methoxyethanol

study

Main

2-EH Treatment group: mg/kg/day: Pregnant females: Corpora lutea SD (+) Total implants’ SD (?) Viable implants” SD (+) Nonviable implants” SD (2) Early resorptions” SD (+) Late resorptions” SD (?) Dead fetuses” SD (&) Percentage of live fetuses SD (?) Sex ratio (% males) SD (+) Fetal body weightg SD (?)

Naive 0 4 11.8 1.26 9.5 3.8 9 4.24 0.5 0.56 0.5 0.58 0 0 91.3 11.8 63.9 18.43 4.31 0.26

Sham 0 6 12 1.26 9 3 9 3 0 0 0 0 100 58.2 II.95 4.39 0.13

420

study

2-EH 1680

6

840 7

2520

11.5 1.76 1.7 3.6 7.5 4.4 0.2 0.41 0.2 0.41 0

1 I.4 2.64 5.4 3.1 5.3 2.9 0.1 0.38 0.1 0.38 0

10.7 3.01 6.3 5.2 5.1 5.2 0.7 0.52 0.7 0.62 0

12

0

0

0

0

98.6 3.4 61. I 17.44 4.56 0.33

98.7 3.44 51.7 20.9 4.77 0.54

80.4 23.9 73.1 30.3 4.78 0.62

97.1 7.65 42.1 24.1 4.54 0.21

6

I

1.29 6.3 4.2 6 4 0.3 0.16 0.3 0.76 0

VPA 400 5

2-ME 420 6

2-ME 1260

11.2 2.28 8 4.5 0.4' 0.9 1.6’ 4.7 7.2c 4.66 0.4' 0.89 0

11.2 1.11 8.5 4 8.2 3.8 0.3 0.41 0.2 0.41 0

11.3 2.07 8.1 1.63 Od

4.4' 9.94 100' 2.51

0.2 0.41 96.9 4.76 33.4" 20.14 4.1 0.34

6

8.7d 1.63 8.1“ 1.63 0

Sham 0 20

252 19

840 23

2520 20

11.6" 1.33 5.9 4.25 5.5 4.24 0.4 0.82 0.4 0.82 0

10.4 2.45 6.7 4.32 6.5 4.36 0.2 0.31 0.2 0.31 0

11.3 1.94 8.3 4.2 8 3.96 0. I 0.42 0. I 0.34 0

10.8 2 7.4 3.3 7.3 3.2 0.1 0.31 0.1 0.31 0

0

0.1 0.29 97.8 4.36 43.1 20.4 4.4 0.32

0

0 Od

86 31.1 62.8 29.3 4.59 0.33

96.8 8.8 41.8 29.4 4.5 1 0.51

99 3.1 53.4 20 4.5 0.38

2-ME 840 20 10.9 1.37 8 3.6 3.4 3.9 4.6d 4.41 2.9' 4.28 0.6' 1.31 I.ld 1.68 44.2d 45 43.5 25.3 3.3d 0.45

’ Means and standard deviations calculated from numbers per animal. bN= 18. ‘p < 0.05 compared with naive control. ‘p < 0.0 1 compared with appropriate sham control. ’ p < 0.05 compared with appropriate sham control. ‘Based on two fetuses from one surviving litter. g Based on numbers of fetuses per litter in grams.

for the main study, compared with controls, is given in Table 6. There were no external, visceral, or skeletal malformations associated with any treatment level of 2-EH. In contrast, 840 mg of 2-ME/kg/day produced significantly increased incidences of two external malformations, i.e. whole body edema and protruding tongue. In addition, the same dose produced significantly increased incidences of two visceral malformations, i.e. missing innominate arteries and dilation of the lateral ventricles of the brain. In the range-finding study, 2EH and 2-ME fetuses were examined for external malformations only; none were found. External and visceral malformations in the surviving fetuses from dams treated with VPA included tail and urogenital defects and imperforate anuses. The incidence of variations by individual fetus or by category is given in Table 7. There were no treatment-related increases in external, visceral, or skeletal variations at any treatment level of 2-EH. There were increases in the numbers of poorly ossified centrum 5 and in bilobed centrum 11 at 840 mg of 2-EH/kg/day only, not attributable to treatment with 2-EH. At 840 mg of 2-ME/kg/day there were increased incidences of ecchymosis of the head (an external variation), fetal atelectasis (a visceral variation), and 54 skeletal variations, relative to the sham control.

DISCUSSION Aspects of the design and interpretation of dermal developmental toxicity studies were discussed in a recent workshop (Kimmel and Francis, 1990). The studies in this report include many of the features recommended by the workshop, such as the use of sham controls, the avoidance of nondermal routes of exposure by the use of occlusive devices, minimization of maternal stress from restraining devices, daily inspection of application sites, and the use of a 6-hr exposure period. The workshop stressed the need to demonstrate maternal toxicity for a study to be usable for risk assessment purposes; in this study systemic maternal toxicity was produced by dermal application of 2-EH, as shown by reductions in weight gain on gds 6 through 9 at 1680 (range-finding) and 2520 (both studies) mg/kg/day. The workshop also recommended that skin irritation at the application site should be minimized. While 2-EH produced encrustation and exfoliation at the application site at all treatment levels, these effects are typical of the effects of alcohols on the skin, and the underlying irritation was quite mild. There was no edema at any treatment level, and erythema scores were mild to moderate at treatment levels of 840 and 2520 mg of 2-EH/ kg/day. The workshop recommended that skin absorption

TERATOGENICITY

Malformations

OF DERMALLY

APPLIED

2-EH

183

TABLE 6 in Fetuses from Pregnant F344 Rats Following Dermal Administration or 2-Methoxyethanol” Fetuses affected

of 2-Ethylhexanol Litters with affected fetuses

2-EH Treatment level: ~~/k/day:

Sham 0

252

840

2-EH 2520

2-ME 840

Sham 0

252

68 5 7.4 32 47.1

18 0 0 0 0

840 - -

2520

2-ME 840

19 0 0 0 0

23 0 0 0 0

20 0 0 0 0

12 4’ 33.3’ 10d 83d

-

Number examinedb Externally Edematous fetus % Protruding tongue % Viscerally Lateral ventricles dilated-tissue % Innominate artery missing

110 0 0 0 0 depressed

90 Skeletally Significant findings %

124 0 0 0 0

185 0 0 0 0

147 0 0 0 0

63 0 0 0 0

33 1 1.5 0 0

97 0 0 0 0

79 2 2.5 0 0

37 5 13.5 8 21.6

18 0 0 0 0

19 1 5.3 0 0

23 0 0 0 0

20 2 10 0 0

12 4’ 33.3’ 4’ 33.3’

53 0 0

58 0 0

88 0 0

68 0 0

31 2 6.5

17 0 0

15 0 0

21 0 0

18 0 0

_ 1’ 2 18.2

Number with malformations External % Soft tissue % Skeletal %

0 0 0 0 0 0

0 0 0 0 0 0

0 0 0 0 0 0

0 0 0 0 0 0

35 51.5 22 59.5 2 6.5

0 0 0 0 0 0

0 0 2 10.5 0 0

0 0 0 0 0 0

0 0 3 15 0 0

Total with malformations %

0 0

2’ 1.6

1 0

3’ 2.0

43 63.2

0 0

2 10.5

0 0

3 15

10d 83.3d 10d 83.3d 2 18.2 12d 1OOd

* Main study only. ’ Number of fetuses or litters. cp < 0.05 compared with sham control. d p < 0.01 compared with sham control. e Includes one fetus with ocular defects. ‘Includes one fetus with a divided olfactory lobe.

and limited pharmacokinetic data be collected in dermal developmental toxicity studies; published data on the skin absorption of 2-EH are not currently available. The F344 rat has been used relatively infrequently for developmental toxicity studies and less so for dermal developmental studies. In consequence, valproic acid was used as a reference compound for performance of the strain against a known developmental toxicant, and 2-methoxyethanol was used as a dermal reference compound. The strain sensitivity was established by the maternal toxicity, post-implantation loss, and reduced fetal body weights produced by daily gavage doses of 400 mg of valproic acid/kg/day. In addition, external and visceral fetal malformations were produced by VPA. These effects are similar to those seen in Sprague-Dawley rats after oral gavage doses of 300-400 mg of VPA on gds

7- 18 (Vorhees, 1987). The sensitivity of the F344 rat to the dermal developmental toxicant 2-ME was shown by the reduced weight gain, increased post-implantation loss, reduced fetal weights, and increased fetal malformations and variations produced by dermal application of 840 and 1260 mg of 2-ME/kg. These results are similar to those of Feuston et al. ( 1990), following dermal exposure of pregnant SpragueDawley rats to single doses of 250, 500, 1000, and 2000 mg of 2-ME/kg on Day 12 and to 2000 mg/kg/day on gds 10 through 14. These studies show that 2-EH is essentially without developmental or teratogenic effects in the presence of demonstrable maternal toxicity, when administered by the dermal route to pregnant F344 rats on gds 6 through 15. Maternal toxicity was limited to reduced body weight gain

184

TYL ET AL. TABLE

7

Total Variations in Fetuses from Pregnant F344 Rats Following Dermal Administration of 2-Ethylhexanol or 2-Methoxyethanol” Fetuses affectedb

Litters with affected fetusesb 2-EH

2-EH Treatment level: mg/kg/day:

Sham 0

252

840

External variations Number examined Number with variations %

110 13 11.8

124 19 15.3

185 37 20

Soft tissue variations Number examined Number with variations %

63 30 47.6

66 36 54.5

Skeletal variations Number examined Number with variations %

53 53 100

58 58 100

97 57 58.8 88 88 100

2520

2-ME 840

Sham 0

252

840

147 21 14.3

68 20 29.4

18 9 50

19 11 57.9

18 13 72.7

19 16 84.2

79 42 53.2 68 68 100

37 37 100 31 31 100

17 17 100

15 15 100

2520

2-ME 840

23 7 73.9

20 12 60

12 9 75

23 21 91.3

20 19 95

12 12 100

18 18 100

11 11 100

21 21 100

a Main study only. b No statistically significant differences from controls. ‘Number of fetuses or litters.

at 1680 and 2520 mg/kg/day and to mild or moderate skin irritation at 840 and 2520 mg/kg/day. Both effects occurred during treatment and were transient or ameliorated after treatment ceased. There were no other clinical findings and no effects on selected organs at necropsy. There were no effects on gestational parameters and no significant differences from controls in the incidence of fetal malformations and variations. The observations in these studies contrast with the findings, discussed above, in the Wistar rat following oral gavage doses of ca. 1300 to 1600 mg of 2-EH/kg (Ritter et al., 1987; US EPA, 1990). The differences are unlikely to be due strain differences, as the F344 rat in these studies and the SpragueDawley rat in the study by Vorhees (1987) responded similarly to almost identical oral gavage doses of VPA, as described above. It is probable that the absence of 2-EH developmental toxicity by the dermal route in the F344 rat is due to the low rate of absorption of 2-EH through the rat skin. In the absence of published data, this may be inferred from its low maternal toxicity after dermal application, when contrasted with the effects of 21-days of oral gavage dosing of female F344 rats (Hodgson, 1987). In this study, daily doses of 950 mg of 2-EH/kg produced a 40% loss in body weight gain and a 40% increase in liver weight compared with controls. This dermal developmental toxicity study with 2-EH was carried out with attention to current concerns about the experimental design of such a study. Part of the study was devoted to establishing that the F344 rat was a valid strain

for the study of developmental toxicity and that it was sensitive to known dermal teratogens. It was concluded that the F344 rat was a valid test strain for the purposes of this study. There are no developmentally toxic or teratogenic effects attributable to 2-EH applied dermally to pregnant F344 rats, at treatment levels up to and including 2520 mg/kg/day. The NOAEL for maternal systemic toxicity is 840 mg/kg/ day, based on reduced body weight gain, and the NOAEL for skin irritation is 252 mg/kg/day. The effects of dermal administration of 2-EH to pregnant F344 rats differ from those of repeated oral gavage administration to pregnant Wistar rats, where maternal toxicity, fetotoxicity, and malformations are produced at 1300 mg/kg/day. The authors are of the opinion that this study demonstrates that human skin exposure to 2-ethylhexanol is unlikely to result in developmental toxicity. ACKNOWLEDGMENT The authors thank Ms. K. A. Rosica, Chemical Manufacturers Association, for advice and help during her service as Program Manager of the Toxicology Research Task Group of the Oxo-Process Panel. REFERENCES Albro, P. W.. Thomas, R., and Fishbein, L. (1973). Metabolism of diethylhexylphthalate by rats. Isolation and chamcterisation of the major metabolites. J. Chromatogr. 76, 32 l-330. Albro, P. W. (1975). The metabolism of 2-ethylhexanol. Xenohiotica 5,625936.

TERATOGENICITY

OF DERMALLY

Brown, M. B., and Forsythe, A. B. (1974). The small sample behaviour of some statistics which test the equality of several means. Technomefrics 16, 129-132. Chemical Manufacturers Association (I 987). Summary of 2-ethylhewnol Uses. Poter~tial Expusure and S&y Procedures. Washington, DC. Crary, D. D. (I 962). Modified benzyl alcohol clearing of alizarin-stained specimens without loss of flexibility. Stain Technol. 37, 124- 125. Doe, J. E.. Samuels, D. M.. Tinston. D. J., and Wickramaratne, G. A. de S. ( 1983). Comparative aspects of the reproductive toxicology by inhalation in rats of ethylene glycol monomethylether and propylene glycol monomethylether. Toxicol. Appl. Pharmacol. 69. 43-47. Feuston, M. H., Kerstetter, S. L., and Wilson, P. D. (1990). Teratogenicity of 2-methoxy-ethanol applied as a single dermal dose. Fundam. Appl. Tosicol.

15, 448-456.

FHSA (Federal Hazardous Substances Act) ( 1985). Method of testing primary irritant substances. US CFR 16, 1500: 41. Fisher, L. C.. Tyl. R. W., and Kubena, M. F. (1989a). Cutaneous developmental toxicity study of 2-ethylhexanol in Fischer 344 rats. Teratology 39(5), 452. Fisher, L. C., Tyl. R. W., Kubena. M. F.. and Vrbanic, M. A. (1989b). Developmental toxicity of 2-ethylhexanoic acid by gavage in Fischer 344 rats and New Zealand White rabbits. Tuxiologist 9( 1). 269. Hardin. B. D.. Schuler. F. L., Burg, J. R., Booth, G. M., Hazelden, K. P., Piccirillo. V. J., and Smith, K. N. (1987). Evaluation of sixty chemicals in a preliminary developmental toxicity screening test. Terutog. Curcinog. Mutagen.

I, 29-48.

Hodgson, J. R. (1987). Results of peroxisome induction studies on tris(2ethylhexyl-mellitate) and 2-ethylhexanol. Toxicol. Ind. Health 3, 49-60. Kimmel. C. A., and Francis, E. Z. (1990). Proceedings of the workshop on the acceptability and interpretation of dermal developmental toxicity studies. Flmdam. Appl. To.yicol. 14, 386-398. Levene, H. (1960). Robust tests for equality of variance. In Cuntributions tu Pr~~bab~/~t~ and Statistics (I. Olkin et al., Eds.), pp. 278-292. Stanford Univ. Press, Stanford, CA. Lindhout, D.. and Meinardi. H. (1984). Spina bifida and in zrtero exposure to valproate. Lancet 2, 396. Nelson, B. K., Brightwell. W. S., Khan. A., Krieg, E. F., Jr., and Hoberman, A. M. (1989). Developmental toxicology evaluation of I-pentanol. I-hexanol and 2-ethyl-1-hexanol Administered by Inhalation to Rats. J Am. Coil. To.uiwl.

8, 405-409.

APPLIED

2-EH

185

Peltzer, M. A., and Schardein, J. L. (1966). A convenient method for processing fetuses for skeletal staining. Stain Technol. 41, 300-302. Price, C. J., Tyl, R. W.. Marr, M. C., Myers, L. B., Morrissey, R. E., Hendel, J. J., and Schwetz, B. A. (1991). Developmental toxicity evaluation of DEHP metabolites in Swiss mice. Teratology 43(5), 457. Ritter, E. J., Scott. W. J., Jr., Randall, J. L., and Ritter, J. M. (1989). Teratogenicity of dimethoxyethylphthalate and its metabolites methoxyethanol and methoxyacetic acin in the rat. Teratology 39, 363-373. Ritter. E. J., Scott, W. J., Jr.. Randall, J. L., and Ritter. J. M. (1987). Teratogenicity of di(2-ethylhexylphthalate), 2-ethylhexanol, 2-ethylhexanoic acid and valproic acid. and potentiation by caffeine. Terutology 35, 4 I46. Rowe. V. K.. and McCollister, S. B. (1982). In Patty’s Industrial Hygiene and Toxicology, (G. D. Clayton and F. E. Clayton, Eds.) third ed., Vol. 2C, p. 4620. Wiley-Interscience. New York. Salewski, E. (1964). Farbemethode zum makroskopischen Nachweis von lmplantionstellen am Uterus der Ratte. Nuunyn-Schmeidebergs Arch. Exp. Pathol.

Pharmacol.

247, 367.

Sokal, R. R., and Rohlf, F. J. (1969). Biometry, pp. 368-37 I, 399-400,470472, and 589-595. Freeman, San Francisco. Staples, R. E. ( 1974). Detection of visceral alterations in mammalian fetuses. Teratology 9, A37. Tyl. R. W., Price. C. J., Marr, M. C., and Kimmel, C. A. (1988). Developmental toxicity of dietary di(2-ethylhexyl)phthalate in Fischer 344 Rats and CD1 Mice. Fundam. .4ppl. To.uicol. 10, 395-412. US EPA (I 990). Office of toxic substances. Chem. Prog. Bzd1. 12(I), 20. Van Hulsingha, E. B., and Bennett, C. G. (1977). A dissecting procedure for the detection of anomalies in the rabbit fetal head. In Meth0d.y in Prenatal Toxicology (D. Neubert, H. J. Merker, and T. E. Kwasigroch. Eds.), pp. I26- 144. PSG Publishing, Littleton. MA. Vorhees, C. R. (1987). Teratogenicity and developmental toxicity of vaiproic acid in rats. Teratology 35, 195-202. Wilson. J. G. (1965). Embryological considerations in teratology. In Terutology-Principles und Techniques (J. G. Wilson and J. Warkany, Eds.). pp. 25 l-277. Univ. Chicago Press, Chicago. Wilson, J. G. (1973). Atlas of free-hand (razor blade) sections of the normal 20-day rat fetus. In Environment and Birth Defects (D. H. K. Lee, E. W. Hewson. and D. Okun. Eds.), Academic Press, New York.