Etiology and pathogenesis
Anencephaly is etiologically heterogeneous (77; 96). At present, known risk factors account for less than one half of cases (02). A number of etiologic factors have been implicated, but most have not been confirmed or are found only in isolated cases. Implicated factors include maternal exposure to certain drugs (eg, aminopterin, salicylates, clomiphene, and gonadotropins), infections (ie, influenza), nitrite-cured meats, blighted potatoes, and hard water. Difficulties in maternal glucose control, ie, diabetes, hyperinsulinemia, obesity, and intake of simple sugars have been associated with increased risk of neural tube defect and may act by affecting gene expression in the developing embryo (165; 154; 145). Periconceptional maternal cannabis use is associated with the development of anencephaly (186; 150). Maternal hyperthermia, treatment of seizures in pregnant mothers with valproic acid, and the fungal product fumonisin are recognized causes of neural tube defects. One study has identified an association of acute conditions, for example, maternal fever, and more chronic ones, such as migraine headaches, with neural tube defects (118). In one study, the greatest attributable risk for anencephaly was Hispanic ethnicity (02).
A genetic basis for some cases of anencephaly seems likely, but it has been difficult to ascertain because the condition appears to be multigenic (71). A genetic contribution has been suggested by the association with other conditions having a recognized genetic basis, such as anal stenosis, anterior sacral meningocele, and Meckel syndrome (55). A genetic basis is also suggested by the identification of numerous mutations whose function causes neural tube defects in mice (44; 71) and the occurrence of familial cases with 10-fold increase in risk to siblings (29). Familial cases can be striking, with recurrence of three to six affected infants per mother. Demenais and colleagues suggested that familial cases involved monogenic inheritance with environmental influences (52), whereas Kalter and Warkany suggested that a genetic model would most likely involve multigenic inheritance (88; 89). Current thinking supports a multifactorial etiology (44; 71). Analysis of sex ratios show a predominance of female anencephalics, and many twins (usually discordant) have been reported. The increased incidence of anencephaly among twins does not appear to be related to mode of conception (ie, natural vs. artificial) (18). Interestingly, this female predominance is more pronounced in those exhibiting localized cranial lesions than those with schisis involving the upper vertebral column (48). A small percentage of individuals with anencephaly may have an abnormal karyotype. In a study, 2% of fetuses were affected, most often by trisomy 18 or trisomy 13 (159). Two cases of anencephaly associated with partial duplication of 2p have been reported, suggesting that genes on the short arm of chromosome 2 could be important to CNS development (178). A homozygous mutation in FOXN1, a member of the forkhead (or Fox) gene family, has been discovered in a human fetus with absent thymus, abnormal skin, anencephaly, and spina bifida (04). This finding suggests that the gene may be involved in neurulation in humans.
The presence of genetic factors is also supported by the observation that some 30% of defects are not prevented by folic acid supplements alone (22). It is possible that obesity (as quantified by basal metabolic index, or BMI) influences the body distribution of folate (182). The relationship of red blood cell folate levels and obesity appears to vary by BMI, but requires further study (107). Genetic factors presumably confer a risk for developing neural tube defects, but this avenue of research is in its earliest stages and does not yet offer substantive information (15). Important avenues for research include studies of genes controlling folate metabolism (51; 26), screens for human homologs to genes that cause anencephaly in laboratory animals (172), and maternal or sex-influenced genetic effects (49). Mutations in planar cell polarity genes (responsible for maintaining epithelial orientation during neural tube closure) have been identified in mice with neural tube defects and may be functional in humans as well (50; 87). The regulatory effects of bone morphogenetic protein and Sonic hedgehog have been implicated in neural tube bending and may play a role in pathogenesis (45), as may maladjustment of microRNAs in the mitogen-activated protein kinase signaling pathway (204).
The pathogenesis of anencephaly remains unclear. Two major schools exist; one holding that the primary mechanism is nonclosure of the rostral neural folds, and the other that cranial mesenchyme is defective primarily (109; 125). Most workers favor nonclosure of the neural tube as the primary mechanism, as evidenced by the widespread popularity of the term "neural tube defect" to designate anencephaly, encephalocele, and myelomeningocele. A third possibility is that a closed neural tube reopens, but supporting evidence is weakest in this area.
In human embryos, primary neurulation (closure of the neural folds) takes place between stage 8 and stage 12 (see Table 1). Anencephaly and other neural tube defects occur when development is abnormal, although the exact mechanism of closure remains unknown. The Hedgehog signaling pathway has been implicated in neural tube closure, for increased Sonic Hedgehog signaling is associated with the appearance of exencephaly (126). Embryonic metabolism appears to be important, as studies using mass spectrometry have identified abnormalities in glucose metabolism during neural tube closure (199).
Some investigators have postulated that the variations of anencephaly and other neural tube defects can be explained by multiple sites of closure of the neural tube (185; 70; 171). Others counter that “accessory loci” for closure exist in human embryos but are highly variable and do not follow a constant pattern (138). To a large extent, knowledge is fortuitous and depends on the study of affected embryos, which are rare and appear to die at very early ages (ie, 5 to 6.5 weeks) (127). Two anencephalic embryos, stages 13 and 22, are exceptions to this, having intact neural crest derivatives (139). The observation suggests that the process responsible for anencephaly began after the initiation of crest cell migration or that crest cells survived the pathogenetic event.
Table 1. Normal and Abnormal Events of Primary Neurulation
|
Stage |
Age* |
Pertinent event |
Results of maldevelopment |
|
8 |
18 days |
Neural folds appear |
Brain and spinal cord exposed and dysplastic (ie, early craniorachischisis) |
|
9 |
20 days |
Neural groove |
Early craniorachischisis |
|
10 |
22 days |
First fusion of neural folds |
Craniorachischisis |
|
11 |
24 days |
Closure of rostral neuropore optic vesicle |
Anencephaly |
|
12 |
26 days |
Closure of caudal neuropore; primary neurulation ends in this stage or stage 13 |
Myelomeningocele; myelocele; encephalocele |
* Postovulatory days are approximate (137; 125; 100).
An understanding of these stages is important in research and for counseling parents. Anencephaly can be said to have a "termination period" (ie, a time after which the anomaly cannot occur); therefore, exposures or other maternal events that occur after the critical periods for anencephaly can be ruled out as causative factors.
Exencephaly, the superior extrusion of a relatively large, bulbous portion of cerebral tissue beyond the calvaria, is a forerunner of anencephaly (198; 125). This has been demonstrated experimentally and by prenatal ultrasonography in fetuses prior to spontaneous degeneration of brain tissue. Holoprosencephaly is rarely associated with anencephaly (98; 167) and may involve aberrant morphogenesis of prechordal and paraxial developmental fields (168). The genetics of neural tube defects and midline craniofacial malformations is becoming understood, as illustrated by studies of the tuft mouse mutant (67).
Pathological findings in the central nervous system depend on the gestational age and extent of the lesion. Rudimentary cerebral hemispheres can be found in 50% of anencephalic newborns, and complete absence can be found in the remainder. Absence of the cerebellum and brainstem has also been reported (187). These infants usually manifest a connective tissue mass adherent to the dura mater. Their cerebral tissue is soft and infiltrated with blood, and scattered islands of neural cells and choroid plexus can often be identified. As a result of this appearance, the term "area cerebrovasculosa" is often applied to the anencephalic brain. This vascular finding is a reactive change to the exposure of neural tissue to amniotic fluid.
Examination of the brainstem reveals variability in the preservation of nuclei and absence of the pyramidal tract. The eyes often bulge dramatically. They may also appear normal but lack central connections to the brain, ending blindly posterior to the optic foramina. Other pathological changes of the eyes include absence or reduction in axis cylinders and colobomata of the optic disc. The spinal cord may show reduced white matter in some tracts, absence of Clarke column, and a lower termination of the conus medullaris.
Cranial bones are also highly abnormal (64; 125): calvarial bones are displaced, rudimentary, or absent; the zygomatic bones may be rotated abnormally; the sphenoid is hypoplastic in the cranial base; and petrous ridges are malpositioned. The maxillae, palatine bones, and vomer may be normal, although the mandible is large and prognathic. Temporal and occipital bones may also be malformed (14). An understanding of this bony anatomy has proven especially helpful in certain forensic situations (58).
Many other interesting associations or pathological findings occur in anencephaly. Not all are clearly related to the mechanism(s) responsible for the development of anencephaly. One study found that liveborn anencephalics had 12.3% additional malformations of other systems, but if there was an associated spina bifida or encephalocele, the prevalence was 87.7% (155). Anomalies of extraocular muscles have been observed in a large percentage of affected fetuses (143). Endocrine system anomalies are present in anencephalic infants with absence or hypoplasia of the pituitary gland, and the adrenal glands are hypoplastic in nearly all cases (114). Diaphragmatic malformations (eg, hernia or eventration), abdominal wall defects, and other thoracic cage abnormalities are common (97; 119). Frequently, the lungs and heart are hypoplastic. One study has demonstrated various degrees of intestinal aganglionosis in anencephaly (112). Anthropometric studies have shown an increased growth of the arms in anencephaly (128). A proximodistal gradient existed with the upper arm being increased by 24%, the forearm by 16%, and the hand by 2%. This is a variable finding, for Barr has observed normal arm length in anencephaly (Barr, personal communication 1997). Other anomalies of the limbs are found, but no one abnormality stands out, except perhaps talipes equinovarus. One detailed anatomic study identified several anomalies of skeletal muscles in a single fetus (03). A sternalis muscle is present in up to 50% of anencephalics but is unusual in the general population (1% to 4%).
