TERATOLOGY 43:583-590 (1991)
Lack of Teratogenicity of trans-2-ene-Valproic Acid Compared to Valproic Acid in Rats CHARLES V. VORHEES, KAREN D. ACUFF-SMITH, WALTER P. WEISENBURGER, DANIEL R. MINCK, JAMES S. BERRY, KENNETH D.R. SETCHELL, AND HEINZ NAU Institute for Developmental Research, Children's Hospital Research Foundation and University of Cincinnati, Cincinnati, Ohio 45229-2899 (C.V.V.,K.D.A.-S., W.P.W., D.R.M.);482 Wellesley Drive, Cincinnati, Ohio 45224 (J.S.B.);Mass Spectrometry Laboratory, Children's Hospital Research Foundation, Cincinnati, Ohio 45229 (K.D.R.S.);Institut fur Toxikologie und Embryopharmakologie, Freie Uniuersitat Berlin, Berlin 33, Germany (H.N.)
ABSTRACT The teratogenicity of trans-2-ene-valproic acid (300 and 400 mg/kg) was compared with that of valproic acid (VPA; 300 mg/kg) and controls (corn oil) administered by gavage to Sprague-Dawley CD rats on embryonic (E) days 7-18. At the 300 mg/kg dose, trans-2-ene-VPA produced no change in maternal weight, number of implantations, proportion of resorptions, proportion of malformations, or fetal weight. By contrast, the same dose of VPA (300 mg/kg) reduced maternal weight during gestation, increased malformations (12.0% vs. 0.7% in controls), and reduced fetal body weight by 25.1%. An even higher dose of trans-2-ene-VPA (400 mg/kg) produced a reduction in maternal body weight during treatment and reduced fetal body weight (by 7.9%)' but did not increase resorptions or malformations in the fetuses. On day E18, maternal serum drug concentrations of VPA were higher in the VPA-treated group compared with those of trans-2-ene-VPA in the trans-2-ene-VPA-treated groups at 1 hr posttreatment. At 6 hr posttreatment the reverse was seen. trans-2-eneVPA may be absorbed more rapidly and distributed differently than VPA. Overall, the data support the view that trans-2-ene-VPA at equal or higher doses than VPA is not teratogenic in rats. The antiepileptic valproic acid (VPA) is associated with a prevalence of spina bifida of -1.3% in infants exposed in utero (Bjerkedel et al., '82; Jager-Roman et al., '86; Lindhout and Meinardi, '84; MastroiaCOVO et al., '83; Robert and Guibaud, '82). In addition, a pattern of minor dysmorphological and dysfunctional effects has been associated with intrauterine VPA exposure and has been termed the fetal valproate syndrome (FVS; DiLiberti et al., '84). Recently, the FVS has been corroborated and the effects associated with it expanded (Ardinger et al., '88). VPA is teratogenic in rats (Ong et al., '83; Vorhees, '87a) and mice (Whittle, '75; Brown et al., '80; Kao et al., '81; Nau et al., '81b; Eluma et al., '84; Paulson et al., '85, '88). One of VPA's major metabolites is trans-2-ene-valproic acid (Nau and Loscher, '84). In mouse models of epilepsy, trans-20 1991 WILEY-LISS, INC.
ene-VPA has been found to be equipotent t o VPA at inhibiting seizure activity (Nau and Loscher, '84; Loscher et al., '84; Loscher and Nau, '85) but is nonteratogenic (Nau, '86). If trans-2-ene-VPA could be confirmed to be therapeutically efficacious, but neither structurally nor functionally teratogenic, then it would have significant clinical potential. The purpose of this investigation was to assess the structural teratogenic potential of trans-2-ene-VPA in rats compared to VPA. Since trans-2-ene-VPA is not commercially available, it was first necessary to synthesize it. Plasma concentrations of the drugs and their major metabolites were also
Received September 24, 1990; accepted December 18, 1990. Address reprint requests to Charles V. Vorhees, PhD, Institute for Developmental Research, Children's Hospital Research Foundation, Cincinnati, OH 45229-2899.
584
C.V. VORHEES ET AL.
determined on the dams during gestation. Here we report that trans-2-ene-VPA does not induce malformations in rat fetuses at doses equal to and higher than one of VPA that does induce malformations. MATERIALS AND METHODS
esterification and distillation of the ethyl 2bromoester, 3) dehydrobromination and distillation of the 2-bromoester, and 4) saponification of the 2-ene-ester and purification of the trans-2-ene-VPA by repeated recrystallization.
Female Sprague-Dawley CD (VAF) rats 2-Bromovalproic acid and their progeny served as subjects Valproic acid (Aldrich Chemical Co., (Charles River, Portage, MI). Rats were ac- 107.5 g, 0.746 mole), bromine (128 g, 0.80 climated to the laboratory for at least 2 mole), and phosphorus tribromide (10 g, 45 weeks prior to breeding. Conception was in- mmole) were heated slowly until hydrogen ferred from discovery of a copulatory plug bromide generation had slowed, then at after overnight pairing of females with 100°C for 6 hr. The mixture was then evacmales of the same stock and supplier (plug uated using an aspirator to remove dissolved hydrogen bromide. The residue was = day EO). Once per day on days E7-18 trans-2-ene- water washed, dried by warming at reduced VPA was administered by gavage to dams pressure (aspirator), and distilled at reat doses of 0, 300, or 400 mg/kg (see below) duced pressure (0.05 mm) to give 140.2 g or VPA at 300 mg/kg (Sigma, St. Louis, MO) (84.3% yield) boiling at 92-100°C. dissolved in corn oil (2 ml/kg). Dams were weighed on days 0 , 7-18, and 20. All dams Ethyl 2-bromovalproate had access to food (Purina Rat Chow 5001) 2-Bromovalproic acid (139 g, 0.623 mole), and water ad libitum and were housed in a ethanol (78.3 g, 1.70 mole), benzene (700 vivarium fully accredited by the American ml), and concentrated sulfuric acid (7 g) Association for the Accreditation of Labora- were combined in a flask equipped with a tory Animal Care. Rats were maintained on Dean-Stark trap. The mixture was refluxed a 12 hr light/dark cycle at 21°C -+ 1°C and overnight to remove water generated by es50% 2 10% humidity. terification. The mixture was washed seOn day E18, half the dams in each group quentially with water, sodium bicarbonate were anesthetized with methoxyflurane solution, and water. After drying over anand a blood sample was drawn from the tail hydrous magnesium sulfate, the solvents vein for determination of drug concentra- were removed by evaporation under aspirations at 1 and 6 hr after their last dose. On tor pressure and the residue distilled E20 all dams were administered an anes- through an 8 in. glass helices-packed colthetic overdose and a laparotomy per- umn. Ethyl 2-bromovalproate was obtained formed. The uterine horns were exteriorized in 78% yield (122 g) boiling at 95-101°C and the contents examined for implantation (pressure 6 mm). sites. Each fetus was removed, sexed, weighed, and examined externally, and its Ethyl cis and trans-2-propyl-2-pentenoate uterine position was recorded. Two of three In a flask equipped in an 8 in. glass helifetuses were placed in Bouin’s solution for a ces-packed fractionating column was placed minimum of 2 weeks and then sliced free- ethyl 2-bromovalproate (120 g, 0.478 mole) hand and examined for visceral defects by and quinoline (185.2 g, 1.43mole). The mixthe method of Wilson (‘65). Every third fetus ture was heated with stirring up to 160in sequence was immersed in hot water and 170°C. After an hour, pressure was reduced the skin was removed, then the fetus was t o distill the mixture of esters. The distillate placed in 95% ethanol and later eviscerated. was washed sequentially with water, dilute Fetuses were then cleared in potassium hy- hydrochloric acid, and water. droxide and double stained with alizarin red acid S and alcian blue for skeletal examination. Trans-2-propyl-2-pentenoic The mixture of unsaturated esters (61.8 g, Synthesis 0.363 mole) was stirred in 300 ml 3 N sotrans-2-ene-Valproic acid (trans-2-propyl- dium hydroxide for l week. The aqueous so2-pentenoic acid) was prepared by the fol- lution was acidified with dilute hydrochloric lowing steps: 1)phosphorus tribromide cat- acid and the organic acids extracted by fouralyzed bromination of valproic acid, 2) fold extractions with 100 ml portions of
TERATOGENICITY OF trans-2-ENE-VALPROIC ACID IN RATS
ether. The combined extracts were water washed and the ether removed in a rotary evaporator. The organic residue was dissolved in an equal volume of chloroform and stored in the freezer (ca. -25°C) for 3 days. The crystalline trans acid was removed. The cis acid, largely in the liquid portion, was again stored without solvent removal in the freezer to harvest a smaller crop of trans crystals. By repeating the process using concentrated chloroform or ethyl acetate solutions and combining and recrystallizing the solid trans acid a total of six times the trans2-propyl-pentenoic acid was purified. The liquidus which contained the cis isomer along with some trans was not further purified.
585
tization by trimethylsilylation, and analysis by a GC-MS computer method employing a Megabore DB 1701 column.
Statistical procedures Frequency data were analyzed by Fisher's test for uncorrelated proportions (twotailed). Body weight, sex ratio, and incidences of defects per litter were analyzed using fixed-effect, factorial analyses of variance (ANOVA; general linear model). For maternal weights during gestation, a splitplot ANOVA was used, with day of gestation as a repeated measures factor. Orthogonal polynomials of the variance-covariance matrix in the split-plot analyses were tested for compound symmetry using a test for sphericity. Nonspherical analyses had F raStructural verification The product was converted to the trimeth- tios calculated using the Geisser-Greenmethod. Significant interactions were ylsilyl (TMS) ether derivative by reaction house with 50 p,1 of a mixture of hexamethyldisi- further analyzed by orthogonal decomposilazaneltrirnethylchlorosilaneipyridine(3:2:1 tion for trend. A posteriori comparisons of by volume) at 60°C for 30 min and analyzed significant group effects were conducted usby gas chromatography-mass spectrometry ing Duncan's multiple range test. (GC-MS). GC-MS was carried out on a FinRESULTS nagan 4635 quadrupole instrument housing a 30 m x 0.4 mm SE54 capillary column, For maternal body weight during gestaoperated with a temperature programmed tion the only noteworthy effect was a signifoperation from 50 to 150°C in increments of icant group X day interaction (F = 2.13 5"C/min after an initial isothermal period of [39/4421; P < .02; Fig. 1). Simple-effect 5 min. Helium was the carrier gas at a flow ANOVAs revealed no group differences on rate of 2 ml/min. Electron ionization (70 eV) any single day of gestation; therefore, ormass spectra were continuously recorded thogonal decomposition analyses were conover the mass range 50-800 Daltons (2 sec/ ducted. This analysis revealed a marginally significant linear trend (P = .059), a signifcycle). The mass spectrum of the TMS ether de- icant quadratic trend (P < .Ol), and no sigrivative revealed a molecular ion at m/z nificant cubic trend (trends of higher order 214, in agreement with the molecular were not considered). To understand better weight of the parent compound, trans-&pro- these effects, each treatment group was pyl-2-pentenoic acid. The base peak was at compared with controls. No trend effects m/z 199 resulting from the loss of methyl were found for the trans-2-ene-VPA 300 mg/ ([M-CH,l+) and prominent ions at m/z 185 kg vs. control comparison. For the truns-2([M-C,H,l+), m/z 169, and m/z 124 (M- ene-VPA 400 mg/kg vs. control analysis the OTMS) were evident. The mass spectrum of only significant effect was a linear trend (P this compound was identical t o the mass < .05; Fig. 2). For the VPA vs. control analspectrum of the TMS ether derivative pre- ysis both the linear and quadratic trends viously published (Nau et al., '81b) for the were significant (P < .05 and P < .01, retrans-2-ene isomer of valproic acid. spectively). The linear component was similar to that seen in the trans-2-ene-VPA 400 Serum drug determinations mg/kg group and is shown in Figure 3 (left). Drug and metabolite concentrations in The quadratic trend in the VPA group is maternal serum were determined by the shown in Figure 3 (right). All curves repremethod of Nau et al. ('81a) using ethyl sent the lines of best fit calculated by the acetate extraction of the samples at pH 5.0 method of least squares. (following addition of 2-methyl-2-ethylFor fetal body weight on E20, significant hexanoic acid as internal standard), deriva- effects were obtained for treatment group (F
586
C.V. VORHEES ET AL
Gestation Weights
"'1
0-0
-
Control
.--.
A -A
100
0
7
8
0
VPA-300 t-A2-WA-300 t-A2-WA-400
10 11 12 13 14 15 16 17 18
20
Day
Fig. 1. Mean t SEM maternal body weight in dams during gestation. There was a significant treatment group X day interaction (see Figs. 2 and 3 for further details).
Gestation Weights
Day
Fig. 2. Mean ? SEM maternal body weight in dams during gestation with the best linear fit for each group by the method of least squares. The lines are significantly different by orthogonal decomposition analysis at P < .05. = 48.24 [3/66];P < .0001) and sex (F = 6.82 [1/661; P < .02), whereas the group x sex interaction was not significant. The sex effect was caused by the fact that males were heavier than females, but, since the interaction was not significant, group effects are shown with males and females combined (Fig. 4). A posteriori Duncan comparisons revealed that the VPA-300 group was the lightest and significantly lighter than controls (25.1%;P < .Ol), the trans-2-ene-VPA400 group was next lighter and also signif-
icantly lighter than controls (7.4%;P < .Ol), whereas the trans-2-ene-VPA-300 group was not different from controls. There were no group differences in number of implantddam, number of resorptions, number of live fetusedlitter, or sex ratio (not shown) among the various treatment groups (Table 1).The proportion of fetuses malformed, both total and visceral, was increased in the VPA group (both P < .01)but was not different from controls in either of the trans-2-me-VPA groups. Skeletal de-
587
TERATOGENICITY OF trans-2-ENE-VALPROICACID IN RATS
Gestation Weights 400
4
I
360 -
,. m
v
,
1
.
r
.-
rn 320-
5
: 280:
I :: 0
8
10 12
14
16
18 20
0
8
10
12
14
16
18 20
Day
Fig. 3. Mean 2 SEM maternal body weight in dams during gestation with the best linear fit (left)and best quadratic fit (right)for each group by the method of
least squares. The linesicurves in both panels are significantly different by orthogonal decomposition analysis at P < .05 (left) and P < .01 (right).
Fetal Weight 0 Control
aVPA-300 3.500
I t-A*-VPA-300
W l t-Az-VPA-400
3.000 2.500 A
0
v
+
r cn .-
2.000
0
3
1.500 1 .ooo 0.500
0.000
Fig. 4. Mean f SEM fetal body weight on E20 by litter. The effect of treatment group was significant by ANOVA (P < .0001). **P< .01 compared to controls by Duncan comparisons.
fects were not increased in any group. The same pattern was seen when malformations were analyzed by litter (proportion affected per litter; Table 1).Visceral abnormalities seen in the VPA group were hydronephrosis (eight cases), edema (seven cases), and one case of cardiac septa1 defect. The two skeletal defects in this group were one with extra 14th rib and one with missing lumbar and sacral vertebrae. In the trans-2-eneVPA-400 group the one malformation seen was an omphalocele. In the trans-2-eneVPA-300 group there were three fetuses with edema, one with hydronephrosis, and
one with shortened tail. Among the controls the only abnormality seen was a single case of edema. Maternal serum drug and metabolite concentrations obtained at 1 and 6 hr after the last dosing on E l 8 are shown in Table 2. Concentrations of the administered compounds were higher at 1hr posttreatment in the VPA group than in either of the trans2-ene-VPA groups, whereas at 6 hr posttreatment, the pattern was reversed, although the differences were smaller. Very little of the administered VPA was transformed to cis or trans-2-ene-VPA,and very
588
C.V. VORHEES ET AL.
TABLE 1. Effects of trans-2-ene-ualproic acid (trans-2-ene-VPA) on reproductive outcome and fetal development in dams treated on days E7-El8 of gestation Group VPA-300 t-2-ene-VPA-300 t-2-ene-VPA-400 (300 rng/kg) (400 mglkgf Control (300 mglkg) Measure 8 8 10 12 No. of litters No. of implants 151 174 122 119 No. of implantddam 15.1 i 0.5% 14.5 t 0.6 15.3 t 1.2 14.9 f 0.6 No. resorbeditot 121151 24/174 71122 8/119 7.9 13.8 5.7 6.7 Resorbed (%) No. live fetuses 139 150 115 111 14.4 t 1.2 13.9 2 0.8 No. live fetusesidam 13.9 i 0.6" 12.5 t 0.8 Aggregate 1/111(0.9) 5/115 (4.31 181150 (12.0)** No. (%) total 11139 (0.7) 2/47 (4.3) 1/35(2.9) 0134 (0.0) No. (%) skeletal 0142 (0.0) 1/77 (1.3) 16/103 (15.5)** 4/80 (5.0) No. (%) visceral 1/97 (1.0) By litter 3.8 i 1.6 0.7 f 0.7 0.8 t 0.8 13.4 4 2.8** Total 1.0 4 1.0 4.2 ? 2.2 1.1 f 1.1 16.8 4 3.4** Visceral *Mean +- SEM. **P< .01 compared with control by Fisher's test for uncorrelated proportions for aggregate proportions or by ANOVA and Duncan group comparisons for mean percent (?SEMI malformed per litter.
TABLE 2. Serum trans-2-ene-valproic acid (trans-2-ene-VPA) and valproic acid (VPAI concentrations in pg lml in dams on E l 8 1 and 6 h r after receiving their last dose VPA trans-2-ene-VPA trans-2-ene-VPA (300 m$kg) (300 mg/kg) (400 mgikg) Groiin (6) (4) 14) 1 Hour posttreatment 1.4 t 0.9 0.2 t 0.04 241.7 i 17.9 VPA (tot.) -* -* 134.8 f 6.3 VPA (unbound) 140.2 t 23.3 4.4 t 0.8 128.6 ? 25.0 trans-2-ene (tot.) -* 37.6 t 11.1 40.4 t 11.5 trans-2-ene (unb.) -* 14.8 t 7.0 6.1 t 1.4 cis-2-ene 1.1 t 0.1 1.0 i 0.1 1.8 t 0.3 2,3-diene 0.6 t 0.1 0.9 4 0.1 1.3 i 0.5 2,4-diene 1.4 t 0.2 1.5 t- 0.3 0.2 ? 0.2 3-ene 0.3 t 0.1 0.02 t 0.02 0.4 t 0.1 3-keto 6 Hours posttreatment 0.2 t 0.1 0.4 t- 0.05 61.9 t 12.2 VPA (tot.) -* -* VPA (unbound) 22.7 t 7.7 92.3 t 17.4 3.8 i 0.7 86.1 t 3.0 trans-2-ene (tot.) -* 14.2 t 4.0 trans-2-ene (unb.) 13.5 i 2.2 -* 2.4 t 0.2 2.7 t- 0.4 cis-2-ene 2.3 t 0.1 0.6 t 0.2 1.7 lr 0.1 2,3-diene * 0.7 t 0.1 0.8 t 0.1 2,4-diene * 1.5 t 0.1 1.2 t 0.4 3-ene 0.6 t 0.1 0.8 f 0.3 0.02 i 0.01 3-keto 'Number of dams per group from whom blood samples were obtained. *Concentrationsat or below detection limit (mean ? SEMI.
little of the administered trans-2-ene-VPA was transformed to VPA. Other metabolites were found in very low concentrations at both time points. DISCUSSION
The present data on VPA confirm previous findings of the teratogenicity of this agent in rats (Ong et al., '83; Vorhees, '87a). By contrast, equal and larger doses of trans2-ene-VPA produced no evidence of teratogenicity in rats. This finding is in agree-
ment with similar data obtained in mice (Nau and Loscher, '84; Nau, '86). Since trans-2-ene-VPA is equipotent with VPA at inhibiting induced seizures in mice (Nau and Loscher, '84; Loscher et al., '84), the current data add support to the view that trans2-ene-VPA represents a n important alternative antiepileptic therapy to VPA. However, it should be noted that VPA is also associated with a syndrome of effects, the fetal valproate syndrome, which includes delayed neurobehavioral develop-
TERATOGENICITY OF trans-2-ENE-VALPROIC ACID IN RATS
ment in humans (DiLiberti et al., '84;Ardinger et al., '88). In rats, VPA is also associated with evidence of behavioral teratogenicity at doses below those inducing malformations (Vorhees, '87b). Therefore, caution about the developmental toxicity of trans-2-ene-VPA should be exercised until evidence concerning its behavioral teratogenic potential is available. The pharmacokinetic data obtained in the present experiment were limited but suggested different patterns for VPA and trans2-ene-VPA. It has previously been shown in mice that NaVPA and trans-2-ene-VPA given S.C.produce similar peak concentrations at about 0.5 hr posttreatment (Nau, '86). However, clearance rates of trans2-ene-VPA were higher than for VPA and involved two phases rather than one (tllz,, = 0.5 and tll,,p = 1.4 hr. for trans-2-eneVPA vs. t,,, = 1.3 hr for VPA). Furthermore, in mice, unbound trans-2-ene-VPA was cleared more rapidly than unbound VPA (Nau, '86). In the present experiment, VPA was higher at 1 hr and lower at 6 h r in the VPA group than was trans-2-ene-VPA in the trans-2-ene-VPA groups. Although the mouse and rat data are not directly comparable, one possibility is that trans-2-eneVPA may be more rapidly absorbed than VPA. However, because of its different protein binding and distribution patterns, trans-2-ene-VPA may be at a different point in its elimination phase at 6 hr than VPA. Another possibility is that the difference is related to the stage of development at which measurements were made. In mice they were taken on E8 and E9, whereas here they were taken in rats on E18. Since liver and kidney function and volume of distribution in the dam change late in gestation, the present findings may reflect such inf luences. Further pharmacokinetic data on trans-2-ene-VPA in rats during gestation will be required to clarify this point. ACKNOWLEDGMENTS
The authors thank Dr. Frank Abbott, Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, for providing advice and methods for the synthesis of trans-2-ene-valproic acid and to Dr. Leland C. Clark, Jr., for use of his laboratory in which to conduct the synthesis. The authors also thank M.S. Moran and C.A. Blanton for assistance in the in vivo, teratological, and statistical aspects of this re-
589
search. The research was supported in part by Public Health Service grant HD21806 from the NIH. LITERATURE CITED Ardinger, H.H., J.F. Atkin, R.D., Blackston, L.J. Elsas, S.K. Clarren, S. Livingstone, D.B. Flannery, J.M. Pellock, M.J. Harrod, E.J. Lammer, F. Majewski, A. Schnizel, H.V. Toriello, and J.W. Hanson (1988)Verification of the fetal valproate syndrome phenotype. Am. J . Med. Genet., 29:171-185. Bjerkedel, T., A. Czeizel, J . Goujard, B. Kallen, P. Mastroiacovo, N. Nevin, G. Oakley, and E. Robert (1982) Valproic acid and spina bifida. Lancet, 2:1096-1097. Brown, N.A., J. Kao, and S. Fabro (1980) Teratogenic potential of valproic acid. Lancet, 1 :660-661. DiLiberti, J.H., P.A. Farndon, N.R. Dennis, and C.J.R. Curry (1984) The fetal valproate syndrome. Am. J . Med. Genet., 19:473-481. Eluma, F.O., M.E. Sucheston, T.G. Hayes, and R.B. Paulson (1984) Teratogenic effects of dosage levels and time of administration of carbamazepine, sodium valproate, and diphenylhydantoin on craniofacial development in the CD-1 mouse fetus. J . Craniofac. Genet. Dev. Biol., 4:191-210. Jager-Roman, E., A. Deichl, S. Jakob, A.-M. Hartmann, S. Koch, D. Rating, R. Steldinger, H. Nau, and H. Helge (1986) Fetal growth, major malformations, and minor anomalies in infants born to women receiving valproic acid. J . Pediatr., 108:997-1004. Kao, J., N.A. Brown, B. Schmid, E.H. Goulding, and S. Fabro (1981)Teratogenicity of valproic acid: In vivo and in vitro investigations. Teratogen. Carcinogen. Mutagen., 1 :367-382. Lindhout, D., and H. Meinardi (1984) Spina bifida and in utero exposure to valproate. Lancet, 2:396. Loscher, W., and H. Nau (1985) Pharmacological evaluation of various metabolites and analogues of valproic acid-Anticonvulsant and toxic potencies in mice. Neuropharmacology, 24:427-435. Loscher, W., H. Nau, C . Marescaux, and M. Vergnes (1984) Comparative evaluation of anticonvulsant and toxic potencies of valproic acid and 2-en-valproic acid in different animal models of epilepsy. Eur. J . Pharmacol., 99:211-218. Mastroiacovo, P., R. Bertollini, S. Morandini, and G. Segni (1983) Maternal epilepsy, valproate exposure, and birth defects. Lancet, 2:1499. Nau, H. (1986) Transfer of valproic acid and its main active unsaturated metabolite to the gestational tissue: Correlation with neural tube defect formation in the mouse. Teratology, 33:21-27. Nau, H., and W. Loscher (1984) Valproic acid and metabolites: Pharmacological and toxicological studies. Epilepsia, 25[Suppl. l]:S14S22. Nau, H., W. Wittfoht, H. Schafer, C. Jakobs, D. Rating, and H. Helge (1981a) Valproic acid and several metabolites: Quantitative determination in serum, urine, breast milk and tissues by gas chromatography-mass spectrometry using selected ion monitoring. J . Chromatogr., 226;69-78. Nau, H., R. Zierer, H. Spielmann, D. Neubert, and Ch. Gansau (1981b) A new model for embryotoxicity testing: Teratogenicity and pharmacokinetics of valproic acid following constant-rate administration in the mouse using human therapeutic drug and metabolite concentrations. Life Sci., 29:2803-2814. Ong, L.L., J.L. Schardein, J.A. Petrere, R. Sakowski, H. Jordan, R.R. Humphrey, J.E. Fitzgerald, and F.A. de
590
C.V. VORHEES ET AL.
la Iglesia (1983) Teratogenesis of calcium valproate in rats. Fund. Appl. Toxicol., 3r121-126. Paulson, R.B., M.E. Sucheston, T.G. Hayes, and G.W. Paulson (1985) Teratogenic effects of valproate in the CD-1 mouse fetus. Arch. Neurol., 42:980-983. Paulson, R.B., M.E. Sucheston, T.G. Hayes, H.S. Weiss, L.A. Sachs, M. Oca, B. Kernan, and S. Weiss (1988) Effects of sodium valproate and oxygen on the CD-1 mouse fetus. J. Craniofac. Genet. Dev. Biol., 8:35-45. Robert, E., and P. Guibaud (1982) Maternal valproic acid and congenital neural tube defects. Lancet, 2: 937. Vorhees, C.V. (1987a) Teratogenicity and developmental toxicity of valproic acid in rats. Teratology, 35: 195-202.
Vorhees, C.V. (1987b) Behavioral teratogenicity of valproic acid: Selective effects on behavior after prenatal exposure to rats. Psychopharmacology, 92:173-179. Whittle, B.A. (1975) Pre-clinical teratological studies on sodium valproate (Epilim) and other anticonvulsants. In: Clinical and Pharmacological Aspects of Sodium Valproate (Epilim) in the Treatment of Epilepsy. N.J. Legg, ed. MCS Consultants Press, Tunbridge Wells, England, pp. 106-111. Wilson, J.G. (1965) Embryological considerations in teratology. In: Teratology: Principles and Techniques. J.G. Wilson and J. Warkany, eds. University of Chicago Press, Chicago, pp. 251-277.
Lack of Teratogenicity of trans-2-ene-Valproic Acid Compared to Valproic Acid in Rats CHARLES V. VORHEES, KAREN D. ACUFF-SMITH, WALTER P. WEISENBURGER, DANIEL R. MINCK, JAMES S. BERRY, KENNETH D.R. SETCHELL, AND HEINZ NAU Institute for Developmental Research, Children's Hospital Research Foundation and University of Cincinnati, Cincinnati, Ohio 45229-2899 (C.V.V.,K.D.A.-S., W.P.W., D.R.M.);482 Wellesley Drive, Cincinnati, Ohio 45224 (J.S.B.);Mass Spectrometry Laboratory, Children's Hospital Research Foundation, Cincinnati, Ohio 45229 (K.D.R.S.);Institut fur Toxikologie und Embryopharmakologie, Freie Uniuersitat Berlin, Berlin 33, Germany (H.N.)
ABSTRACT The teratogenicity of trans-2-ene-valproic acid (300 and 400 mg/kg) was compared with that of valproic acid (VPA; 300 mg/kg) and controls (corn oil) administered by gavage to Sprague-Dawley CD rats on embryonic (E) days 7-18. At the 300 mg/kg dose, trans-2-ene-VPA produced no change in maternal weight, number of implantations, proportion of resorptions, proportion of malformations, or fetal weight. By contrast, the same dose of VPA (300 mg/kg) reduced maternal weight during gestation, increased malformations (12.0% vs. 0.7% in controls), and reduced fetal body weight by 25.1%. An even higher dose of trans-2-ene-VPA (400 mg/kg) produced a reduction in maternal body weight during treatment and reduced fetal body weight (by 7.9%)' but did not increase resorptions or malformations in the fetuses. On day E18, maternal serum drug concentrations of VPA were higher in the VPA-treated group compared with those of trans-2-ene-VPA in the trans-2-ene-VPA-treated groups at 1 hr posttreatment. At 6 hr posttreatment the reverse was seen. trans-2-eneVPA may be absorbed more rapidly and distributed differently than VPA. Overall, the data support the view that trans-2-ene-VPA at equal or higher doses than VPA is not teratogenic in rats. The antiepileptic valproic acid (VPA) is associated with a prevalence of spina bifida of -1.3% in infants exposed in utero (Bjerkedel et al., '82; Jager-Roman et al., '86; Lindhout and Meinardi, '84; MastroiaCOVO et al., '83; Robert and Guibaud, '82). In addition, a pattern of minor dysmorphological and dysfunctional effects has been associated with intrauterine VPA exposure and has been termed the fetal valproate syndrome (FVS; DiLiberti et al., '84). Recently, the FVS has been corroborated and the effects associated with it expanded (Ardinger et al., '88). VPA is teratogenic in rats (Ong et al., '83; Vorhees, '87a) and mice (Whittle, '75; Brown et al., '80; Kao et al., '81; Nau et al., '81b; Eluma et al., '84; Paulson et al., '85, '88). One of VPA's major metabolites is trans-2-ene-valproic acid (Nau and Loscher, '84). In mouse models of epilepsy, trans-20 1991 WILEY-LISS, INC.
ene-VPA has been found to be equipotent t o VPA at inhibiting seizure activity (Nau and Loscher, '84; Loscher et al., '84; Loscher and Nau, '85) but is nonteratogenic (Nau, '86). If trans-2-ene-VPA could be confirmed to be therapeutically efficacious, but neither structurally nor functionally teratogenic, then it would have significant clinical potential. The purpose of this investigation was to assess the structural teratogenic potential of trans-2-ene-VPA in rats compared to VPA. Since trans-2-ene-VPA is not commercially available, it was first necessary to synthesize it. Plasma concentrations of the drugs and their major metabolites were also
Received September 24, 1990; accepted December 18, 1990. Address reprint requests to Charles V. Vorhees, PhD, Institute for Developmental Research, Children's Hospital Research Foundation, Cincinnati, OH 45229-2899.
584
C.V. VORHEES ET AL.
determined on the dams during gestation. Here we report that trans-2-ene-VPA does not induce malformations in rat fetuses at doses equal to and higher than one of VPA that does induce malformations. MATERIALS AND METHODS
esterification and distillation of the ethyl 2bromoester, 3) dehydrobromination and distillation of the 2-bromoester, and 4) saponification of the 2-ene-ester and purification of the trans-2-ene-VPA by repeated recrystallization.
Female Sprague-Dawley CD (VAF) rats 2-Bromovalproic acid and their progeny served as subjects Valproic acid (Aldrich Chemical Co., (Charles River, Portage, MI). Rats were ac- 107.5 g, 0.746 mole), bromine (128 g, 0.80 climated to the laboratory for at least 2 mole), and phosphorus tribromide (10 g, 45 weeks prior to breeding. Conception was in- mmole) were heated slowly until hydrogen ferred from discovery of a copulatory plug bromide generation had slowed, then at after overnight pairing of females with 100°C for 6 hr. The mixture was then evacmales of the same stock and supplier (plug uated using an aspirator to remove dissolved hydrogen bromide. The residue was = day EO). Once per day on days E7-18 trans-2-ene- water washed, dried by warming at reduced VPA was administered by gavage to dams pressure (aspirator), and distilled at reat doses of 0, 300, or 400 mg/kg (see below) duced pressure (0.05 mm) to give 140.2 g or VPA at 300 mg/kg (Sigma, St. Louis, MO) (84.3% yield) boiling at 92-100°C. dissolved in corn oil (2 ml/kg). Dams were weighed on days 0 , 7-18, and 20. All dams Ethyl 2-bromovalproate had access to food (Purina Rat Chow 5001) 2-Bromovalproic acid (139 g, 0.623 mole), and water ad libitum and were housed in a ethanol (78.3 g, 1.70 mole), benzene (700 vivarium fully accredited by the American ml), and concentrated sulfuric acid (7 g) Association for the Accreditation of Labora- were combined in a flask equipped with a tory Animal Care. Rats were maintained on Dean-Stark trap. The mixture was refluxed a 12 hr light/dark cycle at 21°C -+ 1°C and overnight to remove water generated by es50% 2 10% humidity. terification. The mixture was washed seOn day E18, half the dams in each group quentially with water, sodium bicarbonate were anesthetized with methoxyflurane solution, and water. After drying over anand a blood sample was drawn from the tail hydrous magnesium sulfate, the solvents vein for determination of drug concentra- were removed by evaporation under aspirations at 1 and 6 hr after their last dose. On tor pressure and the residue distilled E20 all dams were administered an anes- through an 8 in. glass helices-packed colthetic overdose and a laparotomy per- umn. Ethyl 2-bromovalproate was obtained formed. The uterine horns were exteriorized in 78% yield (122 g) boiling at 95-101°C and the contents examined for implantation (pressure 6 mm). sites. Each fetus was removed, sexed, weighed, and examined externally, and its Ethyl cis and trans-2-propyl-2-pentenoate uterine position was recorded. Two of three In a flask equipped in an 8 in. glass helifetuses were placed in Bouin’s solution for a ces-packed fractionating column was placed minimum of 2 weeks and then sliced free- ethyl 2-bromovalproate (120 g, 0.478 mole) hand and examined for visceral defects by and quinoline (185.2 g, 1.43mole). The mixthe method of Wilson (‘65). Every third fetus ture was heated with stirring up to 160in sequence was immersed in hot water and 170°C. After an hour, pressure was reduced the skin was removed, then the fetus was t o distill the mixture of esters. The distillate placed in 95% ethanol and later eviscerated. was washed sequentially with water, dilute Fetuses were then cleared in potassium hy- hydrochloric acid, and water. droxide and double stained with alizarin red acid S and alcian blue for skeletal examination. Trans-2-propyl-2-pentenoic The mixture of unsaturated esters (61.8 g, Synthesis 0.363 mole) was stirred in 300 ml 3 N sotrans-2-ene-Valproic acid (trans-2-propyl- dium hydroxide for l week. The aqueous so2-pentenoic acid) was prepared by the fol- lution was acidified with dilute hydrochloric lowing steps: 1)phosphorus tribromide cat- acid and the organic acids extracted by fouralyzed bromination of valproic acid, 2) fold extractions with 100 ml portions of
TERATOGENICITY OF trans-2-ENE-VALPROIC ACID IN RATS
ether. The combined extracts were water washed and the ether removed in a rotary evaporator. The organic residue was dissolved in an equal volume of chloroform and stored in the freezer (ca. -25°C) for 3 days. The crystalline trans acid was removed. The cis acid, largely in the liquid portion, was again stored without solvent removal in the freezer to harvest a smaller crop of trans crystals. By repeating the process using concentrated chloroform or ethyl acetate solutions and combining and recrystallizing the solid trans acid a total of six times the trans2-propyl-pentenoic acid was purified. The liquidus which contained the cis isomer along with some trans was not further purified.
585
tization by trimethylsilylation, and analysis by a GC-MS computer method employing a Megabore DB 1701 column.
Statistical procedures Frequency data were analyzed by Fisher's test for uncorrelated proportions (twotailed). Body weight, sex ratio, and incidences of defects per litter were analyzed using fixed-effect, factorial analyses of variance (ANOVA; general linear model). For maternal weights during gestation, a splitplot ANOVA was used, with day of gestation as a repeated measures factor. Orthogonal polynomials of the variance-covariance matrix in the split-plot analyses were tested for compound symmetry using a test for sphericity. Nonspherical analyses had F raStructural verification The product was converted to the trimeth- tios calculated using the Geisser-Greenmethod. Significant interactions were ylsilyl (TMS) ether derivative by reaction house with 50 p,1 of a mixture of hexamethyldisi- further analyzed by orthogonal decomposilazaneltrirnethylchlorosilaneipyridine(3:2:1 tion for trend. A posteriori comparisons of by volume) at 60°C for 30 min and analyzed significant group effects were conducted usby gas chromatography-mass spectrometry ing Duncan's multiple range test. (GC-MS). GC-MS was carried out on a FinRESULTS nagan 4635 quadrupole instrument housing a 30 m x 0.4 mm SE54 capillary column, For maternal body weight during gestaoperated with a temperature programmed tion the only noteworthy effect was a signifoperation from 50 to 150°C in increments of icant group X day interaction (F = 2.13 5"C/min after an initial isothermal period of [39/4421; P < .02; Fig. 1). Simple-effect 5 min. Helium was the carrier gas at a flow ANOVAs revealed no group differences on rate of 2 ml/min. Electron ionization (70 eV) any single day of gestation; therefore, ormass spectra were continuously recorded thogonal decomposition analyses were conover the mass range 50-800 Daltons (2 sec/ ducted. This analysis revealed a marginally significant linear trend (P = .059), a signifcycle). The mass spectrum of the TMS ether de- icant quadratic trend (P < .Ol), and no sigrivative revealed a molecular ion at m/z nificant cubic trend (trends of higher order 214, in agreement with the molecular were not considered). To understand better weight of the parent compound, trans-&pro- these effects, each treatment group was pyl-2-pentenoic acid. The base peak was at compared with controls. No trend effects m/z 199 resulting from the loss of methyl were found for the trans-2-ene-VPA 300 mg/ ([M-CH,l+) and prominent ions at m/z 185 kg vs. control comparison. For the truns-2([M-C,H,l+), m/z 169, and m/z 124 (M- ene-VPA 400 mg/kg vs. control analysis the OTMS) were evident. The mass spectrum of only significant effect was a linear trend (P this compound was identical t o the mass < .05; Fig. 2). For the VPA vs. control analspectrum of the TMS ether derivative pre- ysis both the linear and quadratic trends viously published (Nau et al., '81b) for the were significant (P < .05 and P < .01, retrans-2-ene isomer of valproic acid. spectively). The linear component was similar to that seen in the trans-2-ene-VPA 400 Serum drug determinations mg/kg group and is shown in Figure 3 (left). Drug and metabolite concentrations in The quadratic trend in the VPA group is maternal serum were determined by the shown in Figure 3 (right). All curves repremethod of Nau et al. ('81a) using ethyl sent the lines of best fit calculated by the acetate extraction of the samples at pH 5.0 method of least squares. (following addition of 2-methyl-2-ethylFor fetal body weight on E20, significant hexanoic acid as internal standard), deriva- effects were obtained for treatment group (F
586
C.V. VORHEES ET AL
Gestation Weights
"'1
0-0
-
Control
.--.
A -A
100
0
7
8
0
VPA-300 t-A2-WA-300 t-A2-WA-400
10 11 12 13 14 15 16 17 18
20
Day
Fig. 1. Mean t SEM maternal body weight in dams during gestation. There was a significant treatment group X day interaction (see Figs. 2 and 3 for further details).
Gestation Weights
Day
Fig. 2. Mean ? SEM maternal body weight in dams during gestation with the best linear fit for each group by the method of least squares. The lines are significantly different by orthogonal decomposition analysis at P < .05. = 48.24 [3/66];P < .0001) and sex (F = 6.82 [1/661; P < .02), whereas the group x sex interaction was not significant. The sex effect was caused by the fact that males were heavier than females, but, since the interaction was not significant, group effects are shown with males and females combined (Fig. 4). A posteriori Duncan comparisons revealed that the VPA-300 group was the lightest and significantly lighter than controls (25.1%;P < .Ol), the trans-2-ene-VPA400 group was next lighter and also signif-
icantly lighter than controls (7.4%;P < .Ol), whereas the trans-2-ene-VPA-300 group was not different from controls. There were no group differences in number of implantddam, number of resorptions, number of live fetusedlitter, or sex ratio (not shown) among the various treatment groups (Table 1).The proportion of fetuses malformed, both total and visceral, was increased in the VPA group (both P < .01)but was not different from controls in either of the trans-2-me-VPA groups. Skeletal de-
587
TERATOGENICITY OF trans-2-ENE-VALPROICACID IN RATS
Gestation Weights 400
4
I
360 -
,. m
v
,
1
.
r
.-
rn 320-
5
: 280:
I :: 0
8
10 12
14
16
18 20
0
8
10
12
14
16
18 20
Day
Fig. 3. Mean 2 SEM maternal body weight in dams during gestation with the best linear fit (left)and best quadratic fit (right)for each group by the method of
least squares. The linesicurves in both panels are significantly different by orthogonal decomposition analysis at P < .05 (left) and P < .01 (right).
Fetal Weight 0 Control
aVPA-300 3.500
I t-A*-VPA-300
W l t-Az-VPA-400
3.000 2.500 A
0
v
+
r cn .-
2.000
0
3
1.500 1 .ooo 0.500
0.000
Fig. 4. Mean f SEM fetal body weight on E20 by litter. The effect of treatment group was significant by ANOVA (P < .0001). **P< .01 compared to controls by Duncan comparisons.
fects were not increased in any group. The same pattern was seen when malformations were analyzed by litter (proportion affected per litter; Table 1).Visceral abnormalities seen in the VPA group were hydronephrosis (eight cases), edema (seven cases), and one case of cardiac septa1 defect. The two skeletal defects in this group were one with extra 14th rib and one with missing lumbar and sacral vertebrae. In the trans-2-eneVPA-400 group the one malformation seen was an omphalocele. In the trans-2-eneVPA-300 group there were three fetuses with edema, one with hydronephrosis, and
one with shortened tail. Among the controls the only abnormality seen was a single case of edema. Maternal serum drug and metabolite concentrations obtained at 1 and 6 hr after the last dosing on E l 8 are shown in Table 2. Concentrations of the administered compounds were higher at 1hr posttreatment in the VPA group than in either of the trans2-ene-VPA groups, whereas at 6 hr posttreatment, the pattern was reversed, although the differences were smaller. Very little of the administered VPA was transformed to cis or trans-2-ene-VPA,and very
588
C.V. VORHEES ET AL.
TABLE 1. Effects of trans-2-ene-ualproic acid (trans-2-ene-VPA) on reproductive outcome and fetal development in dams treated on days E7-El8 of gestation Group VPA-300 t-2-ene-VPA-300 t-2-ene-VPA-400 (300 rng/kg) (400 mglkgf Control (300 mglkg) Measure 8 8 10 12 No. of litters No. of implants 151 174 122 119 No. of implantddam 15.1 i 0.5% 14.5 t 0.6 15.3 t 1.2 14.9 f 0.6 No. resorbeditot 121151 24/174 71122 8/119 7.9 13.8 5.7 6.7 Resorbed (%) No. live fetuses 139 150 115 111 14.4 t 1.2 13.9 2 0.8 No. live fetusesidam 13.9 i 0.6" 12.5 t 0.8 Aggregate 1/111(0.9) 5/115 (4.31 181150 (12.0)** No. (%) total 11139 (0.7) 2/47 (4.3) 1/35(2.9) 0134 (0.0) No. (%) skeletal 0142 (0.0) 1/77 (1.3) 16/103 (15.5)** 4/80 (5.0) No. (%) visceral 1/97 (1.0) By litter 3.8 i 1.6 0.7 f 0.7 0.8 t 0.8 13.4 4 2.8** Total 1.0 4 1.0 4.2 ? 2.2 1.1 f 1.1 16.8 4 3.4** Visceral *Mean +- SEM. **P< .01 compared with control by Fisher's test for uncorrelated proportions for aggregate proportions or by ANOVA and Duncan group comparisons for mean percent (?SEMI malformed per litter.
TABLE 2. Serum trans-2-ene-valproic acid (trans-2-ene-VPA) and valproic acid (VPAI concentrations in pg lml in dams on E l 8 1 and 6 h r after receiving their last dose VPA trans-2-ene-VPA trans-2-ene-VPA (300 m$kg) (300 mg/kg) (400 mgikg) Groiin (6) (4) 14) 1 Hour posttreatment 1.4 t 0.9 0.2 t 0.04 241.7 i 17.9 VPA (tot.) -* -* 134.8 f 6.3 VPA (unbound) 140.2 t 23.3 4.4 t 0.8 128.6 ? 25.0 trans-2-ene (tot.) -* 37.6 t 11.1 40.4 t 11.5 trans-2-ene (unb.) -* 14.8 t 7.0 6.1 t 1.4 cis-2-ene 1.1 t 0.1 1.0 i 0.1 1.8 t 0.3 2,3-diene 0.6 t 0.1 0.9 4 0.1 1.3 i 0.5 2,4-diene 1.4 t 0.2 1.5 t- 0.3 0.2 ? 0.2 3-ene 0.3 t 0.1 0.02 t 0.02 0.4 t 0.1 3-keto 6 Hours posttreatment 0.2 t 0.1 0.4 t- 0.05 61.9 t 12.2 VPA (tot.) -* -* VPA (unbound) 22.7 t 7.7 92.3 t 17.4 3.8 i 0.7 86.1 t 3.0 trans-2-ene (tot.) -* 14.2 t 4.0 trans-2-ene (unb.) 13.5 i 2.2 -* 2.4 t 0.2 2.7 t- 0.4 cis-2-ene 2.3 t 0.1 0.6 t 0.2 1.7 lr 0.1 2,3-diene * 0.7 t 0.1 0.8 t 0.1 2,4-diene * 1.5 t 0.1 1.2 t 0.4 3-ene 0.6 t 0.1 0.8 f 0.3 0.02 i 0.01 3-keto 'Number of dams per group from whom blood samples were obtained. *Concentrationsat or below detection limit (mean ? SEMI.
little of the administered trans-2-ene-VPA was transformed to VPA. Other metabolites were found in very low concentrations at both time points. DISCUSSION
The present data on VPA confirm previous findings of the teratogenicity of this agent in rats (Ong et al., '83; Vorhees, '87a). By contrast, equal and larger doses of trans2-ene-VPA produced no evidence of teratogenicity in rats. This finding is in agree-
ment with similar data obtained in mice (Nau and Loscher, '84; Nau, '86). Since trans-2-ene-VPA is equipotent with VPA at inhibiting induced seizures in mice (Nau and Loscher, '84; Loscher et al., '84), the current data add support to the view that trans2-ene-VPA represents a n important alternative antiepileptic therapy to VPA. However, it should be noted that VPA is also associated with a syndrome of effects, the fetal valproate syndrome, which includes delayed neurobehavioral develop-
TERATOGENICITY OF trans-2-ENE-VALPROIC ACID IN RATS
ment in humans (DiLiberti et al., '84;Ardinger et al., '88). In rats, VPA is also associated with evidence of behavioral teratogenicity at doses below those inducing malformations (Vorhees, '87b). Therefore, caution about the developmental toxicity of trans-2-ene-VPA should be exercised until evidence concerning its behavioral teratogenic potential is available. The pharmacokinetic data obtained in the present experiment were limited but suggested different patterns for VPA and trans2-ene-VPA. It has previously been shown in mice that NaVPA and trans-2-ene-VPA given S.C.produce similar peak concentrations at about 0.5 hr posttreatment (Nau, '86). However, clearance rates of trans2-ene-VPA were higher than for VPA and involved two phases rather than one (tllz,, = 0.5 and tll,,p = 1.4 hr. for trans-2-eneVPA vs. t,,, = 1.3 hr for VPA). Furthermore, in mice, unbound trans-2-ene-VPA was cleared more rapidly than unbound VPA (Nau, '86). In the present experiment, VPA was higher at 1 hr and lower at 6 h r in the VPA group than was trans-2-ene-VPA in the trans-2-ene-VPA groups. Although the mouse and rat data are not directly comparable, one possibility is that trans-2-eneVPA may be more rapidly absorbed than VPA. However, because of its different protein binding and distribution patterns, trans-2-ene-VPA may be at a different point in its elimination phase at 6 hr than VPA. Another possibility is that the difference is related to the stage of development at which measurements were made. In mice they were taken on E8 and E9, whereas here they were taken in rats on E18. Since liver and kidney function and volume of distribution in the dam change late in gestation, the present findings may reflect such inf luences. Further pharmacokinetic data on trans-2-ene-VPA in rats during gestation will be required to clarify this point. ACKNOWLEDGMENTS
The authors thank Dr. Frank Abbott, Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, for providing advice and methods for the synthesis of trans-2-ene-valproic acid and to Dr. Leland C. Clark, Jr., for use of his laboratory in which to conduct the synthesis. The authors also thank M.S. Moran and C.A. Blanton for assistance in the in vivo, teratological, and statistical aspects of this re-
589
search. The research was supported in part by Public Health Service grant HD21806 from the NIH. LITERATURE CITED Ardinger, H.H., J.F. Atkin, R.D., Blackston, L.J. Elsas, S.K. Clarren, S. Livingstone, D.B. Flannery, J.M. Pellock, M.J. Harrod, E.J. Lammer, F. Majewski, A. Schnizel, H.V. Toriello, and J.W. Hanson (1988)Verification of the fetal valproate syndrome phenotype. Am. J . Med. Genet., 29:171-185. Bjerkedel, T., A. Czeizel, J . Goujard, B. Kallen, P. Mastroiacovo, N. Nevin, G. Oakley, and E. Robert (1982) Valproic acid and spina bifida. Lancet, 2:1096-1097. Brown, N.A., J. Kao, and S. Fabro (1980) Teratogenic potential of valproic acid. Lancet, 1 :660-661. DiLiberti, J.H., P.A. Farndon, N.R. Dennis, and C.J.R. Curry (1984) The fetal valproate syndrome. Am. J . Med. Genet., 19:473-481. Eluma, F.O., M.E. Sucheston, T.G. Hayes, and R.B. Paulson (1984) Teratogenic effects of dosage levels and time of administration of carbamazepine, sodium valproate, and diphenylhydantoin on craniofacial development in the CD-1 mouse fetus. J . Craniofac. Genet. Dev. Biol., 4:191-210. Jager-Roman, E., A. Deichl, S. Jakob, A.-M. Hartmann, S. Koch, D. Rating, R. Steldinger, H. Nau, and H. Helge (1986) Fetal growth, major malformations, and minor anomalies in infants born to women receiving valproic acid. J . Pediatr., 108:997-1004. Kao, J., N.A. Brown, B. Schmid, E.H. Goulding, and S. Fabro (1981)Teratogenicity of valproic acid: In vivo and in vitro investigations. Teratogen. Carcinogen. Mutagen., 1 :367-382. Lindhout, D., and H. Meinardi (1984) Spina bifida and in utero exposure to valproate. Lancet, 2:396. Loscher, W., and H. Nau (1985) Pharmacological evaluation of various metabolites and analogues of valproic acid-Anticonvulsant and toxic potencies in mice. Neuropharmacology, 24:427-435. Loscher, W., H. Nau, C . Marescaux, and M. Vergnes (1984) Comparative evaluation of anticonvulsant and toxic potencies of valproic acid and 2-en-valproic acid in different animal models of epilepsy. Eur. J . Pharmacol., 99:211-218. Mastroiacovo, P., R. Bertollini, S. Morandini, and G. Segni (1983) Maternal epilepsy, valproate exposure, and birth defects. Lancet, 2:1499. Nau, H. (1986) Transfer of valproic acid and its main active unsaturated metabolite to the gestational tissue: Correlation with neural tube defect formation in the mouse. Teratology, 33:21-27. Nau, H., and W. Loscher (1984) Valproic acid and metabolites: Pharmacological and toxicological studies. Epilepsia, 25[Suppl. l]:S14S22. Nau, H., W. Wittfoht, H. Schafer, C. Jakobs, D. Rating, and H. Helge (1981a) Valproic acid and several metabolites: Quantitative determination in serum, urine, breast milk and tissues by gas chromatography-mass spectrometry using selected ion monitoring. J . Chromatogr., 226;69-78. Nau, H., R. Zierer, H. Spielmann, D. Neubert, and Ch. Gansau (1981b) A new model for embryotoxicity testing: Teratogenicity and pharmacokinetics of valproic acid following constant-rate administration in the mouse using human therapeutic drug and metabolite concentrations. Life Sci., 29:2803-2814. Ong, L.L., J.L. Schardein, J.A. Petrere, R. Sakowski, H. Jordan, R.R. Humphrey, J.E. Fitzgerald, and F.A. de
590
C.V. VORHEES ET AL.
la Iglesia (1983) Teratogenesis of calcium valproate in rats. Fund. Appl. Toxicol., 3r121-126. Paulson, R.B., M.E. Sucheston, T.G. Hayes, and G.W. Paulson (1985) Teratogenic effects of valproate in the CD-1 mouse fetus. Arch. Neurol., 42:980-983. Paulson, R.B., M.E. Sucheston, T.G. Hayes, H.S. Weiss, L.A. Sachs, M. Oca, B. Kernan, and S. Weiss (1988) Effects of sodium valproate and oxygen on the CD-1 mouse fetus. J. Craniofac. Genet. Dev. Biol., 8:35-45. Robert, E., and P. Guibaud (1982) Maternal valproic acid and congenital neural tube defects. Lancet, 2: 937. Vorhees, C.V. (1987a) Teratogenicity and developmental toxicity of valproic acid in rats. Teratology, 35: 195-202.
Vorhees, C.V. (1987b) Behavioral teratogenicity of valproic acid: Selective effects on behavior after prenatal exposure to rats. Psychopharmacology, 92:173-179. Whittle, B.A. (1975) Pre-clinical teratological studies on sodium valproate (Epilim) and other anticonvulsants. In: Clinical and Pharmacological Aspects of Sodium Valproate (Epilim) in the Treatment of Epilepsy. N.J. Legg, ed. MCS Consultants Press, Tunbridge Wells, England, pp. 106-111. Wilson, J.G. (1965) Embryological considerations in teratology. In: Teratology: Principles and Techniques. J.G. Wilson and J. Warkany, eds. University of Chicago Press, Chicago, pp. 251-277.
