Timothy George, M.D.
Richard H Finnell, Ph.D.
Bermans Iskandar, M.D.
Graham Fieggen, M.D.
Rick Abbott, M.D.
Two areas in human neurulation and neurulation defects have been investigated recently. The first hypothesizes that heterogeneity of human neural tube defects could originate from differences in sites of neural tube fusion. It has been proposed that there must be multiple fusion sites (22) since only a single fusion site would predict that human neural tube defects would localize to the caudal and rostral neuropores, which is not the case. Although the model of multiple sites of fusion was attractive to explain the various location of defects along the neuraxis, experimental observations of human embryos using light microscopy and laser scanning electron microscopy strongly corroborate the hypothesis of a single site of fusion and a zipper-like process of neural tube closure (12).
The second area of research involves the investigation of the placode in patients with myelomeningocele. The current hypothesis is that the open neural placode has fairly normal internal architecture and connectivity and that most loss of spinal cord function is due to secondary injury. This proposition has been the basis and rationale of fetal surgery for myelomeningocele. Small studies of the myelomeningocele placode have demonstrated that even though there was secondary injury, abnormal architecture was present. No study has been able to examine the placode using physiological parameters to correlate function to structure (15).
With the global efforts in dietary fortification and maternal periconceptional administration of folic acid, it has been demonstrated that folic acid significantly reduces the recurrence risk for neural tube defects by 50–70%. Yet, the recurrence risk is not entirely eliminated by folate supplementation, suggesting that additional genetic or undetermined exogenous factors may be responsible for the development of neural tube defects.
The mechanism underlying the association between neural tube defects and folate has not been established. However, folate participates in two metabolic pathways that, if disrupted, could adversely affect the development of the embryo. One of these pathways is important for nucleic acid synthesis, and the other for a vast range of methylation reactions. Disruptions in folate metabolism can also result in elevated homocysteine levels, which are teratogenic to the neural tube in some species (23).
Current research is focused on methylation, and, recently, global methylation deficits have been suggested in patients with myelomeningocele. Methylation may play a role in altered gene expression that results in neural tube defects. Jirtle et al. have revealed that the impact of nongenetic factors may be more important in development than previously thought (24). The role of these epigenetic factors has yet to be elucidated.
Folate deficiency leads to upregulation of folate receptors, which are ubiquitous and mediate folate uptake at a physiological level. A recent study by Rothenberg et al. showed that some mothers with a pregnancy complicated by a fetal neural tube defect produced autoantibodies that bind to folate receptors on the placental membrane and therefore blocked the binding of folic acid (9). The authors further suggest that the periconceptional administration of folate would bypass the autoantibodies that mediate a placental folate receptor blockage. Indeed, folate has a high affinity for its receptor and might displace the autoantibody when administered at high doses.
Recent reports have revealed that the effect of folate supplementation has apparently reached a plateau. Since it appears that no more than 70% of neural tube defects can be prevented by folate administration, other factors must be important. The aqueous B vitamins, particularly B12, appear to be important as cofactors in many pathways with and without folate that cause neural tube defects. Copp and Greene demonstrated that the curly tail mouse, which is folate resistant, responds to inositol (20). Inositol is now being investigated in humans.
A number of different genes involved in folic acid metabolism, including those encoding folate receptors, 5,10-methylenetetrahydrofolate reductase (MTHFR), and cystathionine (beta)-synthase, have been investigated to determine possible roles in the development of neural tube defects. Recent studies have implicated homozygosity for the C677T thermolabile variant of the MTHFR gene as a risk factor for neural tube defects (25), and others have suggested that the effect may depend on the level of the lesion (11).
Meta-analysis found a pooled odds ratio for infants homozygous at C677T of 1.7 (95% CI, 1.4–2.2), with a pooled attributable fraction of 6% for homozygosity. While the paternal effect was nonsignificant, the odds ratios for maternal genotype, either homozygous or heterozygous for the T allele, were consistent with a trend for MTHFR involvement (odds ratio for homozygosity was 2.1 [95% CI, 1.5–2.9] and for heterozygosity was 1.2 [95% CI, 0.9–1.5]) (10). In addition, other mutations in the MTHFR gene have been investigated, including A1298C, and other genes, such as cystathionine β-synthase, that in combination with the C677T allele may increase the risk for neural tube defects (26). Reports have failed to demonstrate the association seen with the C677T MTHFR allele and neural tube defects (25).
Myriad exogenous causes for neural tube defects have been postulated and investigated (47). Factors that had been previously reported to have no significant association with neural tube defects, including maternal and paternal age, maternal periconceptional infections, number of prior "successful" pregnancies, recreational drug use, caffeine intake, smoking, exercise, and alcohol use, have been recently reinvestigated and may indeed play a role. Hyperthermia (fever and/or hot tub use) has been implicated to increase risk for neural tube defects, although most of these studies are limited due to recall bias and have yielded inconsistent results. Paternal exposure to Agent Orange in Vietnam veterans has been implicated, as has water chlorination by-products and maternal exposure to solvents through house-cleaning occupations. Exposure to fumonisins, a fungal metabolite commonly found in maize, has also been implicated, and in vivo and in vitro studies show an association with neural tube defects (27) and may act by disturbing folate metabolism.
Increased risk for neural tube defects has been strongly associated with maternal diabetes and maternal obesity (both associated with glucose metabolism), and maternal use of anticonvulsant medications. For example, antiepileptic drugs administered to pregnant mothers can increase the incidence of congenital malformations from 3% without drugs to 9% with drugs. These numbers can rise to 28% when three or more antiepileptic drugs were given to epileptic mothers. Valproic acid has the greatest teratogenicity when given to pregnant women, and its administration results in 1–2% incidence of spina bifida (17).