Plasma vitamin values and antiepileptic therapy: Case reports of pregnancy outcomes affected by a neural tube defect



BACKGROUND: Folic acid supplementation reduces the occurrence of neural tube defects (NTDs); however, it is not clear whether it protects against teratogenic effects of antiepileptic drugs. METHODS: We report the cases of four pregnant women receiving valproic acid therapy, who all had NTD-affected offspring, despite periconceptional 5 mg/day of folic acid supplementation (cases), and investigated homocysteine metabolism, linked with folate metabolism. Their plasma homocysteine, folates, and vitamin B6 and B12 results were compared with values of two other women, who were also receiving valproic acid and folic acid complement, but who had normal pregnancies (valproic acid controls), and values of 40 pregnant women who had normal pregnancies and were not receiving any therapy (controls without therapy). Because of the possible existence of a genetic susceptibility, polymorphisms in homocysteine metabolism were sought. RESULTS: Two cases showed a decreased phosphopyridoxal level, compared with levels in the controls not receiving therapy. The genotype TT (C677T) is an NTD genetic susceptibility, but it was observed in only one valproic acid control. Various polymorphisms were observed in the cases, but were also common in the controls. Several studies have reported that valproic acid therapy lowers vitamin B6 levels. Our case with the greatest decrease in plasma phosphopyridoxal, who was taking periconceptional folic acid plus pyridoxine therapy, had a normal second pregnancy outcome. CONCLUSIONS: In addition to folates, other vitamins, such as vitamin B6, may have played a role in NTDs in our patients taking an antiepileptic drug. Birth Defects Research (Part A) 2007. © 2006 Wiley-Liss, Inc.


During pregnancy, along with folate deficiency (MRC Vitamin Study Research Group, 1991; Czeizel and Dudas, 1992), antiepileptic therapy is another acquired risk factor for NTDs (Ubeda et al., 2002; Yerby et al., 2003). Valproic acid is used for treatment of seizures because it increases cerebral γ-aminobutyric acid, but its adverse effects include antifolate activity (Hendel et al., 1984), hepatotoxicity, and alterations in the homocysteine (Hcy) cycle (Ubeda et al., 2002). Despite this, it is not recommended that antiepileptic drug therapy be stopped during pregnancy. Because folate metabolism is linked with Hcy metabolism, we investigated the possibility of an interaction with valproic acid therapy.

We report the cases of four women receiving valproic acid therapy, who all had an NTD-affected offspring, despite periconceptional folic acid supplementation. Plasma Hcy, folates, and vitamins B6 and B12 were assayed. The results of these four cases were compared with results from two other pregnant women, who were also receiving valproic acid and folic acid supplementation, but who had normal pregnancies (valproic acid controls), and results from 40 healthy pregnant women who had normal pregnancies, who did not receive therapy or vitamin supplementation (controls without therapy). Various polymorphisms in Hcy metabolism were sought in all women and in additional controls not receiving therapy in an attempt to identify a genetic susceptibility.


The four women (cases) who had been receiving valproic acid therapy (500 mg, three times a day) were 24, 28, 30, and 33 years old. These four women had been taking a periconceptional supplement consisting of 5 mg/day of folic acid. Blood samples were obtained during the second trimester of pregnancy. As plasma values of vitamins and Hcy are normally moderately decreased in pregnancy due to modifications in metabolism, dilution, and increased fetal needs, the results of these four women were compared with those of two other pregnant women, ages 27 and 35 years old, at the second trimester of pregnancy, who were also taking valproic acid (treatment dose identical to those of the cases), and who had normal pregnancies (valproic acid controls). Results of the four cases were also compared with those of 40 healthy pregnant women who had normal pregnancies, were not receiving any therapy, and who did not take folic acid supplements. These 40 women (controls without therapy) were 18–40 years old (median, 29 years old) and blood samples from these controls were also obtained during the second trimester of pregnancy. Levels of plasma Hcy (assayed by FPIA; Abbott, Rungis, France), folate vitamins B12 (assayed by RIA on Bayer Centaur) and B6 (pyridoxal phosphate, assayed by HPLC, with fluorimetric detection; Chromsystems, Toulouse, France), and red blood cell folate (assayed by RIA) were determined. Various polymorphisms involved in Hcy metabolism, which is linked with folate metabolism, were sought because we hypothesized that they might increase genetic susceptibility to vitamin deficiency. These polymorphisms were sought in 58 control women who were not receiving therapy (the group of 40, plus 18 others), at the first or at the third trimester of pregnancy.

PCR enzymatically digested products were analyzed electrophoretically to search for C677T in the methylenetetrahydrofolate reductase (MTHFR) gene, A2756G in the methionine synthase (MTR) gene, A66G in the methionine synthase reductase (MTTR) gene, and C677G in the transcobalamine II (TCII) gene (Frosst et al., 1995; Zhu et al., 2003; Gueant-Rodriguez et al., 2003).

All subjects gave written consent.


Results are summarized in Tables 1 and 2. Cases and valproic acid controls were receiving folic acid therapy, so, as shown in Table 1, their homocysteine levels were normal. Two of the cases had decreased plasma pyridoxal phosphate values (the active form of vitamin B6), compared to the controls who did not receive therapy. A similar observation was made for one valproic acid control, who had a normal pregnancy.

Table 1. Plasma Homocysteine (Hcy), Plasma and Red Cell Folates, and Vitamin B12 and B6 Values in Our Population
 Hcy μmol/LPlasma folate ng/mLRed cell folate ng/mLB12 pg/MLPLP mmol/L
  • a

    Subjects receiving folic acid (FA) therapy, 5 mg/day.

  • Cases: pregnant women with NTD-affected offspring; controls: pregnant women with normal pregnancies.

  • + VA: receiving valproic acid therapy.

  • PLP: pyridoxal phosphate.

Case 1 + VA4.8a15a580a3618
Case 2 + VA4.4a69a1,100a44213
Case 3 + VA4.8a38a530a20214
Case 4 + VA3.711.3a412a4109
Control 1 + VA4.7a55a2,400a3149
Control 2 + VA4.3a37a1,516a35028
Controls (n = 40) (no FA supplement)3.8–8.12.9–15.2220–770149–84810–52

Various other abnormalities were observed in the four cases. Whereas plasma and erythrocyte folate values were elevated under folic acid supplementation, in three of the cases, levels were lower than in the two valproic acid controls. This might reflect a dysfunction in folate metabolism.

The results of polymorphism studies are summarized in Table 2. The genotype TT (C677T) is associated with a genetic susceptibility for NTDs, but it has only been described in certain populations. Paradoxically, the TT genotype was observed in only one of the valproic acid controls. Several heterozygous and one homozygous polymorphism were also observed in the four cases, but these same polymorphisms were also common in the controls without therapy.

Table 2. Genotypes for Homocysteine Metabolism in Pregnant Women
PolymorphismsC677TA 2756GA66GTC II
  1. C677T: in the methylenetetrahydrofolale reductase gene, CC is the wild genotypep.

  2. A2756G: in the methionine synthase gene, AA is the wild genotype.

  3. A66G: in the methionine synthase reductase gene, AA is the wild genotype.

  4. Codon 259: in the transcobalamin II gene, PRO-PRO is the wild genotype.

Controls (n = 58)TT 15%GG 4%GG 22%ARG.ARG 28%
TC 46%AG 45%AG 52%PRO.ARG 37%
CC 39%AA 51%AA 26%PRO.PRO 35%

When periconceptional pyridoxine therapy was added to folic acid supplementation, the case that had showed a dramatic decrease in plasma vitamin B6 had a normal second pregnancy and gave birth to a healthy child.


Periconceptional folic acid reduces the risk of NTDs but does not afford complete protection. To correct a possible folate deficiency, official government policies in some countries recommended that women planning a pregnancy should take folic acid supplementation daily. In practice, however, few women actually follow this advice. A number of countries, including France and other European nations (EUROCAT Register), currently recommend that women receiving antiepileptic therapy, who are at high risk of having an NTD-affected child, take a higher dose of folic acid. Valproic acid is known to increase the rate of major malformations such as NTDs in offspring exposed to the drug during pregnancy (Wyszynski et al., 2005). Valproic acid is suspected of having an antifolate activity, responsible for NTDs, but the exact mechanism is unknown. In animals, folinic acid has been shown to decrease the incidence of NTDs induced by valproic acid (Hansen et al., 1991), whereas folinic acid, associated with vitamins B12 and B6, reduced valproic acid–induced NTDs, although the protection was not complete (Elmazar et al., 1992). Among its other properties, valproic acid has been shown to be neuroprotective in ischemia (Ren et al., 2004), but it may also alter gene expression in the brain (Yildirim et al., 2003).

A decrease in plasma vitamin B12 is a known risk factor for NTDs (Kirke et al., 1993; Candito et al., 2004). Further studies are needed to determine whether periconceptional supplements with vitamin B12 might prove helpful. As suggested, the addition of vitamin B6 might be important for patients receiving valproic acid therapy. This might also be true for other antiepileptic drugs such as lamotrigine (Candito et al., 2006). Vitamin B6 deficiency decreases nicotinic acid synthesis from tryptophan and, in so doing, lowers the synthesis of NAD+ and NADP+, which are necessary for pyrimidic nucleotide synthesis. However, vitamin B6 supplements must be administered with caution. Whereas PLP activates the synthesis of GABA from glutamate (Jansonius et al., 1998), it also activates GABA transaminase, the enzyme that degrades GABA (Mueller et al., 2001), and an increase in cerebral GABA is the base of action of valproic acid.

A genetic susceptibility linked to vitamin deficiency might exist. Genotype TT (in the MTHFR gene) is a risk factor for NTDs in some, but not all, populations (Kirke et al., 2004; Johnson et al., 1999). Furthermore, it has been hypothesized that homozygosity confers a survival advantage in populations with adequate folic acid consumption (Rosenberg et al., 2002). Paradoxically, this genotype was observed in only one pregnant valproic control woman, and it was not detected in any of the four cases described here.

NTDs are multifactorial disorders. Other vitamin deficiencies, in addition to folate deficiency, may have played a role in NTDs in our patients receiving an antiepileptic drug. It remains to be determined whether it would be useful to assay B6 levels in pregnant women who are receiving valproic acid as antiepileptic therapy.