Neural Tube Defects and Spina Bifida

Literature and regulatory information updated: 6 October 2026.

Open neural tube defects (anencephaly, craniorachischisis, myelomeningocele and myelocele) and closed defects (encephalocele, iniencephaly, meningocele and closed spinal dysraphism), with emphasis on open spina bifida; multifactorial inheritance, genetic susceptibility factors, teratogens, folic acid prevention, prenatal diagnosis and fetal surgery

0. Initial Reference Table

1. Introduction

Neural tube defects constitute the group of congenital central nervous system malformations resulting from disturbed formation of the embryonic neural tube. The Oxford Desk Reference distinguishes open defects, resulting from failure of closure of part or all of the neural tube during primary neurulation (anencephaly, craniorachischisis, myelomeningocele, or open cystic spina bifida, and myelocele, or non-cystic open spina bifida), from closed defects, a heterogeneous group in which the early neural tube is abnormal but is not exposed to the external environment (encephalocele, iniencephaly, meningocele and spinal dysraphism) [10]. Orphanet defines open spinal dysraphism with myelomeningocele as a form of cystic spina bifida characterised by a non-neurulated spinal cord, the neural placode, on the surface of a cystic extension of dysplastic meninges, with nerve roots visible within the sac and a characteristic association with Chiari II malformation [3]. In anencephaly, failure of anterior neuropore closure produces exencephaly, in which brain tissue exposed to amniotic fluid degenerates until the brain is absent [2].

The modern history of these malformations is the history of three discoveries. The first was epidemiological and genetic: in the mid-1960s the multifactorial threshold model was proposed to explain familial aggregation of common malformations, including neural tube defects, which did not follow Mendelian laws [9]. The second was prenatal diagnosis, made possible in the early 1970s by the discovery of raised alpha-fetoprotein (AFP) concentrations in the amniotic fluid of pregnancies with anencephaly or myelomeningocele, followed by demonstration of elevated maternal serum AFP and parallel developments in ultrasound [8]. The third was prevention: Smithells and colleagues observed that women with an affected pregnancy had mild micronutrient deficiencies, including folate, and a multivitamin supplement containing folate reduced recurrence [8]; the randomised, double-blind Medical Research Council (MRC) trial, conducted at 33 centres in seven countries, demonstrated in 1991 a 72% protective effect of folic acid against recurrence (relative risk 0.28; 95% confidence interval [CI] 0.12–0.71) [25], and the Hungarian trial by Czeizel and Dudás showed in 1992 that a multivitamin containing 0.8 mg folic acid prevented first occurrence (six cases in the control group and none in the supplemented group) [33]. A fourth stage, in the twenty-first century, was fetal repair of myelomeningocele, whose benefit was established in 2011 by the randomised Management of Myelomeningocele Study (MOMS) [26].

The association between exposure to teratogens and these malformations also has its own historical milestones: Robert and Guibaud documented in 1982 the association between maternal valproic acid intake and myelomeningocele, and DiLiberti and Hanson described in 1984 the broader malformation pattern of fetal valproate syndrome [34]. This article addresses together neural tube defects as a general finding and spina bifida, because they share sources, mechanisms, prevention and counselling, while retaining dedicated subsections where the clinical features differ.

2. Molecular Genetics and Pathophysiology

2.1. Embryology: primary and secondary neurulation

In humans, primary neurulation begins on day 22 after fertilisation at the boundary between the future hindbrain and cervical spine, from which the neural tube closes bidirectionally in a “zipper-like” manner; a second closure site begins at the rostral end of the forebrain, the rostral neuropore closes on day 24 and the caudal neuropore on day 26, completing the spinal cord as far as the upper sacral level [8]. Oxford places primary neurulation between days 18 and 26 after fertilisation, with failure of cranial neural fold closure on day 18 in anencephaly and of the lower spinal folds on days 24–26 in open spina bifida [10]; Smith states that the neural tube is normally closed by 28 days [35]. Differences in days reflect the dating conventions of the sources and do not alter the sequence. Craniorachischisis, the earliest defect, results from failure to initiate closure and leaves almost the entire brain and spine open; anencephaly results from subsequent failure of the cranial folds; open spina bifida results from failure of the lower spinal folds [8, 10]. Evidence of multiple anterior closure sites explains variations in location, recurrence risk and aetiology [35].

Secondary neurulation forms the neural tube in the lower sacral and coccygeal regions from the caudal eminence (“tail bud”), whose mesenchymal core canalises into a hollow epithelial tube; as there is no closure stage, the defects are skin-covered (closed spinal dysraphism) and appear to result from failure of specification and spatial separation of neural and mesodermal tissues, explained by the existence of a bipotential neuromesodermal precursor; the frequent association with intradural lipoma remains unexplained [8, 10]. Smith describes the same sequence as the “occult spinal dysraphism sequence” or tethered cord malformation, involving structures of mesodermal and ectodermal origin (mesodermal hamartomas, sacral vertebral anomalies, filum terminale hyperplasia, distal spinal cord abnormalities) [36].

The French PNDS (National Diagnosis and Care Protocol) on lipoma of the filum terminale details the caudal chronology: secondary neurulation begins after closure of the posterior neuropore, between days 25 and 27 after ovulation, from the caudal cell mass, which gives rise to the neural tube, the vertebrae below S2 and, with the underlying endoderm, the lower part of the digestive and genitourinary tracts; from day 43 the conus medullaris begins to ascend, first by regressive differentiation (before day 54, when the caudal neural tube thins and forms the fibrous bands of the filum) and then by differential growth between the spinal cord and the spine, mainly between 8 and 25 weeks, possibly continuing into the first two months of life [28]. The conus medullaris normally lies above the L2–L3 disc; a conus below the body of L2 defines a low-lying cord [28]. Fatty infiltration of the filum during regressive differentiation makes it thick and inelastic, prevents ascent of the cord and may cause a low-lying cord and tethered cord syndrome; the same caudal origin explains the association with anorectal, urogenital and sacral malformations [28]. Other failures of secondary neurulation include terminal lipomas and terminal myelocystocele [28].

2.2. Pathophysiology of open spina bifida: the “two-hit” hypothesis

In myelomeningocele, the bifid neuroepithelium initially undergoes relatively normal neuronal differentiation, with development of motor and sensory function even below the lesion level; as gestation advances, the exposed cord becomes haemorrhagic, neurons die owing to amniotic fluid toxicity, axonal connections are interrupted and function is lost [8, 10]. Neurological disability is therefore understood as the result of two insults, failure of closure and intrauterine neurodegeneration, a concept that provided the rationale for fetal repair [8, 26]. Cerebrospinal fluid (CSF) leakage through the lesion is associated with Chiari II malformation, a downward protrusion of the medulla oblongata, vermis and fourth ventricle through the foramen magnum, in a normally sized cerebellum housed within a small posterior fossa [8, 10].

2.3. Multifactorial threshold model and genetic architecture

Most open neural tube defects are sporadic and isolated, common in the population and rarely present in multigenerational families, suggesting multifactorial causation with environmental (for example, maternal folate status) and genetic contributions [10]. The multifactorial threshold model, as presented by Smith, predicts familial aggregation, risk to first-degree relatives approximating the square root of the population risk, a steep fall in risk among second-degree relatives, increasing risk with the number of affected relatives, an increase with consanguinity, an increase with malformation severity, higher risk for relatives of the less frequently affected sex, and concordance in dizygotic twins similar to that in non-twin siblings [9]. The same book acknowledges that reanalyses suggest that susceptibility is determined by a small number of genes with multiplicative effects interacting with the environment, and that the model is probably not biologically exact, but that the empirical risks underpinning it remain the basis of counselling [9]. Oxford illustrates, in Falconer’s model, that the risk to children of an affected individual would be approximately 1/30 (3%), falling to 1/100 (1%) in second-degree relatives, values close to those observed in neural tube defects [37].

Copp and colleagues estimate heritability at 60–70%, based on sibling prevalence surveys conducted in the 1960s in South Wales, Glasgow and London, and indicate that fewer than 10% of cases are syndromic; concordance is higher in same-sex than opposite-sex twin pairs, there is a female excess in anencephaly but not in spina bifida, and differences between ethnic groups sometimes persist after migration [8]. More than 200 genes are required for neural tube closure in the mouse: exencephaly occurs in more than 150 mutants and open spina bifida in more than 40 [8].

The French PNDS on adult urology, drawing on the review by Au and colleagues, reports a heritability of 60%, a risk of 3–8% for siblings of an affected child, concordance of 7.7% in monozygotic and 4.4% in dizygotic twins, and a North American survey in which more than 16% of children with spina bifida had a family history of neural tube defect; the authors read these data as an indication that the genetic component weighs relatively more in populations already supplemented with folic acid [17]. The 3–8% range for siblings is wider than the 2–5% of Copp and colleagues because it aggregates studies from different periods and baseline prevalences; section 8.2 presents risks by degree of relationship and by baseline prevalence [8, 10].

2.4. Planar cell polarity pathway

The planar cell polarity (PCP) pathway is a non-canonical Wnt signalling cascade required for initiation of neural tube closure; in mouse PCP mutants, including Vangl2 (the Looptail model), convergent extension fails, the body axis remains short and broad and the neural folds are too far apart to initiate closure, resulting in craniorachischisis [6, 8]. In humans, rare missense variants in PCP genes, including CELSR1 and VANGL2, have been identified in cases but not controls [10]. Kibar and colleagues described in 2007 three VANGL1 variants in familial (V239I and R274Q) and sporadic (M328T) cases, one of them (V239I) arising spontaneously in a familial setting (published protein designations, = NM_138959.3:c.715G>A p.(Val239Ile), NM_138959.3:c.821G>A p.(Arg274Gln) and NM_138959.3:c.983T>C p.(Met328Thr) on the MANE Select transcript, confirmed in VariantValidator and ClinVar, where they are classified, respectively, as of uncertain significance, likely pathogenic and risk factor) [38, 39]; V239I abolished the interaction of VANGL1 protein with Dishevelled proteins 1, 2 and 3, and the authors interpreted the gene as a risk factor [40]. In a multiethnic cohort of 673 patients with familial and sporadic neural tube defects, six rare missense variants of VANGL2 were found in seven patients, at a frequency of approximately 2.5% in closed spinal defects (5 in 202) and 0.4% in open spina bifida (2 in 453) [41]. In the review by Juriloff and Harris, rare putatively deleterious variants of VANGL2, SCRIB, DACT1 and CELSR1 contributed cumulatively to more than 20% of craniorachischisis cases, but only approximately 6% of myelomeningoceles (combining CELSR1, FUZ, FZD6, PRICKLE1, VANGL1 and VANGL2), with no contributory variants in PRICKLE1 or PTK7 in craniorachischisis [11]. Variants from these studies are reproduced here in the abbreviated protein nomenclature used by the authors, without conversion to current reference transcripts.

GeneGuidance curation summarises the current status of these associations: VANGL1 and VANGL2 are susceptibility genes rather than fully penetrant Mendelian genes; OMIM records them under the multifactorial entry 182940 within braces (susceptibility); ClinGen has no gene–disease validity curation; GenCC submissions for susceptibility to neural tube defects range from limited to moderate; and penetrance is incomplete, with variants frequently inherited from unaffected parents [6]. A heterozygous missense variant in VANGL1 or VANGL2 in a neural tube defect case should therefore be described in terms of susceptibility rather than as a diagnostic explanation, and does not predict a malformation in a carrier or a pregnancy [6]. For CELSR1, the strongest association is with autosomal dominant primary lymphoedema; rare variants enriched in craniorachischisis, spina bifida and caudal agenesis cohorts are risk variants [6]. DACT1 causes Townes-Brocks syndrome type 2, and five rare missense variants were described in 167 Chinese fetuses with neural tube defects, presented by the authors as a risk factor [6]. FZD6 causes isolated autosomal recessive nail dysplasia and PRICKLE1 progressive myoclonic epilepsy; their presence in neural tube defect gene lists derives from aggregation of rare variants rather than Mendelian neural tube diseases [6, 11].

2.5. Folate and one-carbon metabolism: MTHFR, MTHFD1, GLDC and AMT

Methylenetetrahydrofolate reductase (MTHFR) generates 5-methyltetrahydrofolate, a methyl donor for remethylation of homocysteine to methionine [8, 42]. The common MTHFR variant NM_005957.5:c.665C>T, p.(Ala222Val), historically designated 677C>T, encodes a thermolabile enzyme; the variant NM_005957.5:c.1286A>C, p.(Glu429Ala), corresponds to the historical designation 1298A>C; both descriptions were confirmed on the MANE Select transcript using VariantValidator (GRCh38: NC_000001.11:g.11796321G>A and NC_000001.11:g.11794419T>G, respectively) [38, 42]. Approximately 25% of Hispanics and 10–15% of white North Americans are homozygous for the thermolabile variant, which does not cause homocystinuria due to MTHFR deficiency [42]. In ClinVar, the aggregate record for c.665C>T (VCV000003520) has the germline classification “drug response”, reviewed by an expert panel, and that for c.1286A>C (VCV000003521) has conflicting classifications; neither constitutes a pathogenic variant for neural tube defects [39].

Oxford attributes an approximately 1.8-fold higher risk of neural tube defects to the C677T and A1298C polymorphisms (historical numbering, = NM_005957.5:c.665C>T p.(Ala222Val) and NM_005957.5:c.1286A>C p.(Glu429Ala)), detected only in non-Hispanic populations [10]. The meta-analysis by Amorim and colleagues (37 populations, 3530 cases and 6296 controls) found an association of the TT genotype with neural tube defects overall (odds ratio [OR] 1.41; 95% CI 1.24–1.59) and in populations of non-Latin European ancestry (OR 1.62; 1.38–1.90), but not in those of Latin European ancestry (OR 1.16; 0.95–1.43), and found that, among non-Latin populations, the association was established only in Irish populations [43]. Copp and colleagues add that the maternal or fetal 677TT genotype is a risk factor mainly when folate status is low, and that Mthfr knockout in the mouse does not produce neural tube defects, calling the specificity of the association into question [8]. The 2015 report from the Centers for Disease Control and Prevention (CDC) links the T allele to lower plasma and erythrocyte folate concentrations in Hispanics [22]. The clinical interpretation of these data is that MTHFR c.665C>T is a small-effect risk modifier dependent on population and folate status, with no utility as a diagnostic or predictive test (see section 5) [6, 44]. The USPSTF considers that further research is needed into how variants such as those in MTHFR affect supplementation strategies [19].

The glycine cleavage system genes, AMT and GLDC, mitochondrial enzymes of one-carbon metabolism, showed missense variants in neural tube defect cases but not controls [10], and loss of function of the mouse orthologues produces these defects; as mitochondrial activity supplies approximately 70% of the cell’s one-carbon units in the form of formate, this pathway may contain relevant risk factors [8]. The GeneReviews on non-ketotic hyperglycinaemia emphasises, however, that neural tube defects have not been observed in heterozygous parents of children with that disease [45]. The common MTHFD1 variant NM_005956.4:c.1958G>A p.(Arg653Gln) (MANE Select transcript, confirmed in VariantValidator) has been associated at population level with folate-sensitive neural tube defects (OMIM susceptibility entry 601634), a modest association with variable replication, distinct from MTHFD1 deficiency, a very rare autosomal recessive disease of combined immunodeficiency and megaloblastic anaemia [6, 38]. CD320 (transcobalamin receptor) polymorphisms were associated with increased risk in an Irish cohort [46]. Saudubray reports biallelic variants of SLC25A32, the mitochondrial FAD transporter, in an anencephalic fetus, and neural tube defects in mice with Slc25a32 inactivation, which may reflect a second function of the transporter as a folate carrier; in patients with myopathy due to deficiency of this transporter, riboflavin improved the clinical and biochemical picture [47].

2.6. De novo variants and rare monogenic genes

Exome sequencing of 43 trios with sporadic myelomeningocele or anencephaly identified 42 coding de novo variants in 25 cases, six with loss of function, a rate higher than in control cohorts; two de novo truncating variants in SHROOM3, a de novo nonsense variant in PAX3, and potentially causal missense variants in GRHL3 and PTPRS were found [48]. Pathogenic heterozygous IRF6 variants were identified in fewer than 1% of individuals with neural tube defects, and two individuals with spina bifida were carriers; these defects are clinically indistinguishable from those of other aetiologies [49]. In Waardenburg syndrome type 1 (PAX3), GeneReviews recommends folic acid supplementation for women at risk of having an affected child, given the possible increase in neural tube defect risk, although no human study has defined the dose [50]. In HES7-related spondylocostal dysostosis, two of seven individuals from three families had neural tube defects [51]. Neural tube defects also occur in a small proportion of individuals with holoprosencephaly and in serine synthesis defects [52, 53].

2.7. Chromosomal and syndromic aetiology

Neural tube defects form part of several chromosomal abnormalities [10]. Chen’s review lists associated complete aneuploidies, including trisomies 18, 13, 9, 2, 21, 7, 8, 14, 15 and 16, triploidy, tetraploidy, monosomy X and mosaic trisomies 5, 11 and 20, and recommends cytogenetic investigation of a perinatal neural tube defect [54]. The Brazilian book on genetic syndromes lists Meckel–Gruber syndrome, Waardenburg syndrome and trisomies 13 and 18 [7]. In trisomy 18, neural tube defects are among the fetal anomaly scan findings [55]. Meckel syndrome, autosomal recessive and lethal, combines occipital encephalocele, enlarged kidneys with multicystic dysplasia, hepatic fibrosis and postaxial polydactyly; the genes involved (MKS1, TMEM216, TMEM67, RPGRIP1L, CEP290) participate in ciliary structure or function [10, 56].

2.8. Teratogenic, nutritional and maternal factors

Valproate increases the relative risk of neural tube defects approximately 10-fold when taken in the first trimester; it was considered a folate antagonist, but it is a potent histone deacetylase inhibitor, a possible mechanism of neurulation failure [8, 10]. Valproate-induced defects mainly affect the lower lumbar and sacral regions [10]. In its chapter on fetal anticonvulsant syndrome, Oxford estimates a risk of approximately 5% with valproate, approximately 1% with carbamazepine and no increase with other medicines, with a dose–response relationship above 900 mg/day [57]; the US Depakote® label estimates spina bifida risk following intrauterine valproate exposure at approximately 1–2% (100 to 200 per 10,000 births), compared with 0.06–0.07% in the general population [58]. The difference between these estimates reflects different denominators and outcomes (all neural tube defects in clinical series versus spina bifida in registry data), and both should be presented with their respective source. Exposures to known folate antagonists, such as sulfasalazine, triamterene or trimethoprim, should also be identified [10]. In France, the Agence nationale de sécurité du médicament et des produits de santé (ANSM) contraindicates sodium valproate in girls and women of childbearing potential, unless alternative treatment is impossible, with signature of an informed consent form and effective contraception, and recommends regular neurological follow-up of patients of childbearing potential; carbamazepine is identified as a risk factor of lesser magnitude [17].

Maternal pregestational diabetes is associated with an approximately 10-fold increase in relative risk according to the Oxford neural tube defects chapter [10]; the diabetic embryopathy chapter in the same book, using data from the population-based cohort of Bell and colleagues (401,149 singleton pregnancies, 1677 in diabetic women), reports a relative risk of 5 for neural tube defects, 170 for caudal dysplasia sequence and 57 for laterality disorders [12, 59]. In that cohort, the rate of major non-chromosomal congenital anomalies was 71.6 per 1000 pregnancies (relative risk 3.8; 95% CI 3.2–4.5), with a 30% increase in risk for each 1% of glycated haemoglobin (HbA1c) above 6.3% (45 mmol/mol) [59]. The risk is confined to women who are diabetic at conception and is not observed in gestational diabetes [12]. The mechanism of diabetic embryopathy is unclear; altered inositol, prostaglandin and reactive oxygen species metabolism and disturbed Pax3 expression have been proposed [8, 12].

Severe maternal pre-pregnancy obesity is associated with an approximately 3-fold higher risk [10]: in the meta-analysis by Rasmussen and colleagues (12 studies), the unadjusted ORs were 1.22 (0.99–1.49), 1.70 (1.34–2.15) and 3.11 (1.75–5.46) for overweight, obesity and severe obesity [13]. The risk is stronger for spina bifida than anencephaly, ranges between 1.5- and 3-fold in different studies, increases with body mass indices (BMI) above 35 and may not be reduced by folic acid [8]. Severe maternal hyperthermia in the first third or half of gestation is teratogenic in animal models; in retrospective human studies with more than 170 cases, approximately 10% of mothers had a history of hyperthermia during the week of neural tube closure (days 21–28), compared with none in controls, and in a prospective study 2 of 34 women (5.9%) with fever ≥38.9 °C for at least 24 hours during the critical period had fetuses with anencephaly [14]. Other factors mentioned include fumonisin, a fungal toxin responsible for an outbreak of neural tube defects in South Texas linked to contaminated tortilla flour, zinc deficiency and social class [7–9]. Oxford found no increase in neural tube defects with oral contraceptives or hormonal pregnancy tests, and lists mycophenolate mofetil among teratogens associated with myelomeningocele [60]. The French PNDS adds vitamin B12 deficiency, a cofactor in the conversion of homocysteine to methionine and in DNA synthesis, as a possible folate-independent risk factor, based on Canadian experience, and mentions the contribution of epigenetic factors [17].

3. Clinical Spectrum and Manifestations by System

3.1. Distribution of types and sex ratio

According to Oxford, approximately 40% of open neural tube defects are anencephaly, 50% lumbosacral spina bifida and 10% other open defects, including high spina bifida and craniorachischisis; females are more frequently affected by anencephaly (ratio 3:1), with no marked distortion in open spina bifida [10]. Smith gives male:female ratios of 1:3 for anencephaly and 1:1.5 for myelomeningocele [9]. The Brazilian book indicates that anencephaly accounts for approximately one-third of cases and that defects are more common in females and children of mothers aged 20–24 years [7]. The proportion of each type depends on the source and inclusion of pregnancy terminations, which is why the attributed values are presented.

3.2. Central nervous system in open spina bifida

Hydrocephalus occurs in more than 90% of open spina bifida cases and Chiari II malformation in more than 70% of myelomeningoceles, potentially causing lower cranial nerve palsies and central apnoea and producing increasing symptoms with age [10]. Smith considers hydrocephalus and the other manifestations of Chiari II malformation, present in virtually all cases, the greatest determinants of independence and survival [35]. Copp and colleagues describe the spectrum of Chiari II hindbrain anomalies in approximately 90% of cases, midbrain distortion with “tectal beaking” in approximately 65%, an elongated and kinked medulla at the cervicomedullary junction in approximately 70%, and corpus callosum hypogenesis in one-third to half of children; hydrocephalus results mainly from obstruction of CSF flow at the fourth ventricle [8]. Clinically, symptomatic hindbrain herniation may affect up to 30% of cases and manifests as apnoea, swallowing difficulties and stridor in the newborn, or headache, tetraparesis, scoliosis and impaired balance and coordination in the older child [8].

The cognitive profile of myelomeningocele combines strengths in associative processing (procedural learning, word reading, vocabulary, language form, persistence and social activation) with weaknesses in integrative processing (motor adaptability, language comprehension and pragmatics); intellectual disability affects approximately 20–25% of people, frequently following complications of hydrocephalus, and cognitive and motor outcomes are related to the spinal lesion level [8]. Even with a normal intelligence quotient, executive and oral language disorders are frequent and hinder autonomy in self-care (taking medicines, self-catheterisation, booking follow-up appointments), which requires treatment goals to be explained in simple words and repeatedly, and justifies the support of occupational therapists and neuropsychologists in specialised centres [17]. The Brazilian book reports intellectual disability in most neural tube defect cases, partly attributable to central nervous system infections [7]; this divergence results, at least in part, from the population described (all neural tube defects, including encephaloceles, in a clinical series, versus myelomeningoceles in an international review).

3.3. Neuromuscular, orthopaedic, urinary and bowel systems

Paralysis and absent sensation below the lesion level are typical sequelae; neurological function develops in the fetus even below the lesion and is lost by birth through neurodegeneration [10]. Neurological, skeletal and renal deficits are frequent in survivors [10]: orthopaedic deformities (clubfoot, kyphosis, scoliosis), neurogenic bladder with overflow urinary retention and vesicoureteric reflux, which may lead to recurrent urinary tract infections and deterioration of renal function, and bowel dysfunction in the older child [8, 35]. Vertebral anomalies and hydronephrosis frequently arise as secondary pathology in “isolated” defects [10]. Latex sensitisation is a problem to prevent [7, 8]. Open spina bifida may involve limb deformity or paralysis, bladder, bowel and sexual dysfunction, learning difficulties and risk of psychosocial maladjustment [7].

Urinary tract in adults. According to the French PNDS (National Diagnosis and Care Protocol) on adult urology, more than 90% of people with myelomeningocele have urinary disorders, and 40–60% of people with spinal dysraphism, open or closed, have urinary incontinence, with similar prevalence in the two forms in a prospective cross-sectional study of 318 patients [17]. The usual complaints are incontinence of variable type, frequency and volume, overactive bladder syndrome and recurrent urinary tract infections [17]. Risk to the upper urinary tract arises from high bladder pressures and recurrent pyelonephritis, and is greater when detrusor overactivity or a poorly compliant bladder is associated with detrusor–sphincter dyssynergia or detrusor pressures above 40 cmH2O [17]. In a French cohort, urological diseases (end-stage kidney disease, infections) were the leading cause of death in adults with spina bifida and accounted for 17% of deaths; urinary incontinence and voiding disorders were the primary reason for seeking care in 29% of adults who contacted the national reference centre between 2007 and 2015 [17]. As the neurological level is rarely above T6, spasticity and autonomic dysreflexia are rare in this population; loss of bladder sensation, present in most patients, makes lower urinary tract infections difficult to recognise, with atypical symptoms (malaise, altered voiding, new incontinence), and pyelonephritis may occur without fever or loin pain, while mechanical low back pain is frequent [17].

Neurogenic bowel. In a French reference centre, digestive disorders were the main complaint of 16.3% of patients with spina bifida, second only to urological complaints (27.8%); among 228 adults assessed by a gastroenterologist, 35.5% had a history of faecal impaction, 35.0% abdominal pain and 73.7% dyschezia, with evacuation manoeuvres needed in 53.5% (digital rectal evacuation in 90% of these), and 60% had severe anal incontinence, 85% severe constipation and 42% severe neurogenic bowel dysfunction [27]. These figures come from a reference-centre cohort and do not estimate prevalence in the population with spina bifida. Severe constipation is associated with anal incontinence, and bowel disorders do not follow the neurological level [27]. In a French multicentre study of 460 patients, anal and urinary incontinence scores did not correlate with quality of life, which did not improve after surgical treatment of one of the incontinences; in a Canadian study of 31 patients operated on for urinary continence with a continent stoma (with enterocystoplasty in 97%, bladder neck suspension in 84% and continent caecostomy in 77%), incontinence scores improved without change in quality of life at 6 months [27]. The authors conclude that correcting an isolated impairment is not enough to improve quality of life, which depends on physiological, psychological and social components [27].

Sexual function and fertility. Erectile dysfunction is frequent in men with spinal dysraphism, who often have sperm abnormalities and require further investigation; women have less difficulty conceiving, but pregnancy requires specialist follow-up (sections 7.13 and 8.6) [17].

Motor level, walking and musculoskeletal complications. The argumentaire (supporting evidence document) of the French perinatal PNDS takes up the International Myelodysplasia Study Group classification (from level T10, with no lower-limb movement, to level S2–S3, with normal or near-normal strength) and the Shurtleff classification for lumbosacral defects, which links level to walking (table 3.3a) [18]. The PNDS on physical medicine and rehabilitation adds data on evolution in adults [32].

Table 3.3a. Lesion level, absent muscles and expected ambulation, according to the French PNDS.

The same documents indicate that foot malposition (equinovarus foot, clubfoot) affects approximately 75% of children with myelomeningocele or myelocele, that approximately 75% have urinary disorders (probably almost all, given the imprecision of the series), that renal failure affects up to 15% of patients with lumbar myelomeningocele after 20 years of evolution, that 78% of 527 Nordic children aged 4–18 years had bowel disorders and that 30% of children with myelomeningocele have moderate Chiari II symptoms, such as feeding difficulties or gastro-oesophageal reflux [18]. A wheelchair is used by 36–57% of people with dysraphism; in adults, upper-limb overload causes osteoarthritis and rotator cuff injury in 31–71% (shoulders) and osteoarthritis or carpal tunnel syndrome in 49–73% (wrists); scoliosis affects 47.7%, osteoporosis approximately 50% and excess weight 35–37%; hip and knee osteoarthritis are frequent, spasticity rare, and neurogenic osteoarthropathy has an incidence of approximately 1% in the feet [32]. Pressure ulcers have a prevalence above 30% and an incidence of 42–52%, favoured by motor, sensory and orthopaedic deficits and by orthopaedic devices and footwear themselves [32]. These figures come from centre series and reviews cited by the PNDS, with different denominators and periods, and do not estimate population prevalences.

3.4. Closed spinal dysraphism and tethered cord

Spinal dysraphism frequently presents as a hairy patch, pigmented or abnormal skin, subcutaneous mass, dimple or sinus over the lumbosacral spine, with failure of fusion of two or more dorsal vertebral arches, usually at L5/S1 [10]. Tethering of the lower end of the cord may cause lower-limb weakness and particularly functional bladder disturbances, and is frequently associated with anorectal anomalies, including anal stenosis or atresia [10]. Smith describes cutaneous markers (hair tufts, skin tags, dimples, lipomas and aplasia cutis congenita) at the presumed junction between the caudal cell mass and posterior neuropore, around L2–L3, and emphasises that a dimple at the tip of the coccyx is very common and usually not associated with tethered cord [36]. In its chapter on minor congenital anomalies, Oxford reports that small blind-ending sacral dimples occur in 2% of newborns and that, in a study of 75 babies with a sacral dimple or sinus, none had an abnormality on spinal ultrasound, so these lesions, when blind-ending and not accompanied by pigmentation, a hairy patch or a subcutaneous lumbosacral mass, do not indicate a high risk of occult dysraphism [61]; by contrast, hypertrichosis limited to the spinal region requires a search for diastematomyelia, myelomeningocele, dermal cyst or sinus, vertebral anomalies or subdural or extradural lipoma [62]. Presentation ranges from childhood to adulthood, and prevalence is unknown because of the variability of the pathology [10]. The French PNDS (National Diagnosis and Care Protocol) classifies closed dysraphism by the presence or absence of a subcutaneous mass and identifies as the most frequent forms myelolipocystocele (commonly termed myelolipoma), lipoma of the filum, meningocele, diastematomyelia and caudal regression syndrome [17]. In these forms, the neurological deficit is local or locoregional, may be absent at birth and become apparent later, there is no surgical emergency and diagnosis is often late; functional prognosis is generally better than in open forms (better walking, greater autonomy, no associated cognitive disorders) [17, 27]. A defect of a single posterior arch, usually at L5 or S1, is described in up to 17% of the adult population, a figure close to the 15% reported by Oxford, and in most cases is not accompanied by a neurological deficit [10, 17]. The 2021 perinatal PNDS distinguishes, among closed dysraphisms, those with a subcutaneous mass (meningocele or subcutaneous lipoma: LDM, lipomyelomeningocele, terminal myelocystocele) and those with only a cutaneous stigma (LDM without a mass, dermal sinus, intradural lipoma, lipoma of the filum terminale, caudal regression syndrome), describes Myelo-LDM as a borderline form and advises against the term “spina bifida”, which designates only the opening of the posterior arch and may exist with a normal spinal cord [18]. The 2015 PNDS on physical medicine and rehabilitation follows the Tortori-Donati classification and also includes “split notochord syndrome” (enteric fistulas, neurenteric cysts, diastematomyelia, dermal sinus and intestinal duplication) and the sacrococcygeal agenesis of the OEIS and VACTERL associations and of the Currarino triad [32]. In lipomyelomeningocele, two-thirds of lipomas are asymptomatic, but this form has the worst prognosis among conus lipomas, with sphincter disorders predominating over sensorimotor disorders [18].

Lipoma of the filum terminale. This is the simplest closed dysraphism: the fat is confined to the filum, without spinal cord dysplasia or mass effect, unlike conus lipomas [28]. Prevalence is 4–6% in autopsy studies and 1.5–5% on randomly selected lumbar MRI scans; rare isolated familial cases have been described and no genetic basis has been identified [28]. Most patients are asymptomatic when the lipoma is discovered; symptoms are those of tethered cord syndrome (sphincter disorders in 43–59%, neuro-orthopaedic in 35–50% and sensory in 16–32% of associated symptoms), due to axial traction on the spinal cord with altered microcirculation and oxidative metabolism [28]. A low-lying cord (below the inferior endplate of L2) is suspected of being tethered, but a tethered cord may be in a normal position [28]. In infants, examination looks for incomplete distal movements, inequality of length or asymmetry of the lower limbs and feet, abnormal reflexes and amyotrophy; in older children, walking difficulty, instability, painless ulcers due to non-radicular sensory deficit, scoliosis or clubfoot and, more rarely, low back or neuropathic lower-limb pain; in adolescents, pain predominates [28]. Sphincter disorders include delayed or lost continence, recurrent urinary tract infections, urgency and daytime or night-time leakage, and constipation with overflow faecal soiling [28]. Patients with isolated lipoma of the filum, like those with other closed dysraphisms, have no associated cognitive or neurodevelopmental disorders [28].

3.5. Psychosocial and quality-of-life manifestations

Children and adolescents with spina bifida have lower health-related quality of life than unaffected peers and people with other chronic diseases; the strongest predictors are shunted hydrocephalus and lack of mobility, more than lesion level or continence, alongside social class, pain and parental stress; more depressive symptoms, poorer self-concept, social immaturity and greater dependence on adults are observed, and between 10% and 15% of families have clinically significant family dysfunction [8].

4. Clinical Recognition and Warning Signs

Recognition of open neural tube defects is now mainly prenatal, and the Brazilian book states that diagnosis should be prenatal [7]. Cranial ultrasound warning signs (lemon sign, banana sign, biparietal diameter small for gestational age and ventriculomegaly) require detailed examination of the spine and, in many units, tertiary referral [8]. In fetal ventriculomegaly, the spine should be specifically examined for neural tube defects [63]. Raised maternal serum AFP is another warning sign (section 5).

After birth, warning signs include an open dorsal lesion, flaccid paralysis with lower-limb areflexia, clubfoot, rapid head circumference growth and signs of symptomatic hindbrain herniation, such as apnoea, stridor and dysphagia [8, 35]. In closed defects, lumbosacral cutaneous markers (hairy patch, flat haemangioma, lipoma, deep dimple, skin tag, aplasia cutis) should prompt imaging assessment; Smith recommends ultrasound to document normal cord mobility with respiration and magnetic resonance imaging (MRI) in uncertain cases, and emphasises that waiting for urological or neuromuscular complications may render neurological damage irreversible [36]. In adults and parents of an affected child, asymmetric lower motor neuron weakness with atrophy, reduced reflexes or spasticity with hyperreflexia, or altered bladder control, suggests spinal dysraphism [10]. In children, adolescents or adults with unexplained bladder–sphincter or anorectal disorders, the French PNDS (National Diagnosis and Care Protocol) recommends looking for signs of closed dysraphism, such as pes cavus and claw toes (a sign of distal neurological involvement) and subtle cutaneous changes in the sacral region (subcutaneous lipoma, angioma, hair tuft), and referring to a specialised centre if present [17].

For lipoma of the filum terminale, the 2021 PNDS lists the midline lumbosacral cutaneous stigmata that should suggest the diagnosis: atypical coccygeal dimple (wider than 5 mm and more than 2.5 cm from the anus), deviation or bifurcation of the intergluteal cleft, swelling (subcutaneous lipoma), abnormal hair (the “faun tail” is typical of diastematomyelia), flat or tuberous angioma, areas of hyper- or hypopigmentation, aplasia cutis and fleshy appendages [28]. In closed dysraphism in general, these changes are present in 50–92% of cases; in lipoma of the filum, as the fat forms beneath an ectoderm that may be intact, stigmata may be absent, and the probability of diagnosis increases with the number of stigmata [28]. Cutaneous stigmata are present in approximately 3% of children, but only 3–8% of these have a closed dysraphism [28]. Syndromes with anorectal malformation and sacral agenesis (VACTERL association, Currarino syndrome) are accompanied by spinal malformations in 17–50% of cases, of which 30–82.4% are lipomas of the filum; in these children, diagnosis of tethered cord is difficult, because continence and bowel transit may already be altered by the malformation or its repair, and the neuro-orthopaedic examination may be modified by limb anomalies [28].

In a person with known spina bifida, any new urinary symptom or worsening of existing symptoms should suggest neurological deterioration and lead to MRI of the whole spinal axis looking for syringomyelia, particularly if digestive, motor or neurological symptoms coexist [17]. Isolated fever requires exclusion of other sources of infection before treating the bacteria identified on urine culture, and macroscopic haematuria requires full assessment (section 11.5) [17].

In the genetics clinic, signs requiring reconsideration of an isolated defect are additional anomalies (occurring in more than 20% of cases), with particular attention to congenital heart disease, diaphragmatic aplasia and oesophageal atresia, a history of anticonvulsant exposure, pregestational diabetes, hyperthermia or folate antagonists, consanguinity, recurrence despite high-dose folate supplementation and findings such as polydactyly, cystic kidneys or severe intrauterine growth restriction with syndactyly (Meckel syndrome, triploidy) [10].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Prenatal ultrasound

Ultrasound allows detection of anencephaly from 11–12 weeks; in high-risk pregnancies, Oxford recommends serial ultrasound from 16 weeks for spina bifida, in which large defects may be visible from 13 weeks [10]. The prenatal imaging table in the same book gives 11–12 weeks for anencephaly and “from 11 weeks” for spina bifida [64]. In a European study of 18 registries in 11 countries (670,766 deliveries), ultrasound sensitivity in an unselected population was 98% for isolated anencephaly and 75% for isolated spina bifida, with detection rates between centres ranging from 33% to 100% [10, 65]. Copp and colleagues report that routine second-trimester ultrasound detects 90–98% of fetuses with myelomeningocele in countries offering it, and that the minimum standard for the UK programme is 90% [8].

Spinal examination should be systematic, in sagittal, axial and coronal planes, from C1 to the sacrum; skin-covered lesions are rarely identified in utero, and ultrasound does not reliably exclude neural tissue within the sac [8]. The lemon sign (loss of frontal bone convexity) is present in virtually all fetuses with myelomeningocele between 16 and 24 weeks, but only in 30–50% after 24 weeks; the banana sign (cerebellar shape due to downward traction) is detectable from 14 weeks; cerebellar anomalies are present in 95% of fetuses, with predominance of the banana sign (72%) before 24 weeks and a non-visualised cerebellum (81%) later; ventriculomegaly is present in up to 70% of cases in the second trimester and almost all in the third [8]. Three-dimensional ultrasound estimated the defect level within one segment in 86% of cases in one study, but the anatomical level may not coincide with the functional level, and ultrasound did not predict postnatal mobility or intellectual function [8]. The Brazilian book includes nuchal translucency between 11 and 14 weeks as a screening test and two- and three-dimensional ultrasound in the second and third trimesters [7].

Ultrasound screening in France. The National Conference on Obstetric and Fetal Ultrasound (CNEOF, 2016) recommends three routine scans (11 weeks to 13 weeks and 6 days, 20–25 weeks and 30–35 weeks of gestation), with mandatory elements in the report: in the first trimester, biparietal diameter, midline and skull contours; in the second and third, the spine, skull contour, lateral ventricles, posterior fossa and cerebellum [18]. Prenatal detection of open dysraphism exceeds 85%: 81% in the Alsace registry (1995–2009, rising from 75% to 84%), 90.1% in the Paris registry (myelomeningoceles 1992–2000) and 89.1% for myelomeningoceles in Europe (EUROCAT, 2012–2016) [18]. In Alsace, the diagnosis was made before 14 weeks in approximately 10% of cases and between 15 and 28 weeks in the great majority (the published percentages by trimester in the argumentaire add up to more than 100%) [18]. Of the five forms covered by the French PNDS (National Diagnosis and Care Protocol), namely myelomeningocele, myelocele, LDM, Myelo-LDM and lipomyelomeningocele, all are accessible to prenatal ultrasound except non-saccular LDM [18]. A family history or first-trimester anomalies may justify an early anomaly scan, at around 18 weeks [18].

Ultrasound signs according to the argumentaire of the perinatal PNDS. In the first trimester, direct visualisation of the defect detects 50–60% of myelomeningoceles (only from 13 weeks in a German study of 16,164 fetuses), with greater difficulty in low defects and small sacs; indirect posterior fossa signs have highly variable sensitivities and searching for them is not recommended in routine practice (table 5.1a) [18]. In the second and third trimesters, the direct signs are the anechoic posterior fluid-filled mass with a neural placode on the dorsum of the sac and visible roots (sagittal plane), U- or V-shaped splaying of the laminae (axial plane) and loss of parallelism of the spinal margins (coronal plane); in myelocele, with no sac, the direct signs are subtle and the cranial signs guide the examination [18]. Chiari II malformation is present in 100% of myelomeningoceles, ventriculomegaly (more than 10 mm) in 46–90% and microcephaly (head circumference below the 3rd centile) in 46–70%; in a cohort of 70 fetuses, 67% had other brain anomalies, with corpus callosum anomalies in 60%, periventricular nodular heterotopias in 11% and gyration anomalies in 3% [18]. The extracerebral signs are equinovarus foot and leg amyotrophy [18]. In saccular LDM, the defect preserves convergence of the laminae, the sac has a thick covering continuous with the skin, there are one or more stalks between the cord and the covering, with no placode or roots, and there are no cranial signs (the cisterna magna is often small); in lipomyelomeningocele a hyperechoic intracanal mass is seen in contact with the low-lying cord; in the 20 fetuses with Myelo-LDM in the Vande Perre series, the cisterna magna was reduced in 95%, there was frontal bone flattening in 50%, Chiari II in 40%, a short corpus callosum in 37%, moderate ventriculomegaly in 50% (mean 13 mm) and a sac in 90%, with no lower-limb anomalies [18].

Table 5.1a. Indirect first-trimester signs in myelomeningocele (argumentaire of the French perinatal PNDS).

Diagnostic ultrasound. When a defect is suspected, second-line ultrasound should be performed by an operator trained in all forms, including LDM and Myelo-LDM, and should look for the skin covering, the upper level of the lesion, cerebral and extracerebral anomalies and signs of a syndromic form, including an anorectal anomaly suggesting Currarino syndrome; the PNDS advises against establishing a definitive diagnosis at the first contact with the couple [18]. In a French prenatal repair programme, of 126 fetuses referred for myelomeningocele, the diagnosis was corrected to LDM in 11 and to Myelo-LDM in 20; the multidisciplinary clinic at Hôpital Trousseau has received 120 couples since 2014 and refined the diagnosis in 20% [18]. The argumentaire proposes major and minor criteria to distinguish open from closed forms (table 5.1b) [18]. The ultrasound anatomical level correlated with postnatal motor function to within one level in 65% of cases (fetal MRI 59%); in another study, agreement with postnatal MRI was 33% (ultrasound) and 48% (MRI), and a 2021 meta-analysis (14 studies, 655 fetuses) considered the two techniques equivalent; routine functional assessment of the lower limbs is not recommended [18].

Table 5.1b. Major and minor criteria for open and closed dysraphism (argumentaire of the French perinatal PNDS).

5.2. Alpha-fetoprotein and acetylcholinesterase

Maternal serum AFP screening with a threshold of 2.5 multiples of the median (MoM) at 16 weeks has a detection rate of 82% for spina bifida and a false-positive rate of 1.9%; it does not detect closed defects and is less sensitive in women taking valproate, and should therefore be used alongside detailed fetal anomaly ultrasound, particularly in a high-risk pregnancy [10]. Smith reports that serum screening with detailed ultrasound detects more than 90% of affected pregnancies [66]. In maternal diabetes, the laboratory should be informed because AFP reference values require adjustment, and ultrasound detection of fetal anomalies was lower in diabetic women (30%) than in non-diabetic women (73%) in one series, largely because of obesity [12]. Amniotic fluid AFP and acetylcholinesterase are diagnostic, but amniocentesis is now rarely used solely to diagnose neural tube defects, given the effectiveness of ultrasound combined with serum screening [8, 10, 67]. Copp and colleagues consider biochemical screening increasingly redundant, with maternal obesity, which hampers ultrasound, its main current indication [8]. In France, since trisomy 21 screening moved to the first trimester, maternal serum AFP measurement is no longer used, and the 2021 perinatal PNDS (French National Diagnosis and Care Protocol) considers it not indicated, as is non-invasive prenatal screening on maternal blood in the presence of a dysraphism [18]. The PNDS recommends offering amniocentesis to exclude a chromosomal anomaly (karyotype and chromosomal microarray, possibly as a trio with the parents) and to confirm the open nature of the defect by qualitative cholinesterase electrophoresis, performed by a laboratory authorised for prenatal diagnosis: the butyrylcholinesterase band is normal and the acetylcholinesterase band appears only when CSF passes into the amniotic fluid [18]. Acetylcholinesterase is generally absent in LDM and present in 80% of Myelo-LDM, which makes the test useful when imaging is equivocal between forms; in the event of termination of pregnancy, amniocentesis may be performed at the time of feticide, beyond 20–22 weeks [18].

5.3. Magnetic resonance imaging and assessment for fetal surgery

Presurgical assessment for fetal repair includes obstetric assessment, investigation for genetic or chromosomal syndromes, ultrasound to assess lower-limb function, clubfoot and vertebral level, fetal echocardiography and ultrafast fetal MRI to document hindbrain herniation, hydrocephalus and other brain anomalies [8]. In the MOMS trial, hindbrain herniation and a normal karyotype were inclusion criteria (section 7) [26]. The French perinatal PNDS (National Diagnosis and Care Protocol) does not recommend fetal MRI for all dysraphisms, because its spatial resolution is lower than that of ultrasound and it does not correct misdiagnoses; it should be discussed and offered case by case when the pregnancy continues, with or without fetal surgery, to refine the neurodevelopmental prognosis, and when fetal presentation or technical conditions hamper ultrasound [18]. MRI better distinguishes a small posterior fossa from a true Chiari II and better detects subependymal heterotopias, but assesses the corpus callosum less well; in a series of 70 myelomeningoceles, MRI performed in 12 added information in 2 (heterotopias), and it has not been shown that these associated brain anomalies change the prognosis [18].

5.4. Genetic assessment: what to test and what not to test

In clinical practice, neural tube defect predisposition genes are not routinely investigated because known variants are rare missense variants or low-relative-risk polymorphisms [10]. If the neural tube defect is the only anomaly, karyotyping is not routinely indicated according to Oxford; genomic analysis (microarray, gene panel, exome or genome sequencing) is indicated when signs suggest a chromosomal or syndromic aetiology, and radiography or MRI when the defect is poorly defined or parental dysraphism is suspected [10]. Copp and colleagues indicate that karyotyping should be offered when there are other anomalies or risk factors, such as advanced maternal age [8], and Chen’s review recommends cytogenetic investigation for a perinatal neural tube defect [54]. The differences reflect the period and context (postnatal counselling of a family versus prenatal decision-making), and the common point is that the likelihood of a chromosomal cause increases with associated anomalies.

MTHFR genotyping (c.665C>T, c.1286A>C) should not be requested to assess or prevent neural tube defects. The 2013 American College of Medical Genetics and Genomics (ACMG) guideline concluded that MTHFR polymorphism testing has minimal clinical utility and should not form part of routine thrombophilia assessment, and the 2020 addendum reaffirmed this position, including recurrent pregnancy loss [6, 44, 68]. An MTHFR result obtained through direct-to-consumer testing does not, by itself, justify high-dose folate; if there is clinical uncertainty, the appropriate test is plasma total homocysteine rather than genotype [6]. The USPSTF does not alter its supplementation recommendation on the basis of MTHFR variants [19].

When a panel, exome or genome is performed for suspected syndromic disease, the report should state the gene, reference transcript with version (for example NM_005957.5 for MTHFR), HGVS nomenclature, zygosity, classification according to the ACMG/AMP five-class system and limitations (coverage of structural variation, non-coding regions, mosaicism) [38, 42]. A rare heterozygous missense variant in a PCP gene, even if absent from controls, should be reported as a susceptibility factor or variant of uncertain significance and does not become causal through partial segregation [6, 40]. The French perinatal PNDS (National Diagnosis and Care Protocol) summarises the same position for prenatal diagnosis: no major gene has been identified in humans, and the only genetic investigation performed routinely in the presence of a dysraphism is the search for a chromosomal anomaly, because aneuploidies (trisomies 13 and 18) are generally accompanied by other malformations; it gives Currarino syndrome, due to variants of HLXB9 (former name of MNX1), as an example of a monogenic form [18]. Known factors explained only 28% of cases in a population of pregnant women with an affected fetus [18]. The 2015 PNDS on physical medicine and rehabilitation also mentions the statistical association of the MTHFR C677T variant (c.665C>T in current nomenclature) with spina bifida, which describes the epidemiology and does not alter the recommendation not to genotype [32]. For lipoma of the filum terminale, the 2021 PNDS describes rare isolated familial cases and no identified genetic basis [28]. The GeneReviews on severe MTHFR deficiency notes the existence of pseudogenes that may complicate testing [42].

5.5. Diagnostic algorithm

Table 5.5a. Imaging indications when closed dysraphism is suspected (annex 3 of the French PNDS (French National Diagnosis and Care Protocol) for lipoma of the filum terminale).

5.6. Functional urological and bowel assessment in adults

Initial urological assessment. At the first assessment of the adult, the French PNDS (French National Diagnosis and Care Protocol) proposes a specialist neuro-urology consultation with a bladder diary and questionnaires, computed tomography (CT) with or without contrast or renal and bladder ultrasound with or without a plain abdominal radiograph, creatinine clearance (ideally with 24-hour urinary creatinine) or cystatin C measurement, and urodynamic study; urine culture is requested only if there are symptoms [17]. Serum creatinine alone does not reliably assess renal function in this population (grade B), except in closed dysraphism with an ambulant patient, in whom the glomerular filtration rate estimated with the CKD-EPI equation appears sufficient; in the remaining cases, 24-hour urinary creatinine clearance is used if there is no incontinence, or cystatin C-based equations, whose superiority over creatinine-based equations has been demonstrated only in children and which was not reimbursed in France [17]. The prognostic value of bladder wall thickness on ultrasound is controversial and does not replace urodynamics; urinary cytology lacks sufficient evidence for bladder cancer screening (grade C) [17]. Four urinary quality-of-life questionnaires have been validated in spinal cord injury (I-QOL, Qualiveen, SF-Qualiveen and NBSS) and can be used during follow-up [17].

Bowel assessment. The 2021 PNDS directs the history towards the predominant symptom (constipation, anal incontinence or alternating constipation and overflow diarrhoea), the number and consistency of stools, defecation habits and manual evacuation; in constipation, persistence of the urge sensation, the need for abdominal straining or digital manoeuvres (perianal, intra-anal or intravaginal), the sensation of incomplete evacuation and any history of faecal impaction; in incontinence, whether it is passive (without urge sensation) or urge-related, the number of weekly episodes and the use of pads; and also bloating, abdominal pain, a sensation of prolapse or pelvic heaviness, and urinary and sexual complaints [27]. Useful instruments at the initial assessment and during follow-up are the stool diary, the Bristol scale (seven consistency types) and the neurogenic bowel dysfunction score, which correlates with quality of life; digestive quality of life can be measured with the Gastrointestinal Quality of Life Index (GIQLI; 36 items, score from 0 to 144, higher values indicating better quality of life) [27]. The assessment also considers the speed of movement and transfers to the toilet, the cognitive abilities needed for therapeutic education and the family and occupational context, with support from a neuropsychologist or occupational therapist when necessary [27].

Physical examination covers the three perineal compartments (anorectal, urological and gynaecological), in the lateral decubitus position when mobility is reduced, at rest, during straining and during squeeze [27]. It includes sensation and motor function of the sacral segments, digital rectal examination (anal tone, voluntary contraction, dyssynergia on straining; marked hypotonia and a gaping anus suggest a peripheral neurological lesion) and the sacral reflexes (bulbocavernosus and clitoroanal, S3; anal stretch and nociceptive reflex, S3–S4; anal cough reflex, with T6–T12 afferents and S3–S4 efferents), whose interpretation is difficult [27]. Rectal prolapse or high-grade internal rectal prolapse (anoscopy on straining), genital prolapse (cystocele, rectocele, uterine prolapse, enterocele), fissure, haemorrhoids, perianal dermatitis and pressure ulcers are sought; rectal prolapse or high-grade internal prolapse may justify self-catheterisation in a patient who voids by abdominal straining [27].

Colonoscopy has the same indications as in the general population [27]. Anorectal manometry and colonic transit time with radio-opaque markers help to understand the pathophysiology and guide treatment after failure of first-line measures, because neither correlates with the neurological level [27]. In a case–control study of 93 children operated on for myelocele or myelomeningocele and 80 controls, resting anal pressure was lower in patients and higher in those who were continent, and four manometric profiles were defined: normal tracing (61% of patients, 63% incontinent), spontaneous relaxations of the upper anal canal with rectal contractions (13%, 91% incontinent), unstable pressure with ultra-slow waves (18%, 92% incontinent) and anal hypotonia without the rectoanal inhibitory reflex (8%, all under 2 years) [27]. In 132 adults at a reference centre, 62.9% with severe anal incontinence, absent anal canal sensation, a maximum tolerated volume above 330 mL and a rectoanal inhibitory reflex above 75% on insufflation of 50 mL were independently associated with severe incontinence, without association with the neurological level; in 17 patients allergic to latex, rectal perception could not be tested [27]. In 22 patients with myelomeningocele and 22 controls, colonic transit time was longer in patients and did not correlate with stool frequency or neurological level [27]. A lower rectal sensory threshold predicted response to biofeedback, and weak voluntary contraction, low urge and maximum tolerated volumes and slow colonic transit predicted success of transanal irrigation [27]. Defecography or MR defecography is used for complete study of pelvic statics before surgical correction, a plain abdominal radiograph to document faecal stasis in spurious diarrhoea, endoanal ultrasound after vaginal delivery or proctological surgery, and perineal electrophysiological tests to confirm a neurological origin in case of doubt [27].

5.7. Lipoma of the filum terminale: confirmation and assessment of severity

Spinal MRI is the investigation of choice: fat in the filum gives spontaneous T1 hyperintensity; on sagittal slices, the filum appears as a white intradural line between the conus and the bottom of the dural sac, but it may be thinner than the slice, so the PNDS (French National Diagnosis and Care Protocol) recommends systematic axial T1 lumbosacral slices (the lipomatous component appears as a hyperintense dot) and sagittal T2 slices to locate the end of the conus [28]. MRI requires sedation in the newborn and young child and is necessary before any surgery [28]. Ultrasound, without sedation, serves as screening in infants: it assesses the position of the conus, cord pulsatility and filum thickness, with a conventional threshold of 2 mm, and a threshold of 1.1 mm at 4–6 months correlated with lipoma; it is less sensitive than MRI, operator-dependent, most useful in the first weeks and no longer possible after 6 months [28]. Plain radiographs are of almost no use for detecting closed dysraphisms and are used only for associated bony anomalies [28]. Cord tethering can be documented by reduced or absent oscillation on dynamic ultrasound (M and B mode) or by absence of change in conus position between prone and supine MRI [28].

Filum lipoma is an incidental finding in 0.24–6% of lumbosacral MRI scans performed for other reasons, and only about 5% of these patients have symptoms [28]. The initial assessment includes history, complete examination of the spine and extremities, peripheral neurological examination and systematic simple urological assessment: a 2-day voiding diary (times, number and volume of voids), ultrasound of the upper and lower urinary tract with post-void residual according to age, and uroflowmetry with surface perineal electromyography [28]. If the child has no symptoms and all three investigations are normal according to International Children's Continence Society (ICCS) standards, nothing further is done; if there are urinary symptoms or any abnormality, urodynamics are arranged (uroflowmetry with electromyography, cystometry with bladder and rectal pressures, and urethral pressure profilometry), whose most frequent abnormality in tethered cord is detrusor overactivity; urodynamics can be performed at any age, but are difficult to interpret before the age of continence, when physiological overactivity persists [28]. Radiographs of the spine or lower limbs are requested only if the examination suggests deformity; somatosensory evoked potentials (reduced amplitude in tethered cord) may be performed at baseline and during follow-up, but not systematically; the Necker Enfants Malades (NEM) personal score, with versions for the newborn and infant and for the child and adult, is used to measure initial severity and follow progression [28]. The diagnosis is disclosed at a dedicated paediatric neurosurgery or multidisciplinary consultation, explaining the MRI, the risk of progressive neurological symptoms, warning signs, the relative simplicity of filum section, the prompt indication if symptoms are present and the possibility of prophylactic surgery, and the investigations and frequency of follow-up [28].

6. Differential Diagnosis

The first differential step is to distinguish a primary neural tube defect from a defect occurring as a component of a syndrome, chromosomal abnormality or teratogenic exposure [10]. The Brazilian book organises the differential by region: in the brain, besides the defects themselves (anencephaly, acrania, acephaly, acalvaria, iniencephaly, encephalocele) and Arnold–Chiari malformation, microcephaly, congenital hydrocephalus, Dandy–Walker malformation, holoprosencephaly, lissencephaly, pachygyria, hydranencephaly, septo-optic dysplasia, megalencephaly, porencephaly, schizencephaly and polymicrogyria; in the spinal cord, Currarino syndrome, diastematomyelia and syringomyelia [7].

7. Treatments and Risk Management Strategies

7.1. Primary prevention with folic acid: evidence

The Cochrane review summarised by the World Health Organization (WHO) included five trials (7391 women) and showed that periconceptional folate supplementation reduces the risk of a pregnancy affected by a neural tube defect by more than two-thirds (relative risk [RR] 0.31; 95% CI 0.17–0.58; 5 trials, 6708 births), with a similar effect on recurrence (RR 0.34; 0.18–0.64; 4 trials, 1846 births); in the only first-occurrence trial there were no cases in the supplemented group (RR 0.07; 0.00–1.32); there was no effect on other congenital anomalies or difference in miscarriage risk (RR 1.10; 0.94–1.28), and evidence quality was high for neural tube defects [76]. Participants were Caucasian and Indian women studied before widespread fortification [76]. The systematic update for the USPSTF (12 observational studies, n = 1,244,072) found additional evidence of benefit and no evidence of harm relating to multiple pregnancy, autism or maternal cancer [77].

These trials have limitations relevant to counselling: event numbers are small, the 72% protection in MRC is relative and applies to high-risk women, and the first-occurrence trial used a multivitamin, preventing isolation of the folate effect; evidence for first occurrence with folic acid alone is statistically imprecise in the meta-analysis [33, 76]. Protection is incomplete: Oxford adopts a 10% risk following a recurrence despite high-dose folate and recommends reconsidering a syndromic aetiology [10]. Wald and colleagues argue for a logarithmic relationship between serum folate and risk, without a threshold, and estimate that the 4 mg/day trial raised median serum folate from 5 to 44 ng/mL with an 83% preventive effect, whereas current fortification (serum folate 10–16 ng/mL) would produce an approximately 20% effect; this is modelling rather than a new trial [79].

7.2. Recommended doses and divergences between sources

In the sources consulted, folic acid for prevention does not have a medicine-label-regulated indication comparable to those of prescription medicines; the following doses derive from the cited guidelines and trials.

The 1992 US Public Health Service recommendation set 0.4 mg/day for women of childbearing age and, for women with a previous affected child, 4.0 mg/day from 1 month before conception until 3 months of pregnancy [22, 35]. The 4 mg dose corresponds to that used in the MRC trial; 5 mg is the British and Portuguese formulation, and Oxford states that no known or suspected adverse effects have been recorded with 5 mg/day, with no contraindications, but that women with epilepsy should review their anticonvulsant treatment; no maximum dose for periconceptional supplementation is defined in the sources cited [10, 25]. For second-degree relatives, Oxford estimates that 5 mg/day reduces an absolute risk of 1–2% (versus 0.1–0.5% in the population) by approximately 85%, compared with approximately 36% with 400 µg [10]. In fetal anticonvulsant syndrome, the same book recommends 5 mg/day before conception [57]. The US valproate label recommends folic acid supplementation before conception and during pregnancy in women of childbearing age taking it, but states that it is unknown whether supplementation reduces neural tube defect risk or decreased IQ in exposed offspring [58]. In Waardenburg syndrome type 1, supplementation is recommended without a dose defined by human studies [50]. High-dose folate use for indications not listed by the above sources (for example, isolated MTHFR genotype) is not supported by the sources consulted [6, 44]. The 12 weeks of gestation of the French recommendation correspond to about 10 weeks after conception, a limit slightly earlier than the 12 weeks of Oxford and the DGS and the 2–3 months of the USPSTF; the urology PNDS (French National Diagnosis and Care Protocol) is the only source consulted that extends 5 mg/day until at least 2 months after conception when one of the parents has spinal dysraphism (section 7.15) [10, 17, 19, 20]. The PNDS points out that, because the malformation occurs before the woman knows she is pregnant, the prescription should be made when the wish for pregnancy is expressed, which explains why few women follow it correctly [17]. The 2021 perinatal PNDS advises the same population dose (0.4 mg/day), ideally 4 weeks before the start of pregnancy and until 12 weeks of gestation, and 5 mg/day with a history of a child with a neural tube defect, pregestational diabetes or antiepileptic treatment; after a pregnancy with dysraphism, it recommends 5 mg/day for all types, although efficacy has only been evaluated after myelomeningocele; and it maintains 0.4 mg/day in obesity, for lack of data on higher doses [18]. The argumentaire cites the 2006 WHO recommendations (start ideally 2 months before, until 12 weeks, 0.4 mg and 5 mg in the same risk situations), a 2010 meta-analysis with a 70% reduction in secondary prevention and 62% in primary prevention, and the 2010 Cochrane review (RR 0.28, whatever the dose) [18]. In France, folic acid intake before pregnancy rose from 14.8% in 2010 to 23.2% in 2016 (National Perinatal Survey), and pharmacy dispensing in 2006–2016 was 14.3% in the periconceptional period and 6.4% before conception, values probably underestimated because they do not include purchases without prescription [18].

Folic acid should begin before conception because the neural tube closes during the first month of gestation [8, 20]. In Brazil, more than 55.4% of pregnancies are unplanned and adherence to supplementation was 4.3% during the periconceptional period and 31.8% during pregnancy in the studies cited by Rodrigues and colleagues [80]; in the United States, 17% of Hispanic women and 30% of non-Hispanic white women reported taking ≥400 µg/day [22]. These figures explain the complementary role of fortification.

7.3. Food fortification and population targets

At population level, WHO recommends erythrocyte folate concentrations above 400 ng/mL (906 nmol/L) in women of reproductive age to achieve the greatest reduction in neural tube defects (strong recommendation, low-quality evidence), emphasising that this threshold does not predict individual risk, that no serum threshold is recommended and that the microbiological assay is the reference method; in the underlying Irish study, values <340 nmol/L were associated with 66 neural tube defects per 10,000 live births and values above 906 nmol/L with 8 per 10,000 [21].

Oxford records that fortification was associated with variable reductions in different regions, and Copp and colleagues describe reductions of approximately 25% in the United States and greater reductions (>50%) in Chile, Costa Rica, Canada, South Africa and Saudi Arabia, approximately proportional to initial prevalence [8, 10]. Lipomyelomeningocele did not decrease after fortification in Canada, suggesting no folic acid protection in closed dysraphisms [10]. Introduction of mandatory programmes in Europe remains under discussion: some authors question whether existing programmes reach susceptible pregnancies, while others advocate weighing preventive benefit against possible risks to other population groups [8]. The French PNDS (French National Diagnosis and Care Protocol), citing Mills and Signore (2004), reports reductions of 26% in the United States and of about 50% in Canada after the 1998 fortification, and 87 countries with mandatory fortification of industrially processed cereals in 2018, mainly of wheat flour and sometimes of maize flour or rice; the 26% approaches the 28% of the 2015 CDC report, obtained with a different period and method [17, 22].

7.4. Management of teratogens and maternal conditions before conception

Antiepileptics. Oxford recommends optimising treatment before pregnancy: reassessing the indication and considering gradual withdrawal in women in remission with a low relapse risk; choosing the medicine with the lowest teratogenic potential, without prescribing valproate to girls, adolescents, women who could become pregnant or pregnant women unless other treatments are ineffective or not tolerated; preferring monotherapy at the minimum effective dose and avoiding valproate doses ≥700 mg/day wherever possible; measuring plasma concentrations; avoiding switching antiepileptics once pregnancy is established; and offering prenatal diagnosis [57]. The mother should not stop or alter medication without prior discussion with the neurologist [57]. European and US regulatory measures are summarised in table 7.8.

Diabetes. Glycaemic control should be optimised before conception, and diabetic women require specialist maternal–fetal medicine obstetric follow-up [12]. In Bell’s cohort, periconceptional glycaemia was the most important modifiable risk factor (adjusted OR 1.3 for each 1% HbA1c above 6.3%) [59]. The argumentaire of the French perinatal PNDS (French National Diagnosis and Care Protocol) cites a North American and Canadian case–control study (1976–2011; 1154 cases of dysraphism and 9439 controls) with an adjusted OR of 1.84 for pre-existing diabetes, and a Finnish study (2000–2014; 181 cases and 876,672 controls) with diabetes in 2.2% of cases and 0.5% of controls (OR 4.35); the risk approaches that of the general population when glycaemic control is optimal during organogenesis, and caudal regression syndrome is about twice as frequent with maternal diabetes [18]. The same document recommends reviewing the indication for the antiepileptic with the neurologist before pregnancy (valproate appears riskier than carbamazepine), preferring the lowest effective dose as monotherapy and discussing a switch to a lower-risk drug [18].

Hyperthermia. It is recommended that women of reproductive age who might be pregnant limit hot tub exposure to less than 15 minutes in water at 39 °C and less than 10 minutes at 40 °C [14].

Obesity. The Brazilian book includes prevention of maternal obesity, prevention of gestational diabetes, correction of zinc deficiency and B-complex vitamins among prenatal care measures [7]; the association between obesity and spina bifida may not be modified by folic acid [8]. The French perinatal PNDS considers a body mass index ≥30 kg/m² a risk factor, with no data supporting higher doses of folic acid [18].

Vitamin B12. Vitamin B12 deficiency is a risk factor independent of folate; there are no data justifying systematic measurement before conception, but the perinatal PNDS recommends measuring vitamin B12 and supplementing if necessary in situations at risk of deficiency, such as coeliac disease or bariatric surgery (gastric bypass) [18] [32].

7.5. Obstetric management and mode of delivery

Caesarean delivery before labour was advocated for open spina bifida because of an alleged better neurological outcome (Luthy, 1991), a benefit subsequently questioned (Lewis, 2004) [10]. The Brazilian book indicates elective caesarean delivery [7]; the authors of the published case in section 10 note that several professional guidelines do not consider myelomeningocele a mandatory indication for caesarean delivery [82]. After open fetal repair with hysterotomy, all subsequent pregnancies should end in caesarean delivery before labour begins [26]. Prenatal diagnosis and pregnancy termination have been practised since the 1970s and are the most common outcome of open neural tube defects in the United Kingdom and other countries [10].

The 2021 French perinatal PNDS (French National Diagnosis and Care Protocol) concludes that planned caesarean delivery should not be performed to improve motor prognosis or to reduce infectious risk in myelomeningocele without fetal surgery: studies from the 1980s–1990s suggested a motor benefit, but those of the last 20 years have not confirmed it, there are no data on infection prevention and no excess infection has been described after vaginal delivery; the same recommendation applies to closed dysraphisms [18]. Table 7.5a summarises the proposed organisation of birth [18].

Table 7.5a. Mode, timing and place of delivery according to prenatal repair (2021 French perinatal PNDS).

7.6. Postnatal treatment of myelomeningocele

Postnatal surgical closure remains the main treatment for cases reaching term [10]. Traditional management is surgery within the first 48 hours to reduce the risk of ascending infection and meningitis [8]; the Brazilian book describes protecting the placode with swabs moistened with warm saline and a plastic dressing, prone or lateral positioning, avoidance of latex exposure, antibiotic prophylaxis for meningitis and ventriculitis (ampicillin and gentamicin, without doses specified in the source), closure on the first day of life or as soon as possible, and ventriculoperitoneal shunting if hydrocephalus or intracranial hypertension is present [7]. Neonatal cardiorespiratory stabilisation and frequent head circumference measurement before and after surgery are fundamental [7]. The 2021 French perinatal PNDS (French National Diagnosis and Care Protocol) entrusts the delivery room to a neonatologist or neonatal intensivist and, in unrepaired open dysraphism, requires prepared equipment (warm saline, sterile swabs, impermeable dressing, “doughnut” mattress, latex-free sterile gloves), a warmed table with a sterile drape, lateral decubitus for suction and drying and then a foam mattress cut out under the lesion, which is covered with swabs soaked in warm saline; antibiotic prophylaxis is given only if the sac ruptures, latex gloves are proscribed because of increased sensitisation, and the newborn is admitted to neonatology, transferred in an incubator, because of the risk of acute multisystem complications [18]. In closed dysraphisms or open dysraphisms already repaired in utero, the place of admission is discussed case by case [18]. Closure should be performed in the first days of life: in a prospective study, surgery after the third day was associated with more central nervous system infections, and another group described a fivefold higher risk after 48 hours; the effect on neurological prognosis is discordant between studies [18]. Surgery consists of dissecting the plaque, closing it on itself, replacing it in the canal and closing in several layers (dura mater, fascia, subcutaneous tissue and skin) [18]. After prenatal repair there is no neonatal surgery, except for incomplete repair with CSF leakage [18].

Hydrocephalus is present in about 80–90% of newborns with open dysraphism, but the shunting rate varies from 51% to 91% between teams [18]. In the absence of consensus criteria, the PNDS indicates treating ventricular dilatation when it is associated with deviation of head circumference, signs of intracranial hypertension (bulging fontanelle, suture diastasis, sunsetting gaze), radiological progression, head circumference above the 95th centile, syringomyelia, or postoperative CSF leakage or a tense meningocele; shunting can be performed in the same operation as closure if hydrocephalus is severe from the outset, and closure may transiently worsen dilatation [18]. Shunt complications are very frequent: infections weigh more on cognitive development than hydrocephalus itself, the revision rate reaches 95% in some series, with patients operated on up to 10 times, and repeated revisions worsen adult neurological prognosis; ventriculoperitoneal shunting has more malfunctions and infections, endoscopic third ventriculostomy more failures, and neither is superior [18]. After fetal repair, the indications for shunting remain the same [18].

First six months: rehabilitation, feet, bladder, bowel. Early rehabilitation follows the usual motor milestones (head control at 3 months, rolling at 4 and sitting at 6) [18]. Talipes equinovarus is treated early and conservatively, with physiotherapy centred on the talonavicular joint and undercorrecting splinting (strapping, thermoformed splint or plaster cast), sometimes only at night; surgery, from Achilles tenotomy to extensive releases, is reserved for failure of conservative treatment; because the child does not feel pain, pressure points of the splints must be monitored by the parents and orthotists to prevent ulcers [18]. The initial urological assessment includes clinical examination, observation of voiding from birth, post-void residual, serum creatinine and renal ultrasound, if possible after the first week; urodynamics are recommended, with a variable timeframe depending on the dysraphism [18]. Intermittent catheterisation may follow a proactive (systematic) or reactive strategy (only in the presence of clinical or urodynamic abnormalities or high-grade reflux, with close surveillance), with no demonstrated superiority of either; prolonged antibiotic prophylaxis is not given, only temporary cover while the parents are learning and treatment of symptomatic infections; in a North American series of 299 newborns (2015–2019), 16.1% had a urinary tract infection [18]. When detrusor overactivity is documented, an anticholinergic is started, with oxybutynin as first line [18]. The PNDS does not give a dose; the US label for immediate-release oxybutynin covers only children from 5 years of age (5 mg twice daily, at most 3 times daily) [83], and in a Korean series of 121 children with dysraphism (0–15 years; 34% treated before 5 years), prescribed doses ranged between 3 and 24 mg/day, with 90% between 5 and 15 mg/day, without serious adverse effects (constipation and facial flushing were the most frequent) [84]; in infants, the dose is therefore an extrapolation from the label and observational series, to be determined by the specialist team. In the bowel, constipation is the only problem of the first months and is diagnosed clinically (Bristol scale, Rome criteria), without investigations; osmotic laxatives (macrogol, lactulose) are used first, with a water or Normacol® Enfant enema during exacerbations, and transanal irrigation devices are not recommended before 3 years [18]. The PNDS does not give doses; the reference is the 2014 ESPGHAN/NASPGHAN guideline for functional constipation in infants and children, which excludes children with underlying disease, so its application to neurogenic bowel is an extrapolation, and the guideline found no trials on the optimal dose: for faecal impaction, oral polyethylene glycol (PEG, macrogol) with or without electrolytes 1–1.5 g/kg/day for 3–6 days, or one enema daily for 3–6 days if PEG is not available; for maintenance, PEG with a starting dose of 0.4 g/kg/day, adjusted to response (0.2–0.8 g/kg/day in the dosing table), and lactulose 1–2 g/kg once or twice daily as first line when PEG is not available, considered safe at all ages; maintenance lasts at least 2 months, symptoms should have resolved for at least 1 month before stopping, and tapering is gradual [85]. The Portuguese labels are more restrictive: paediatric macrogol 4000 (Casenlax® 4 g) is authorised only from 6 months to 8 years, with 1 sachet daily from 6 to 12 months, 1–2 sachets from 1 to 4 years and 2–4 sachets from 4 to 8 years, not exceeding 3 months and with prior exclusion of an organic cause, especially under 2 years [86]; lactulose (Duphalac® syrup, 667 mg/mL) allows up to 5 mL/day under 1 year and 5–10 mL/day from 1 to 6 years, in one or two doses, with the caveat that children should use laxatives only in exceptional circumstances and under medical supervision [87]. In infants under 6 months, macrogol is off-label use under the Portuguese label, and the prolonged duration required by neurogenic bowel exceeds the 3 months of the label; the paediatric guideline is the reference and the label the comparison. Psychological support for the family should be offered from diagnosis, if possible before birth [18].

Shunt requirement depends on level: virtually all newborns with thoracic lesions, approximately 85% with lumbar lesions and approximately 70% with sacral lesions; endoscopic third ventriculostomy with choroid plexus coagulation is an alternative in highly selected cases [8]. Severe symptomatic hindbrain herniation may require posterior fossa decompression [8]. Orthopaedic deformities are usually treated shortly after birth; bladder function is monitored with ultrasound and urodynamic studies and treated with clean intermittent catheterisation, medicines and surgery; neurogenic bowel requires suppositories, enemas, laxatives or antegrade colonic enemas; and care should be provided by a coordinated multidisciplinary team (paediatrics, neurosurgery, urology, orthopaedics, physiotherapy, nursing, social work and psychology) [7, 8]. In anencephaly, if the lesion is small, suturing may be undertaken alongside palliative care; in encephalocele, surgery shortly after birth to remove the hernial sac, repair the membrane, close the bony opening and, if necessary, perform plastic surgery [7].

7.7. Fetal repair of myelomeningocele

The MOMS trial compared open in utero repair before 26 weeks with postnatal repair [26]. Inclusion criteria were singleton pregnancy, myelomeningocele with an upper boundary between T1 and S1, hindbrain herniation, gestational age 19.0–25.9 weeks at randomisation, normal karyotype, residence in the United States and maternal age ≥18 years; main exclusion criteria were an unrelated fetal anomaly, severe kyphosis, preterm birth risk (short cervix, previous preterm delivery), placental abruption, BMI ≥35 and a surgical contraindication, including previous hysterotomy in the active segment [26].

The trial was stopped for efficacy after 183 of 200 randomisations; the Bayley mental development index did not differ between groups; there were no maternal deaths; chorioamniotic separation (26%), pulmonary oedema (6%), oligohydramnios (21%) and placental abruption (6%) were more frequent after fetal surgery; and there were two perinatal deaths in each group [26]. In the complete cohort, ankle, knee and hip movements in utero, absence of a sac and a lesion ≤L3 were associated with independent walking in the operated group, and the authors recommend counselling separately about shunt-related outcomes and distal neurological function outcomes [88]. Results should not be generalised to less experienced centres or patients outside eligibility criteria, and not all children benefited [26]. A pregnant woman whose fetus has myelomeningocele before 24 weeks has three options: pregnancy termination, continuation with caesarean delivery near term and postnatal repair, or prenatal surgery if criteria are met [8]. Copp and colleagues recommend restricting fetal surgery to high-volume centres with multidisciplinary teams and standardised protocols [8].

Interpretation. The first primary outcome is composite and dominated by “shunt requirement” according to criteria applied by a committee unaware of allocation, in an unblinded trial [26]. The absolute benefit in shunts actually placed (82% to 40%) coexists with maternal and prematurity risks extending to subsequent pregnancies [26]. The population was 92–94% white and excluded BMI ≥35, although obesity is common in pregnancies with myelomeningocele [26]; applicability to populations with higher obesity prevalence or a different ethnic composition was not tested in the trial. Comparisons between fetoscopic techniques and the open technique are based on non-randomised cohorts and historical controls.

Position of the French perinatal PNDS. Every couple whose fetus has myelomeningocele should receive information about in utero surgery, generally performed before 26 weeks of gestation, and understand that it does not cure the disease but only improves the prognosis compared with postnatal surgery; fetal surgery has been evaluated in open forms and concerns mainly myelomeningocele [18]. Neither technique (open, with laparotomy and hysterotomy, or fetoscopic, through trocars via uterine incisions of less than 1 cm) is superior to the other, but fetoscopy is considered promising because of three-layer closure and the approach after maternal laparotomy, with less preterm birth [18]. The argumentaire summarises the cited results in Table 7.7a [18]. In a single-centre French study, of 28 couples with myelomeningocele eligible for fetal surgery, 21 preferred termination of pregnancy [18].

Table 7.7a. Results of prenatal repair cited by the argumentaire of the 2021 French perinatal PNDS (French National Diagnosis and Care Protocol).

The values were checked against the abstracts of the original publications. The argumentaire text attributes to the MOMS school-age follow-up a shunting rate of 47% versus 70%, which in the original publication corresponds to revisions among children with a shunt (shunts placed were 49% versus 85%), and describes the 2 fetal deaths from uterine rupture in the subsequent-pregnancy registry as neonatal deaths; the table follows the original publications [90, 91]. These figures come from series of different design, period and selection and do not allow direct comparison of techniques; the 2020 analysis (161 children, adaptive behaviour) and the 2021 analysis (154 children, physical function; section 11.2) are distinct publications from the same follow-up [89, 90].

French programmes. The fetal medicine department of Armand-Trousseau Hospital, with the paediatric neurosurgery department of Necker Hospital, began the PRIUM 1 open repair programme in 2014; after 10 fetuses with results similar to those of MOMS, the surgery became routine care in 2018, with 18 children operated on at the time of the PNDS [18]. The PRIUM 1 criteria were: adult pregnant woman, inclusion up to 25 weeks and 6 days, singleton pregnancy, myelomeningocele with an upper boundary between T1 and S1, non-praevia placenta and isolated anomaly (the usual changes of myelomeningocele do not count as associated malformations); exclusions were multiple pregnancy, low-lying or praevia placenta, unrelated malformation, previous preterm birth before 37 weeks or cervix shorter than 26 mm before 26 weeks, active bleeding between 20 and 26 weeks, BMI above 35, previous anterior uterine surgery and surgical or anaesthetic contraindication [18]. PRIUM 2, started in March 2021, evaluates fetoscopic repair of sacral myelomeningoceles (level at or below S1) and of Mye-LDM with cerebral signs, before 26 weeks, excluding cervix shorter than 15 mm, previous late miscarriage before 22 weeks, rupture of membranes, low-lying placenta, BMI above 35, uterine anomalies and transmissible maternal infections (HIV, hepatitis B and C) [18]. The Necker Hospital department uses fetoscopy after maternal laparotomy (Belfort technique), with 6 published cases and criteria similar to those of MOMS [18].

7.8. Comparative regulatory and policy table

No medicine authorised by the FDA or European Medicines Agency (EMA) with an indication to treat neural tube defects was identified in the sources consulted; treatment is surgical and supportive.

7.9. Developments in the last 12 months and research direction

Phase 1 cell therapy results assess safety rather than efficacy; fetoscopic series are not randomised, use historical controls and differ in technique, gestational age and selection; and premature membrane rupture rates with the percutaneous approach are high. None of these data replaces MOMS as the efficacy reference. On ClinicalTrials.gov, the CuRe trial is recruiting (estimated primary completion in March 2027), and there are other studies of fetoscopic, robotic and umbilical cord graft repair [98].

7.10. Neurogenic bladder in the adult: goals and medical treatment

The PNDS (French National Diagnosis and Care Protocol) for adult urology, drawn up by the Rennes reference centre according to the 2012 HAS method and with grades of recommendation based on the 2011 Oxford Centre for Evidence-Based Medicine levels of evidence, sets two goals: preserving renal function, through screening for and treatment of neurogenic lower urinary tract dysfunction with identification of high-risk bladders, and achieving daytime and night-time urinary continence; it adds the person's autonomy in their own care [17]. In the absence of urological recommendations specific to spina bifida, the PNDS adapts those for neurogenic bladder, namely the 2016 European Association of Urology (EAU) guidelines, the 2018 Société Internationale d'Urologie and International Consultation on Urological Diseases (SIU-ICUD) recommendations for spinal cord injury and the 2007 Groupe d'études de neuro-urologie de langue française (GENULF) recommendations, and recommends that follow-up be at least coordinated, and ideally provided, by specialised multidisciplinary centres (grade D) [17]. As most adults were followed by paediatric teams, the PNDS recommends contact with those teams and, where available, a transition booklet [17]. Medical treatment, clean intermittent catheterisation possibly combined with anticholinergics, should ideally start on the basis of the urodynamic study and before the upper urinary tract is affected; early initiation is associated with better tolerance and adherence, a lower risk of renal failure and less recourse to reconstructive surgery [17].

Clean intermittent catheterisation. This is the reference drainage method in the presence of retention, incomplete emptying or high bladder pressures, and has been practised for more than 40 years; it should start as early as possible when bladder function threatens the upper tract (detrusor–sphincter dyssynergia, incomplete emptying) and, if started in childhood, should be maintained into adulthood [17]. The PNDS rules are 5–6 catheterisations per 24 hours, exclusive voiding mode (no spontaneous voiding between catheterisations, even when possible, to reduce the infectious risk), a urine output of 1.5–2 L per day so as not to exceed 400 mL per catheterisation, handwashing with soap and water or alcohol-based hand rub without sterile gloves, clean but not sterile perimeatal cleansing with water or alcohol-free wipes (disinfectants irritate the mucosa without reducing infections) and, preferably, single-use hydrophilic self-lubricating catheters, to reduce urethral trauma [17]. Training should take place within a therapeutic education programme or with an experienced nurse, with videos as a supplement for carers who did not receive the initial training, and with psychological support available at this stage and throughout follow-up [17]. Indwelling catheterisation, urethral or by cystostomy, is a last resort because of the risk of irreversible complications, infection, lithiasis and probable malignant transformation; it does not prevent incontinence and favours perineal pressure ulcers in wheelchair users; overnight catheterisation placed by the patient or carer is harmful and is not indicated [17]. External penile devices are used by many men with urinary incontinence without faecal incontinence or risk to the upper tract, but they are unsatisfactory, because the goal is continence, voiding by abdominal straining is not physiological and the penis is often small; alternatives should be sought first [17].

Anticholinergics. By extension of the neurogenic bladder recommendations and without specific studies in spina bifida, anticholinergics are the first-line medical treatment of detrusor overactivity (grade C); in case of inefficacy the drug may be switched and, if monotherapy fails, two anticholinergics may be combined; with cognitive impairment, particularly in open dysraphism, the PNDS prefers trospium chloride (Ceris®) or fesoterodine (Toviaz®), because of their poor passage across the blood–brain barrier, and points out that trospium should not be taken with meals [17]. Adverse effects are frequent and dose-dependent (dry mouth, constipation, nausea, dry eyes, asthenia) and there is a potential risk of cognitive worsening, documented mainly in older people and in populations with cognitive impairment; a case–control study of 32 children with dysraphism found no behavioural changes with prolonged use; splitting doses across the day may improve tolerance [17]. Isolated low compliance is not an indication for anticholinergics or for alpha-blockers [17]. The PNDS does not specify doses, so the reference doses are those of the labels: immediate-release oxybutynin (Ditropan®) 5 mg 2–3 times a day in adults, with a maximum of 5 mg 4 times a day and a starting dose of 2.5 mg 2–3 times a day in frail older people, and 5 mg twice a day (maximum 3 times) above 5 years of age [83]; trospium chloride 20 mg twice a day, with water, on an empty stomach or at least 1 hour before a meal, and 20 mg at bedtime if creatinine clearance is below 30 mL/min [102]; fesoterodine 4 mg once a day, up to 8 mg, with a US indication also in paediatric neurogenic detrusor overactivity from 6 years of age and above 25 kg (4 mg, increased to 8 mg) [103]. These doses were established for overactive bladder in adults or for paediatric neurogenic detrusor overactivity and are applied to adults with spina bifida by extrapolation.

Intradetrusor botulinum toxin. This is the second line in neurogenic detrusor overactivity refractory to anticholinergics (grade C), after failure of two different anticholinergics or of their combination, or because of intolerance; it may be combined with anticholinergics if efficacy is partial at the maximum dose, and it is effective in low compliance accompanied by detrusor overactivity, but not in isolated low compliance [17]. In spina bifida, a systematic review and a multicentre paediatric series showed improvement in continence and, to a lesser degree, in urodynamic parameters (reduction in maximum detrusor pressure, increase in capacity and compliance), more marked in overactive bladders [17]. The technique described includes urine culture with antibiogram 5–10 days beforehand, targeted antibiotic therapy started 48 hours beforehand if there is bacteriuria, discontinuation of anticoagulants or antiplatelet agents at least 3 days beforehand or their substitution, prior cystoscopy looking for trabeculation, diverticula, tumour, lithiasis or open ureteric orifices, and injection of the dose in a single session at 20–30 sites throughout the wall, with injection of the trigone possible; it is performed under general or local anaesthesia (instillation for 15–30 minutes of 1% or 2% lidocaine with 14‰ bicarbonate), and pregnancy is a contraindication [17]. The initial adult dose is 200 units of onabotulinumtoxinA (BOTOX®), which may be increased to 300 units in case of failure or a duration of effect of less than 6 months, a dose outside the marketing authorisation (grade C) [17]. The US label provides for 200 units in 30 injections of about 6.7 units in adults with detrusor overactivity associated with a neurological condition, 200 units (≥34 kg) or 6 units/kg (<34 kg) in 20 injections in children, and a maximum of 400 units in adults per 3-month interval, summing all indications [104].

Other options. Mirabegron, posterior tibial nerve stimulation and sacral neuromodulation have hardly been studied in spina bifida and cannot be recommended, but they may be discussed with specialised teams before more invasive options [17]. Treatment of bowel disorders, in particular constipation, contributes to bladder–sphincter balance and should precede any urinary continence surgery [17, 27].

7.11. Urinary tract infections in the adult with neurogenic bladder

Asymptomatic bacteriuria is practically constant under self-catheterisation and in continent diversions, and frequent in patients who void by abdominal straining with a residual volume; it should not be sought or treated while the patient is asymptomatic (grade A), routine urine cultures are not recommended, asymptomatic leucocyturia is not treated and, after treatment, a follow-up urine culture is not needed, because the urine becomes recolonised [17]. When symptoms justify antibiotic therapy, urine culture with antibiogram is mandatory before treatment (grade A): empirical treatment without urine culture is only accepted in uncomplicated cystitis in healthy women and does not apply to this population, which has a high risk of complications and resistant strains [17]. Narrow-spectrum antibiotics are preferred, self-catheterisation should not be interrupted (incomplete emptying would worsen the infection) and, in polymicrobial cultures, which are frequent under self-catheterisation and often interpreted as contamination, identification of each strain should be requested and all should be treated [17]. With burning, hypogastric pain or new leakage without fever or loin pain, the first step may be a moderate increase in fluid intake and in catheterisation frequency, followed by urine culture and antibiotic therapy adjusted to the antibiogram [17].

Recurrent symptomatic infections require the search for and correction of contributing factors (grade B): inappropriate voiding mode (in patients who void spontaneously, consider self-catheterisation, as a rule after a urodynamic study), irregular catheterisations or fewer than 5 per day, faulty technique (checked in all positions used, including lying down, with catheter and bag in a dependent position) or inappropriate equipment (try hydrophilic or sleeved catheters), insufficient or excessive urine output (1200–1500 mL per day), volumes per catheterisation above 400 mL (above half a litre they favour vesicoureteric reflux and pyelonephritis), post-void residual, lithiasis or upper-tract involvement (ultrasound or, preferably, CT urography), detrusor overactivity or poor compliance (urodynamic study) and untreated anorectal disorders [17]. Retrograde cystography may be indicated in recurrent pyelonephritis or upper-tract dilatation, although reflux is as a rule a consequence of bladder–sphincter dysfunction, and recurrent symptomatic infections also prompt a search for bladder cancer (section 11.5) [17]. In a 2009 survey of 280 European spina bifida centres, with 41 responses, there were no protocols or consensus on the prevention, diagnosis and treatment of symptomatic urinary tract infections in children [17]. Cranberry extracts have not shown efficacy in prevention in adults with spina bifida, although some experts and many patients use them [17].

Only after all factors have been corrected, and if infections persist in a patient on regular and well-performed self-catheterisation, without lithiasis, without uncontrolled detrusor overactivity, with good compliance, without reflux and with balanced bowel transit, may antibiotic prophylaxis be considered, preferring a weekly cyclic regimen alternating two antibiotics, following the WOCA (weekly oral cyclic antibiotic) model of Salomon and colleagues (grade C) [17]. The PNDS (French National Diagnosis and Care Protocol) specifies, as a single loading dose once a week, amoxicillin 3000 mg, trimethoprim–sulfamethoxazole 320/1600 mg, fosfomycin 6000 mg or cefixime 400 mg, choosing two antibiotics active against the flora of the 6 urine cultures collected one week apart before prescription (for example, amoxicillin 3000 mg alternating with trimethoprim–sulfamethoxazole 320/1600 mg every Monday), avoiding antibiotics with strong selective pressure, particularly fluoroquinolones, and no longer using nitrofurans for prolonged prophylaxis because of the risk of hepatic and pulmonary toxicity, although some teams continue them [17]. These weekly regimens do not correspond to a label indication; as a point of comparison, the US label of fosfomycin trometamol provides for one sachet (3 g) as a single dose in acute uncomplicated cystitis in adult women, half the weekly PNDS dose [105]. The regimen is now supported by a French open-label randomised trial in 45 adults with neurogenic bladder due to spinal cord injury, on self-catheterisation and with recurrent infections: at 6 months, WOCA prophylaxis reduced treated symptomatic infections (median 1.0 versus 2.5; p = 0.024) and febrile infections (0 versus 9 patients), without emergence of resistance in the urinary, digestive and nasal flora [106]. In the trial, each antibiotic was taken on a single day per week (on Sunday), at the dose of one day of treatment: amoxicillin 1000 mg 3 times, cefixime 200 mg twice, fosfomycin trometamol 6000 mg once or trimethoprim–sulfamethoxazole 160/800 mg twice; the daily totals coincide with the PNDS doses, which presents them as a single dose [17, 106]. The most used combination was fosfomycin trometamol with cefixime (14 patients), the most frequent adverse effects were diarrhoea and asthenia (39% each), and the trial was stopped at 45 of the 80 planned patients, because of recruitment difficulties and ethical concerns given the benefit observed [106]. Routine antibiotic prophylaxis without recurrent symptomatic infections, used in the past, has not shown efficacy, increases selection of resistant strains and is not recommended under self-catheterisation (grade A) [17].

In pregnancy, by extension of the HAS recommendation to treat asymptomatic bacteriuria in pregnant women, the PNDS recommends, with scarce data (grade D), urine culture at the start of pregnancy and cyclic antibiotic prophylaxis with antibiotics chosen according to the ecology of the latest urine cultures and without known fetal harm (consulting the Centre de Référence sur les Agents Tératogènes), with monitoring by monthly urine culture or dipstick, given that the combination of leucocytes and nitrites would have acceptable sensitivity and specificity under self-catheterisation; this is the only context in which prophylaxis is proposed without recurrent symptomatic infections, supported by a study of 30 pregnancies under self-catheterisation [17]. Before invasive urological examinations (urodynamics, retrograde cystography, cystoscopy), urine culture with antibiogram is performed in the preceding days; if sterile and with no risk factors (more than 4 infections per year, vesicoureteric reflux, known pre-voiding detrusor pressure above 40 cmH2O), no prophylaxis is given; if it shows asymptomatic bacteriuria, prophylaxis is recommended in the presence of symptoms of infection (a clear change in urinary symptoms that does not resolve with abundant hydration and more catheterisations, or suspected febrile infection) or of immunosuppressant infusion; when indicated, prophylaxis is adapted to the urine culture, starts 24–48 hours beforehand and ends after the examination [17]. Before surgery of any type, asymptomatic colonisation is in principle treated during the preceding 48 hours and the following 48 hours [17].

7.12. Reconstructive urological surgery

Surgery is a last resort after failure of medical treatment, and anorectal disorders should be treated beforehand, otherwise pads will still be needed if anal incontinence persists; the usual indication is detrusor overactivity or poor compliance refractory to anticholinergics and toxin, and upper-tract involvement is an additional argument [17]. The patient must be motivated, able to take on self-catheterisation, both the manoeuvre and its long-term management, and understand what is at stake, with careful assessment of cognitive impairment [17]. The specific literature is scarce and the PNDS (French National Diagnosis and Care Protocol) often relies on data from neurogenic bladder in general or from spinal cord injury; in an analysis of 2011 North American data, augmentation was the most used operation for neurogenic bladder in children (3403 children, versus 772 non-continent diversions) [17]. Voiding by abdominal straining is now discouraged, because of the risk of global perineal deterioration and of genital prolapse in women and rectal prolapse in men [17].

7.13. Male sexual function

Sexological care is part of management and both sexes can be seen in a specialised clinic [17]. In erectile dysfunction, phosphodiesterase 5 inhibitors are the first line (grade C), and the PNDS (French National Diagnosis and Care Protocol) prefers sildenafil because it is the only one studied specifically in spina bifida, with efficacy superior to placebo and dose-dependent; in France, at the time of the PNDS, these drugs were not reimbursed by social security [17]. The PNDS does not specify doses; the label of sildenafil (Viagra®) recommends 50 mg about 1 hour before sexual activity (between 30 minutes and 4 hours), adjustable between 25 and 100 mg, at most once a day, with contraindication of nitric oxide donors, such as nitrates [107]. As second line, or from the outset when the cost of oral drugs is an obstacle, intracavernous injections of prostaglandin E1 (alprostadil; Caverject®, Edex®) give good-quality erections in clinical practice, although without studies in this population (grade C), require training and determination of the minimum effective dose, and were reimbursed at 30% [17]. The Caverject® label recommends, in erectile dysfunction of pure neurogenic cause (spinal cord injury), starting with 1.25 µg, titrating in the office to the minimum effective dose, not exceeding 60 µg and not injecting more than 3 times a week, with at least 24 hours between injections [108]. The patient should be warned of the risk of priapism, and the PNDS proposes giving him a protocol with midodrine (Gutron®) 2.5 mg, 6 tablets as a single oral dose, and attendance at the emergency department after 2 hours of persistent rigid erection [17]. This use of midodrine is extrapolated: the US label authorises it in symptomatic orthostatic hypotension, at 10 mg 3 times a day [109]. The vacuum device is an alternative not studied in this population and limited by the frequently small penis (grade D); the three approaches may be combined after assessment of the efficacy and tolerance of each, and the penile prosthesis, as a rule hydraulic, is the last resort (grade D), does not prevent catheterisation and is only exceptionally placed in this indication [17]. Reinnervation of the glans by anastomosis of the ilioinguinal nerve to the dorsal nerve of the penis (TOMAX procedure) restored sensation in half of the glans in small uncontrolled series, with sensations referred to the inguinal region in almost half of cases, and is not yet intended for clinical practice [17].

7.14. Neurogenic bowel

The 2021 French PNDS (French National Diagnosis and Care Protocol) on bowel disability, an update of the 2011 version, is organised into thematic sheets with grades of recommendation and expert opinions and assigns to the general practitioner, as first resort, screening for chronic constipation and faecal incontinence at each consultation, initiation of first-line treatment and referral to a specialised team in case of failure; the algorithms follow the predominant symptom and a stepwise progression in complexity, and should be adapted to the disability, resources, context and cognitive abilities of each patient [27].

Table source: 2021 PNDS algorithms (pp. 21–22) [27].

Oral laxatives and prokinetics. Osmotic laxatives, in particular polyethylene glycol (PEG) 4000 (macrogol), can be used for constipation in spina bifida (grade A) and are superior to lactulose [27]. The only randomised trial in spina bifida (Rome) randomised 67 children with myelomeningocele (the PNDS states 64, aged 3–14 years) to PEG 4000 0.5 g/kg/day or lactulose 1.5 g/kg/day for 6 months: both increased weekly bowel movements and reduced encopresis, and PEG was superior, with complete remission (3 or more bowel movements per week) in 46% versus 22% at the end of the study (p <0.01) and 5.1 versus 2.9 bowel movements per week [27, 110]; adverse effects were similar and not serious (flatulence in about 50% and 60%, abdominal pain in 45% and 50%, hard stools in 30% and 40%, diarrhoea in just under 20%), and efficacy was lower in lumbosacral lesions (33% with PEG and 13% with lactulose) than in sacral or sacrococcygeal lesions (51% and 26%) [27]. There are no studies in adults with spina bifida [27]. The experts recommend taking into account the risk of liquid stools (about 20% with PEG) and of anal incontinence, preferring PEG in severe constipation documented by transit time with radiopaque markers, performing an oral colonic lavage (PEG or sodium picosulfate) if there is major faecal loading before maintenance treatment, accepting bulk-forming laxatives (mucilages), which carry less exposure to liquid stools and incontinence (expert opinion), and using stimulants occasionally in severe constipation not responding to osmotic agents [27]. For stimulant laxatives, prokinetics and naloxegol, without studies in spina bifida, there is no recommendation and prescription is empirical; the experts accept prucalopride as second line before irrigation and naloxegol as second line in opioid-induced constipation [27]. In a phase II double-blind placebo-controlled pilot trial of 23 patients with spinal cord injury (prucalopride 1 mg in 8, 2 mg in 8, placebo in 7, for 4 weeks), only the 2 mg group increased weekly bowel movements and shortened colonic transit time (measured in 4 patients), and 2 of the 4 patients with moderate to severe abdominal pain on 2 mg discontinued treatment [111]; prucalopride was authorised in France for chronic idiopathic constipation refractory to laxatives, and linaclotide and lubiprostone were not marketed in France at the time of the PNDS; naronapride and velusetrag only had phase 2 data [27]. The paediatric dose from the trial is the only dose specific to spina bifida; for adults, the US label of over-the-counter PEG 3350 indicates 17 g once a day, for at most 7 days without medical advice [112]; prucalopride (Motegrity® in the United States) is authorised for chronic idiopathic constipation in adults at 2 mg per day, or 1 mg if creatinine clearance is below 30 mL/min [113]; and naloxegol (Movantik® in the United States) for opioid-induced constipation at 25 mg per day on an empty stomach, reduced to 12.5 mg if poorly tolerated or with creatinine clearance below 60 mL/min, stopping maintenance laxatives on initiation [114]. Application of these doses to adults with spina bifida is an extrapolation from related indications.

Abdominal massage. It may be offered to constipated patients (grade C), daily and as a complement to conventional treatment, when this is not enough to obtain regular bowel movements, when there is abdominal pain or distension related to constipation, when the patients themselves or someone close to them can perform it and when the patient is motivated; precautions are suspected obstruction, an abdominal mass, recent abdominal surgery or radiotherapy, implanted intra-abdominal material and, in syringomyelia, raised abdominal pressure [27]. The technique is not standardised: as a rule manual, from the ascending to the descending colon, in sessions of 15–20 minutes, performed by the patient, a carer, physiotherapist or nurse, or by rolling a tennis ball [27]. Neurological evidence is limited to 24 patients with spinal cord injury (more frequent bowel movements and faster transit with daily massage) and to a randomised trial of 30 patients with multiple sclerosis over 4 weeks; a trial of 58 non-neurological constipated patients (15 minutes, 5 days a week, 8 weeks) showed improved quality of life and cost-effectiveness when performed by the patient or in an institution; there are no long-term data and no adverse effects described [27].

Biofeedback rehabilitation. In evacuation constipation it may be tried in adults (expert opinion) and in children (grade C) if there is an objective evacuation disorder, voluntary control of the anal sphincter measurable by manometry or electromyography and measurable rectal sensation to distension, conditions exceptionally met in spina bifida; patient adherence also counts [27]. In faecal incontinence it may be tried particularly in children (grade C; in adults, expert opinion), with training of gluteal contraction when achievable and of rectoperineal coordination during distension when rectal sensation persists; electromyographic techniques combined with medium-frequency electrical stimulation (3T) deserve attention [27]. Data are scarce: 19 of 24 children with spina bifida occulta treated with biofeedback and electrical stimulation improved; in a comparative study of 12 children with myelomeningocele, 3 of 8 with biofeedback and 3 of 4 controls improved; in a retrospective Spanish series, 10 of 12 children with myelomeningocele (5–15 years) improved continence; and, in 33 children, a comprehensive programme improved continence by more than 50% in two thirds, but the benefit attributable to biofeedback was limited to one quarter, particularly with at least two bowel movements per day and a level below L2 [27]. In spinal cord injury without major gait impairment, the benefit was 29%, versus 63% in radicular lesions and 62% without neurological lesion [27].

Transanal irrigation. It can be used in constipation and faecal incontinence (grade B), as second line after failure of well-conducted conventional treatment (lifestyle, dietary and defecation measures, manual evacuation, oral and local laxatives), preferring self-administration systems after training in therapeutic education; systems with a rectal balloon catheter (Peristeen®) improve autonomy [27]. In children, irrigation reduces faecalomas, improves continence in more than 60% and satisfies more than 60% of children and parents, with efficacy maintained for up to 3 years; with a balloon catheter, after 2 weeks of training, 66% performed 3 irrigations per week and 60% spent 15–30 minutes [27]. In adults there are no studies exclusive to spina bifida: in a randomised multicentre European trial in spinal cord injury of various causes, irrigation was superior to conventional treatment for constipation, faecal incontinence, neurogenic bowel dysfunction and quality of life, with greater benefit in patients with more reduced mobility; in an Italian study of 32 patients (12 with spina bifida), anal incontinence improved in 68% and constipation in 63% at 3 weeks; about 60% manage to irrigate without help and 75% feel less dependent [27]. In a cohort of 348 patients followed for a mean of 21 months, neurological aetiology predicted success (63%) and the 18 patients with spina bifida had 67% success; in an exclusively neurological cohort, success was 46% at 19 months and 35% at 3 years, stable thereafter, with male sex and slow colonic transit as predictors [27]. The mean volume is 616 mL in children and 700 mL in adults, 1–3 times a week, with about 30 minutes of evacuation [27]. Contraindications are spinal shock, acute inflammatory bowel disease, age under 3 years and pregnancy (medical opinion before starting), and precautions are pelvic irradiation, anorectal surgery less than 6 months previously, chronic inflammatory bowel disease, grade 3–4 haemorrhoids and externalised prolapse [27]. Mortality was zero and perforation rare (2 in 110,000 procedures, less than 0.002%); practical problems were pain on insertion (25%), catheter expulsion (39%), fluid leakage (75%) and need for help (15%), with discontinuation because of expulsion or leakage in 17.7% of 62 patients; above T6, symptoms of autonomic dysreflexia may occur, which in the European trial were not serious and tended to diminish [27]. A British analysis estimated a lifetime saving of £21,768 per patient compared with conventional treatment [27].

Antegrade colonic irrigation. Continent caecostomies (Malone, Monti-Malone, MACE or Macedo-Malone procedures, and percutaneous or endoscopic caecostomy button, CHAIT type) may be offered for incontinence and defecation disorders (grade C) when all the criteria are met: request from a motivated patient who understands the demands of the method, major social and personal disability, technical feasibility and failure of properly performed conventional methods [27]. The literature does not define the ideal timing; the Belgian and French series required a minimum age of 10 years [27]. Pseudocontinence reaches 93% after a mean of 22 months and 89% at 75 months; volumes range from 500 to 1500 mL, with 3 irrigations per week recommended, using water with possible addition of bisacodyl or PEG, and a median weekly time of 150 minutes (60–240) [27]. After the postoperative period, complications occur in more than one third of patients: stomal stenosis (13% in a systematic review; 42% in a paediatric series), stomal leakage (14%), stomal prolapse (6.5%), abdominal pain (11.3%) and caecal volvulus (2 of 164); 12.8–20% abandon the technique within 21–52 months; a left colonic location had similar efficacy, fewer stomal complications (25% versus 42%) and shorter evacuation, and the ileal conduit had three times more faecal leakage than the CHAIT button, which causes more granulomas [27]. In a Belgian cohort of 80 young people with spina bifida (5–18 years), 25% had a caecostomy, performed in 85% of cases at the same time as bladder surgery, and one in five discontinued the technique after a median of 21 months, mainly because of stenosis [27].

Sacral neuromodulation. It may be offered for anal incontinence (grade C) when conventional treatment and treatment of transit are insufficient, the neurological lesion is incomplete (at least one sacral reflex arc present, verifiable in the acute test) and the disease is little or not progressive, with no need for repeated MRI, which is a relative contraindication with a stimulator; in France, this indication was outside the scope of reimbursement of the device, and there is no argument for using it in constipation [27]. In a randomised crossover trial of 33 children older than 5 years, mostly with spina bifida, 78% improved their anal incontinence (reduction of at least 50% in episodes) with active stimulation versus 17% with the stimulator switched off (p = 0.001); in an open series of 10 patients with spina bifida, electrode placement failed in 20% and only 3 had improvement greater than 50% in the test, of whom 2 were implanted [27]. The definitive implant, usually at the third sacral root, is offered after a test with a reduction of at least 50% in leakage on the bowel diary; complications, of 5–26% in non-neurological patients and similar in the two spina bifida series, are pain at the generator site, usually superficial infection and electrode migration [27].

7.15. Divergences between the French PNDSs (French National Diagnosis and Care Protocols) and other sources

7.16. Lipoma of the filum terminale: surgical indication, technique and symptomatic treatment

The aetiological treatment of tethered cord due to a filum lipoma is intradural extramedullary section of the filum, decided at a multidisciplinary meeting of a reference centre; there is no absolute contraindication, and the risks of severe comorbidities (for example, heart disease in the VACTERL association) are weighed against the functional risk [28]. There is strong consensus for operating on patients with symptoms and a lipoma on MRI; the prophylactic indication in the asymptomatic child remains under discussion [28]. In its favour are generally easy surgery, with few complications and good long-lasting results, the possibility of abnormal urodynamics and electrophysiology without apparent deficit, the difficulty of interpreting urodynamics in young children and the fact that, once symptoms appear, surgery offers only about 50% improvement [28]. Against it are the frequency of asymptomatic findings with modern MRI, a poorly defined lifetime risk of symptoms that is considered small, recurrence of tethered cord of 0–8.6% in series that mix dysraphisms, the increase in the absolute number of wound-healing complications (wound infections and meningitis may favour re-tethering) and the possibility of surveillance until the end of growth, the risk of symptoms in adulthood being considered almost nil by the PNDS (French National Diagnosis and Care Protocol) [28]. After section, half of the symptomatic patients improve and the other half stabilise, regardless of other malformations and of the position of the conus; younger children improve more, which gives a better prognosis to cases discovered early through cutaneous stigmata [28]. Neuropathic pain improves first, then sensory and motor deficits, and finally, less clearly, bladder and sphincter disorders; orthopaedic deformities (scoliosis, clubfoot) do not improve without orthopaedic treatment [28]. When symptoms are present with a conus below L2 and a thickened non-lipomatous or even normal filum, section is readily indicated; with a normal lumbar MRI, the indication is debated, because some studies show benefit and others results equal to those of medical treatment [28].

Particular cases. A filum lipoma coexists with another closed dysraphism in 32% of cases, especially in diastematomyelia (in which all the filum specimens examined histologically were pathological), in LDM and in dorsal lipomas; when it appears on preoperative MRI, it is treated as a secondary tethering point and sectioned at the same procedure, even if this requires another incision [28]. In Currarino syndrome (anorectal malformation, presacral mass and sacral anomaly) and in the VACTERL association, and in children at risk of developmental delay that may mask symptoms, prophylactic section is readily indicated to remove doubt about the contribution of the lipoma to bladder and bowel symptoms [28]. Before scoliosis surgery it is usual to section the filum, even when asymptomatic, to reduce the risk of neurological decompensation through traction during correction, a practice decided case by case; the filum lipoma is not considered a cause of scoliosis, unlike distal deformities of the lower limbs, in which release at diagnosis optimises the orthopaedic result [28].

Technique. The classic approach is at L5–S1, away from the conus and roots and above the bottom of the dural sac, through a midline incision of a few centimetres, interlaminar approach under the microscope, opening of the dura mater over a few millimetres, identification of the filum (wavy vessel on the surface, yellowish colour, rigidity, midline and posterior position), coagulation and section as caudally as possible, without downward traction, and watertight dural closure in layers [28]. The unilateral interlaminar and endoscopic approaches reduce postoperative low back pain and are useful in adolescents and adults, but are not suitable for very thick filum and may affect joint growth in young children [28]. The risks are surgical site infection (with dehiscence, CSF leak and meningitis, more frequent in older children), minimal bleeding and neurological deterioration that is practically nil except for errors of structure, traction or level [28]. Extradural trans-hiatal section of the filum, proposed for tethered cord, syringomyelia, Chiari I and scoliosis, is not validated by scientific societies, because tension in the extradural filum is barely transmitted to the conus; the PNDS considers it a procedure without demonstrated benefit for non-specific symptoms [28]. Intraoperative neuromonitoring, useful in conus lipomas, adds little to the safety of filum section [28]. The PNDS proposes a randomised trial between prophylactic section and surveillance [28].

Symptomatic treatment. In the bladder, the most frequent presentation is detrusor overactivity with or without detrusor–sphincter dyssynergia; the anticholinergic is started after urodynamics, if overactivity is documented, with oxybutynin first line (success rate greater than 93%), the only anticholinergic with French authorisation below 12 years, and trospium chloride above that age; intradetrusor botulinum toxin injections are reserved for intolerance or resistance [28]. With significant dyssynergia and incomplete or high-pressure voiding, clean intermittent catheterisation is offered, taught to the family or the child, with at least 5 catheterisations per day to reduce infections, and temporary antibiotic cover only during the learning period (ICCS) [28]. Anticholinergic doses are given in section 7.10 (labels) and section 7.6 (infants). In the bowel, constipation is treated with osmotic laxatives (macrogol, lactulose), then a glycerol suppository if there is rectal stasis and retrograde transanal irrigation if leakage persists, not before 3 years [28]. Neuropathic pain responds well to filum section, but may become chronic if treatment is delayed; the PNDS proposes, as first line, tricyclic antidepressants (amitriptyline, imipramine), serotonin and noradrenaline reuptake inhibitors (duloxetine, venlafaxine) and antiepileptics (gabapentin, pregabalin) and, as second line, opioids (morphine, tramadol) and topical treatments (lidocaine, botulinum toxin type A), with often unsatisfactory efficacy, without indicating doses [28]. In children, tramadol has been contraindicated below 12 years in the US since 2017 [115, 116], whereas the French SmPC for the oral drops authorises it from 3 years, at 1 mg/kg per dose, 3 to 4 times a day [117]. Table 7.16a gives, for comparison, the doses on the US labels for neuropathic pain in adults, by extrapolation of indication.

Table 7.16a. Label doses of two antiepileptics and of duloxetine, proposed by the filum lipoma PNDS for neuropathic pain (extrapolation from adult indications).

Traumatic spinal cord injury is the label indication closest to pain due to tethered cord; in children, none of the three drugs is indicated for neuropathic pain, and the paediatric label doses refer to epilepsy (pregabalin, gabapentin) or adolescent fibromyalgia (duloxetine, US only). For duloxetine, the analogous indication is diabetic peripheral neuropathic pain, and the labels diverge on the maximum (60 mg/day in the US, 120 mg/day in the EU) [120, 121]. Kinesitherapy aims to strengthen deficient muscles and stretch strong antagonists, before a deformity develops and after surgery if there is a motor deficit; psychological support is offered mainly in severe symptomatic forms and when catheterisation or irrigation is introduced [28].

Education and daily life. The diagnosis does not require a change in lifestyle: tethered cord does not contraindicate sport or leisure and does not justify adapted schooling [28]. If daily catheterisation or irrigation is needed, the family and patient receive therapeutic education in programmes authorised by the regional health agency, with progressive autonomy; at school, catheterisation can be performed by nurses with a prescription and an individualised care plan (PAI); 100% coverage (ALD) may be requested, and an application to the MDPH is justified only if there are associated disabilities [28].

7.17. Physical and rehabilitation medicine

The 2015 French PNDS (French National Diagnosis and Care Protocol) for physical and rehabilitation medicine (PRM), prepared by the Rennes spina bifida reference centre, brings together six sheets written in 2011–2012 on the basis of literature searches and expert opinion [32]. The PRM assessment records the patient's history, past medical history, lifestyle and requests, systematically looks for latex allergy, assesses psychological status with the Hospital Anxiety and Depression scale (HAD; anxiety above 12 and depression above 7 suggest a disorder), determines the sensory and motor neurological level by the American Spinal Injury Association (ASIA) score, tone, trophicity, reflexes and signs of syringomyelia, brainstem involvement, intracranial hypertension with a shunt and epilepsy, and looks for neuropsychological disorders, often overlooked by the patient and family [32]. The orthopaedic examination measures gait (Functional Ambulation Classification, FAC, from 0 to 8), sitting and standing stability, deformities of the feet, spine and pelvis, limb length discrepancy and hyperlaxity; pain is assessed with the usual scales (DN4, VAS, NRS); the skin with the scar, pressure areas and the feet; autonomy with the functional independence measure (FIM, 18 items scored 1 to 7) and technical aids, housing, school and work [32].

Table 7.17a. Orthoses and ambulation in children with dysraphism (sheet I of the 2015 PRM PNDS).

The evidence is observational: in a study of 36 pairs of children with high lesions, those who followed a walking programme had fewer fractures and ulcers and better transfers, with longer hospital stays, and 12 were still walking between 12 and 20 years; with the reciprocating gait orthosis, 4 of 15 children reached 500 m at 5 years and 8 were no longer walking at 10; short orthoses reduced oxygen consumption and increased speed and stride length in low lumbar and sacral lesions; in a prospective study of 434 patients, the lower limb axis deteriorated into valgus regardless of walking, orthoses or level [32].

Catheterisation education (sheet II). Intermittent catheterisation, with or without an anticholinergic, is the reference treatment for neurogenic bladder and sphincter disorders (level of evidence 2, grade A), and renal complications remain the leading cause of death in children with open spina bifida [32]. Starting from birth was associated with fewer bladder augmentation operations (17% versus 41% without catheterisation) and better acceptance by parents; catheterisation alone achieves continence in up to 60%, and with an anticholinergic in 75–80% or about 90% depending on the series [32]. In a series of 59 children younger than 13 years, 76% had acquired self-catheterisation, on average at 8 years, and all complications (3.5 per 1000) occurred with catheterisation performed by others; at 12–13 years, 60% remain partly dependent [32]. A child with normal motor dexterity and intelligence can learn self-catheterisation from 6 years; neither age, sex nor IQ predicts difficulty, and the best predictor is the time needed for the procedure, linked to executive functions (level 3) [32]. Adherence determines continent surgery: in a series of 47 patients, stomal stenoses occurred in 56% of non-adherent patients versus 10% of adherent patients [32]. Recommendation: start catheterisation in infancy, all the earlier if there are urological complications or signs of a high-pressure regime, combined with an anticholinergic [32].

Transition to adulthood (sheet III). About 85% of children reach adulthood, with a peak of mortality in the third decade and a mean mortality of 3% per year in a French series; the risk of disruption of care at transition is high [32]. The PNDS recommends (grade D evidence) discussing transfer during adolescence, according to maturity; appointing a coordinator in the paediatric service; writing a transition plan (coordinator, clinical summary, emergency management, skills to be acquired) and a transferable clinical record; seeing the young person alone for part of the consultation and making them an expert in their own health (skin, constipation, sexuality, folic acid, latex allergy); transferring at around 18 years, not compulsorily, during a period of stability, with joint consultations if possible; and training the adult teams [32].

Kinesitherapy and physical activity (sheets IV and V). In the only randomised trial (18 children), treadmill training twice a week doubled walking speed and improved the 6-minute test and endurance, with the effect maintained for 3 months [32]. The PNDS recommends regular lifelong physical activity (adapted sport in children; reconditioning after assessment in adults), with a minimum of 20 consecutive minutes, 3 days a week, at 50% of maximum VO2 or maximum heart rate; walking orthoses in case of instability, falls, fatigue or pain; and kinesitherapy during periods of growth, pregnancy and ageing [32]. The precautions specific to dysraphism are to avoid aggressive massage and mobilisation of the scar, manoeuvres that put traction on the roots and spinal cord, sustained effort in progressive syringomyelia, electrotherapy on insensate areas and polarised direct currents, and to take bone fragility into account in resistance manoeuvres; neuromuscular electrical stimulation is not validated (level 4) [32]. Regular lifelong PRM follow-up is recommended, with special surveillance at the transition to adulthood and during ageing [32].

Pressure ulcers (sheet VI). In the absence of specific studies, the PNDS applies the general HAS recommendations for adults: clinical assessment with the Braden scale (grade B), NPUAP classification into 4 stages, pressure-reducing surfaces (including in the operating theatre), changes of position every 2–3 hours, 30° oblique lateral decubitus, nutritional assessment, cleansing with normal saline without antiseptics, dressings chosen according to the appearance of the wound (hydrocolloids; alginates or hydrogels for debridement) and surgery with a muscle flap for deep lesions [32]. The sites most affected in the authors' experience are the ischia, sacrum and heels; two aspects are specific to dysraphism: pelvic obliquity with spinal deformity, which requires attention to sitting supports (scoliosis surgery or kyphectomy may reduce pelvic ulcers), and the tolerance of insensate skin to orthoses and footwear [32]. The Irish association proposes changing sitting supports every 20 minutes and inspecting the skin at least once a day [32].

8. Reproductive Options and Family Counselling

8.1. Family assessment

History-taking should include a three-generation pedigree with questions about stillbirths and neonatal deaths, diabetes (including gestational diabetes) in previous pregnancies and a detailed pregnancy history covering drug exposure (for example, anticonvulsants) and preconception folic acid intake [10]. Wherever possible, the affected individual should be examined or the autopsy report obtained to confirm the defect type (open or closed) and whether it is isolated; parents, or an intervening relative, should be examined for signs of spinal dysraphism, and spinal MRI should be performed for suggestive findings, in which case counselling is provided as though the person were affected [10]. The multifactorial model helps explain to parents that genetic factors come from both sides of the family [9].

8.2. Recurrence risks

In non-syndromic neural tube defects there is no tendency for the defect site (anencephaly, high or low spina bifida) to recur in subsequent affected pregnancies [10].

Radiological spinal dysraphism with two or more absent arches, with or without tethering symptoms, has an uncertain effect on the risk of open or closed defects, and isolated spina bifida occulta (a single arch, without cutaneous or neurological findings) is a common normal variant without increased risk [10]. The above risks apply to isolated defects; in syndromic, chromosomal or teratogenic forms, risk depends on the cause (for example, 25% in autosomal recessive Meckel syndrome) [7, 10].

8.3. Prenatal diagnosis in a subsequent pregnancy

Even when the recurrence risk is low, prenatal diagnosis for risk is advocated for a pregnant woman who has already had an affected child [7]. Oxford recommends targeted ultrasound (anencephaly at 11–12 weeks; spina bifida on serial scans from 16 weeks) combined with serum AFP, the sensitivity of targeted ultrasound being potentially higher than that observed in an unselected population [10]. A woman taking valproate should be informed of the lower sensitivity of AFP [10]. The French perinatal PNDS (French National Diagnosis and Care Protocol) provides, for any plan for a new pregnancy and whatever the outcome of the previous one, a preconception consultation with genetic counselling to estimate the recurrence risk, folic acid 5 mg/day, a first-trimester scan by a reference sonographer and a possible intermediate scan at 18 weeks; after an operated dysraphism, it advises waiting at least one year before a new pregnancy [18]. In France, instead of AFP, screening in the next pregnancy is based on ultrasound [18].

8.4. Reproductive options and legal framework

Options include high-dose supplementation before conception, ultrasound surveillance, termination of pregnancy after diagnosis, continuation with postnatal repair and, in eligible cases, fetal repair [8, 10]. In Brazil, legislation permits termination of pregnancy in cases of anencephaly [7]. In isolated defects, the known genetic factors are rare variants or polymorphisms of low relative risk, which is why analysis of predisposition genes is not used routinely and does not, by itself, support reproductive decisions [10]. In women with epilepsy, referral to a neurologist before a new pregnancy allows assessment of withdrawal, simplification to monotherapy or dose reduction [57]. In France, termination of pregnancy for medical reasons may be requested by the pregnant woman at any gestational age when two physicians of a CPDPN (multidisciplinary prenatal diagnosis centre) certify a strong probability of a condition of particular severity recognised as incurable at the time of diagnosis (article L2213-1 of the Public Health Code); the PNDS (French National Diagnosis and Care Protocol) considers that myelomeningocele meets this criterion even with fetal surgery, and that in LDM and Mye-LDM the decision is made case by case [18]. According to EUROCAT 2018, 78% of dysraphisms diagnosed in France ended in termination, and in the Alsace registry the proportion was 97% of myelomeningoceles with prenatal diagnosis in 1995–2004 [18]. After termination, a fetopathological examination should always be offered, with consent, to confirm the type, level and open or closed character, assess cerebral and extracerebral repercussions, look for a multiple malformation syndrome and collect samples for genetic study [18]. In a Nancy series (2004–2014; 60 fetuses, 46 terminations), autopsy corrected several diagnoses, the ultrasound vertebral level coincided in only 42% and the associated anomalies had been detected on ultrasound in 58.8%; the PNDS proposes a minimum report (lesion, brain, extracerebral findings, microscopy of the placode and radiographs) [18]. In isolated dysraphism, the aetiology is generally sporadic and the PNDS describes a very low recurrence risk without a teratogenic factor, whereas in a syndromic form counselling depends on the gene or chromosomal cause [18]; this qualitative wording should be read in the light of the empirical risks in section 8.2.

8.5. Counselling after fetal repair

Mothers undergoing open fetal surgery should understand that all subsequent pregnancies should end in caesarean delivery before labour because of the risk of uterine dehiscence and rupture [26].

8.6. Pregnancy and parenthood in people with spina bifida

To prevent recurrence, which is raised by the future child's first-degree history, the French PNDS (French National Diagnosis and Care Protocol) recommends, when one of the parents has spinal dysraphism, folic acid 5 mg/day from at least 1 month before conception until at least 2 months after [17]. Pregnancy in a woman with spina bifida and neurogenic bladder is a period of risk for urinary tract infections that are sometimes complicated, with scarce scientific data; the pregnant woman should be referred to a specialised centre as soon as pregnancy is diagnosed and followed in a multidisciplinary setting, with urine culture at the start of pregnancy and cyclical antibiotic prophylaxis with drugs without known fetal harm (section 7.11) [17]. Pregnancy contraindicates intradetrusor botulinum toxin injections, and transanal irrigation should be started in a pregnant woman only after medical advice [17, 27]. With a continent cystostomy, the possibility of a future pregnancy is taken into account in choosing the stoma site [17]. Men with spinal dysraphism often have sperm abnormalities and need further investigation, and women, although they conceive with less difficulty, need close specialised surveillance [17].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Measuring prevalence: the denominator problem

Birth prevalence depends on whether pregnancy terminations are included: in the United Kingdom most cases end in termination, whereas in Ireland all were liveborn during the period described by Oxford, so epidemiological comparisons must consider both outcomes; frequency in miscarriages is also relevant but difficult to measure [10]. Omitting terminations underestimates prevalence and may bias aetiological studies [8]. Combining live births and terminations, the average is approximately 1 per 1000 pregnancies, with values of 0.94 in Europe, 1.3 in Australia, 2.0 in Peru and 4.5 in China (per 1000 live births and terminations, during the decade preceding the Oxford edition) [10]. Copp and colleagues estimate 0.5–0.8 per 1000 births in the United States and many European countries and values more than 20 times higher in some Chinese regions, where rare subtypes (craniorachischisis, iniencephaly) are disproportionately frequent [8]. Encephalocele occurs in approximately 1 per 10,000 established pregnancies, and frontoethmoidal encephalocele is frequent in Southeast Asian populations [10]. Radiological single-arch spina bifida occulta occurs in approximately 15% of the population [10].

9.2. United States

In the 2016–2020 national estimate, based on 13 population-based surveillance programmes covering 4,141,961 live births (22% of the country’s births), race/ethnicity-adjusted prevalence was 3.48 per 10,000 live births for spina bifida without anencephaly (1 in 2875; approximately 1278 cases per year), 1.91 for anencephaly (1 in 5246; approximately 700 per year) and 0.95 for encephalocele (1 in 10,365) [16]. Spina bifida remained stable from 1999–2001 (3.68), whereas anencephaly decreased in the most recent period [16]. By maternal group, crude spina bifida prevalence was 4.21 in Hispanics, 3.68 in non-Hispanic whites, 2.35 in non-Hispanic blacks and 1.94 in Asians or Pacific Islanders; Hispanic mothers had the highest prevalence of anencephaly and spina bifida [16]. These values refer to live births and exclude terminations. The CDC data page, revised in September 2026, gives approximately 1251 babies per year (1 in 2857) and estimates that at least 92,551 insured people with spina bifida live in the United States [122]; the slight difference from the 2024 article results from the calculation presented on the page and does not alter the order of magnitude. The USPSTF estimates approximately 3000 affected pregnancies per year, with spina bifida at 3.9, anencephaly at 2.5 and encephalocele at 1 per 10,000 live births [19]. In the United States, open spina bifida is more prevalent in pregnancies of Hispanic mothers and less prevalent in black mothers [8, 10]; hypotheses include differences in folic acid intake and the MTHFR T allele, which is more frequent in Hispanics [22, 42].

9.3. Europe

In 28 EUROCAT registries from 19 countries (approximately 12.5 million births, 1991–2011), with 11,353 non-chromosomal neural tube defects (4162 anencephaly, 5776 spina bifida and 1415 encephalocele cases), total prevalence, including live births, fetal deaths from 20 weeks and terminations, was 9.1 per 10,000 births (95% CI 8.94–9.27), with 3.34 for anencephaly and 4.63 for spina bifida; prevalence showed no downward trend over the two decades despite supplementation recommendations, and live-birth prevalence, below half the total, decreased mainly for anencephaly [15]. The authors estimate approximately 5000 affected fetuses annually in Europe, most diagnosed before birth and with termination as the most common outcome [15]. Prevalence varied between registries, from 4.79 in Zagreb to 15.08 in Wales and 18.72 in Mainz [15]. In Scotland, between 2000 and 2021, total prevalence was 9.8 per 10,000 births, without temporal change, lower in children of mothers aged 30–39 years and from less deprived areas, with a recent increase in anencephaly attributed to earlier detection [81]. European evidence illustrates the main limitation of a strategy based solely on voluntary supplementation [15].

Portugal. In the Southern Portugal EUROCAT registry (1991–2010; 316,853 births), total prevalence was 4.73 per 10,000 births, with 1.96 for anencephaly and 2.40 for spina bifida, among the network’s lowest [15]. These regional data do not necessarily represent the whole country. National supplementation guidance is the 2015 DGS guidance (section 7) [20].

France. The French PNDSs (French National Diagnosis and Care Protocols) cite, for 2005–2015 in EUROCAT, a European prevalence of open spinal dysraphisms of 5.1 per 10,000 pregnancies (live births, fetal deaths and terminations of pregnancy for medical reasons), with 1.8 per 10,000 live births and 3.2 per 10,000 terminations; in France, prenatal diagnoses were slightly more frequent (6.6 per 10,000), with fewer live births (1.2) and more terminations (5.3) [17, 27]. These values refer only to open dysraphisms and to a period different from that of the series by Khoshnood and colleagues (4.63 per 10,000 births for all spina bifida in 1991–2011), so the two estimates are not directly comparable [15, 17]. The same registry showed the highest prevalence in Ireland in 1980–1986 and, in 2004–2008, values higher than those of the other European countries on the island of Réunion and in Ukraine [17]. More than 8000 people with spinal dysraphism are estimated to be living in France, a figure that the authors consider difficult to measure and probably underestimated, because few studies include closed forms and late diagnoses are frequent [17]. France has not adopted mandatory fortification, and prevention relies on individual prescription by physicians and midwives; the PNDS cites the EUROCAT analysis of 1980–2012 (Morris and colleagues, 2018), with no reduction in prevalence in Europe, consistent with the 1991–2011 series [15, 17].

The 2021 perinatal PNDS places the total prevalence of dysraphisms at about 5.0 per 10,000 pregnancies in France and in Europe (about 9 per 10,000 for all neural tube defects), with a birth prevalence below 2 per 10,000 live births in 2012 because of the proportion of terminations; in the Paris registry (1992–2000), 90.1% of open dysraphisms had a prenatal diagnosis, with 85.1% terminations [18]. The regional data cited vary between registries and periods, and the evidence document notes that the term “spina bifida” in registries does not distinguish the types of dysraphism [18]. The 2015 PRM PNDS reproduces the EUROCAT series for 1980–2011 (spina bifida in all pregnancies: 18.46 per 10,000 in 1980, 5.57 in 2000, 4.57 in 2010 and 5.33 in 2011; live births: 14.36, 2.41, 1.89 and 1.72), explains the 1980 peak by the weight of the Irish registry and the initial decline by the entry of registries with lower prevalence, and notes higher prevalences on Réunion Island (about 1.7 per 1000 in 2004–2008) and in Ukraine [32]. Worldwide, the evidence document cites total prevalences below 1 per 10,000 in some Asian countries (Japan, Thailand, Vietnam) and of 19.7 per 10,000 in Turkey [18].

9.4. Latin America and Brazil (ECLAMC)

The Latin American Collaborative Study of Congenital Malformations (ECLAMC), a hospital network, analysed 3,347,559 births in 77 hospitals in Chile, Argentina and Brazil between 1982 and 2007 and concluded that, among 52 anomaly types, only the reduction in neural tube defects was attributable to folic acid fortification, with the greatest reduction in cephalic (cervicothoracic) spina bifida, followed by caudal (lumbosacral) spina bifida, anencephaly and cephalocele [123]. Before fortification, South American rates were approximately 18 per 10,000 female births and 12 per 10,000 male births (male/female ratio 0.67); in Chile and Argentina, the reduction after fortification was greater in females, particularly in anencephaly and cervicothoracic spina bifida, altering the sex ratio, while Venezuela, without fortification, served as a control [124]. A Latin American review of 1990–2010 found prevalences between 0.2 and 9.6 per 1000 live births, dependent on ascertainment method, and declines of 33% to 59% after fortification [125]. As hospital network data, ECLAMC series measure birth frequency in participating hospitals rather than national prevalence.

In Brazil, the genetic syndromes book reports prevalences of 1–5 per 1000 live births, varying by region [7]. In the population-based study of national information systems (Central-West, Southeast and South; 17,925,729 live births and 194,858 stillbirths, 2001–2014), total neural tube defect prevalence fell from 0.79 to 0.55 per 1000 births after mandatory fortification (−30.1%; prevalence ratio 1.43; 95% CI 1.38–1.50), from 0.57 to 0.44 per 1000 live births and from 17.74 to 11.70 per 1000 stillbirths; total anencephaly decreased by 38.3% and spina bifida by 23.8% [23]. Before fortification, children of adolescent mothers had a 34% higher risk and females a 40% higher risk, and reductions in these groups were 37.7% and 36.5% [23]. Between 2010 and 2019, 13,443 live births with neural tube defects were recorded in Brazil (52% spina bifida, 33% anencephaly and 15% encephalocele), and fortification is estimated to have prevented 3499 affected live births [80]. An earlier evaluation, comparing 2001–2004 with 2006–2010 data, had found a more modest reduction in liveborn spina bifida (prevalence ratio 0.937) and a greater reduction in fetal deaths from anencephaly (0.727), attributing the smaller impact to dietary habits, folic acid concentration in flour and ethnic composition [126]. Copp and colleagues’ statement that Brazil had recorded no reduction after fortification [8] was superseded by subsequently published population data [23].

9.5. Special populations and situations

Prevalence differences between ethnic groups may persist after migration [8]. Monozygotic twins have an excess of early defects, including anencephaly, with concordance of only 5% to 20% [75]. Obesity, more frequent in pregnancies with myelomeningocele, limits ultrasound detection and was a MOMS exclusion criterion (BMI ≥35) [12, 26]. In US Hispanic populations, lower folic acid intake prompted the proposal to fortify masa corn flour, estimated to prevent approximately 40 cases per year [22]. In fetal repair, the Latin American experience with laparotomy-assisted fetoscopy (Mexico and Colombia) was justified by the authors on the basis of high spina bifida prevalence and less favourable perinatal outcomes in the region [101].

9.6. Reference networks and patient associations

ERN ITHACA is the European reference network for rare malformation syndromes and intellectual disability, responsible for the European consensus on the fetal valproate spectrum [29, 34]. No European reference network specifically dedicated to isolated neural tube defects was identified in the sources consulted. The International Federation for Spina Bifida and Hydrocephalus (IFSBH) is the international federation of associations [30]; in Portugal, the Associação Spina Bífida e Hidrocefalia de Portugal (ASBIHP) [31]; in the United Kingdom, SHINE (Spina bifida, Hydrocephalus, Information, Networking, Equality) and, for fetal anticonvulsant syndrome, the FACS Association [10, 57].

In France, the spina bifida PNDSs (French National Diagnosis and Care Protocols) were prepared by the Centre de référence Spina Bifida – Dysraphismes of the Rennes University Hospital, a constituent site of the C-MAVEM rare disease reference centre (vertebral and spinal cord malformations) and part of the NeuroSphinx health network (filière), and reviewed with competence centres, the MAREP (anorectal and pelvic malformations) and syringomyelia reference centres, scientific societies and the patient associations ASBH (Association Spina Bifida et Handicaps associés), FFASB (Fédération française des associations du spina bifida) and, in French-speaking Belgium, ASBBF (Association Spina Bifida belge francophone) [17, 27]. The French series is organised by specialty: in addition to the PNDSs on adult urology (2020) and the bowel (2021, an update of the 2011 one), there is a PNDS on physical and rehabilitation medicine (2015) [17, 27]. The PNDS can serve as a reference for the attending physician, together with the specialist, for the care protocol of an application for exemption from the ticket modérateur as a condition outside the list of long-term conditions (ALD hors liste) and identifies products used off-label or not reimbursed: at the time, phosphodiesterase 5 inhibitors and cystatin C were not reimbursed, intracavernous alprostadil was reimbursed at 30%, sacral neuromodulation was outside the scope of reimbursement in spina bifida and the 300-unit dose of botulinum toxin was off-label [17, 27]. Both PNDSs stress the role of the general practitioner as first point of contact and close coordinator for the patient, the need for contact with the regional reference centre and referral to patient associations [17, 27]. For the perinatal period and for filum lipoma, the 2021 PNDSs were prepared by the C-MAVEM reference centre (vertebral and spinal cord malformations): the Armand-Trousseau Hospital site (AP-HP, Paris) follows patients from fetal life until transfer to adulthood, with fetal medicine, paediatric radiology, neurosurgery, orthopaedics and paediatric PRM, and the Necker-Enfants Malades Hospital site coordinates paediatric neurosurgery; the MAVEM network includes centres in Tours, Besançon, Lille, Marseille, Montpellier, Paris (Rothschild Hospital), Nancy and Toulouse [18] [28]. Every prenatal suspicion is discussed in a CPDPN [18]. Parents are directed to the ASBH association and to the eNorme platform, which puts parents of sick or disabled children in contact, and the social services support recognition of disability by the MDPH, the education allowance for a disabled child (AEEH) and its supplements, the daily parental presence allowance, specialised childcare arrangements and carer status [18]. The PRM PNDS was reviewed by the French societies of PRM, coloproctology, urodynamics, neurosurgery, gynaecology, paediatric surgery, neuro-urology, urology, paraplegia, paediatric and adult orthopaedics, paediatrics and paediatric neurology, and by the ASBH, FFASB and ASBBF associations [32].

10. Illustrative Clinical Case

Anghelina and colleagues published in 2025, in Romania, a case of severe myelomeningocele in the fourth pregnancy of a 29-year-old woman [82]. The woman had three previous normal pregnancies, a non-consanguineous marriage, no concomitant diseases or congenital anomalies in the family, low socioeconomic status, and had not taken folic acid before conception or during the first trimester; she had no prenatal care and no AFP measurement between 16 and 18 weeks [82]. She first presented at 28–29 weeks in poor general condition, with lower abdominal pain and transient lumbar pain; ultrasound showed a female fetus in breech presentation with lumbosacral myelomeningocele, and the team opted for caesarean delivery [82].

The newborn weighed 1100 g, measured 36 cm, had a head circumference of 27 cm and was born in a severe condition, with generalised cyanosis, intermittent apnoea, bradycardia at 36 beats per minute and an Apgar score of 1 at the first minute, progressing to cardiorespiratory arrest [82]. In the lumbosacral region there was a defect of approximately 8 × 8 cm, 1 cm deep, with exposed and herniated spinal cord and meninges without skin or meningeal covering (article figure 1: https://pmc.ncbi.nlm.nih.gov/articles/PMC12831564/figure/F1/); the upper limbs had active movements and the lower limbs were paralysed and areflexic [82]. Transfontanellar ultrasound revealed Chiari II malformation and marked hydrocephalus. Despite intubation, chest compressions, adrenaline and saline boluses, with severe mixed acidosis (pH 6.659) and lactate of 12.43 mmol/L, the child died shortly after birth [82]. Autopsy showed prematurity, moderate internal hydrocephalus, leukomalacia, bilateral pulmonary atelectasis and intracardiac thrombosis, without other malformations; fetal karyotyping was not performed [82].

The case illustrates several points of this article. First, a multifactorial defect may arise after three normal pregnancies without a family history. Second, absent periconceptional supplementation was the modifiable risk factor identified by the authors, although causation cannot be asserted in an individual case. Third, lack of prenatal care prevented AFP screening, second-trimester fetal anomaly ultrasound and consequently counselling about termination, postnatal repair or fetal repair before 26 weeks [26, 82]. Fourth, the combination of an extensive non-epithelialised lesion, Chiari II and hydrocephalus describes the severe end of the spectrum. The decision for caesarean delivery in this context was the team’s choice; the authors themselves note that several guidelines do not consider myelomeningocele a mandatory indication for caesarean delivery, which should not be read as a routine recommendation [82]. Future counselling of this family would apply the recurrence risk after one affected child and high-dose supplementation (section 8) [10, 20].

11. Prognosis and Long-Term Follow-up

11.1. Prognosis by defect type

Anencephaly and craniorachischisis are lethal [7, 10]. In encephalocele, the greater the cranial content in the herniated sac, the worse the prognosis, and there may be motor and sensory deficit [7]. In open spina bifida, prognosis depends on the spinal level and on the number and severity of associated anomalies; it is guarded at thoracic levels or in the presence of kyphosis, hydrocephalus, early hydronephrosis and associated anomalies, and better below the first lumbar vertebra without severe kyphoscoliosis [7]. The mortality rate of live-born infants with myelomeningocele was about 10% in the period described by MOMS [26]. Rodrigues and colleagues report that, with advances in prenatal imaging and neonatal care, 75% of children with spina bifida can reach adulthood [80]. In adults, urological diseases (end-stage renal failure, infections) remain the leading cause of death in a French cohort, accounting for 17% of deaths, which supports the urological surveillance in section 11.4 [17].

11.2. Long-term cohorts

Hunt and Oakeshott studied the complete cohort of 117 babies with open spina bifida whose defects were closed at Addenbrooke’s Hospital, Cambridge, between 1963 and 1971, without any attempt at selection, following complete neurological examination and with ventriculoatrial shunting when hydrocephalus required it [127]. At review in spring 2002, by confidential postal questionnaire supplemented by telephone contact with the patient or carer and Office for National Statistics death data, ascertainment was 100%: 63 (54%) had died, mainly the more severely affected, from cardiorespiratory (19) or renal failure (18), hydrocephalus (10), central nervous system infection (10), seizures (2), aspiration of vomit (2), sudden infant death (1) and thrombocytopenic purpura (1); the 54 survivors had a mean age of 35 years (32–38) and a male:female ratio of 1:1.3, the same as at birth [127]. Of the 54 survivors, 46 had CSF shunts, 39 had IQ ≥80 (recorded between 5 and 15 years), 16 could walk at least 50 m with or without aids, only 11 were fully continent without catheters or devices, 30 had experienced pressure ulcers and 30 were overweight [127]. Twenty-two lived independently in the community (seven of them wheelchair-dependent), 12 in sheltered accommodation with help available and 20 needed daily help, mainly from a parent (then aged 52–77 years), partner or social services; these 20 were all severely disabled, two were blind after shunt malfunction and two had respiratory support [127]. Twenty drove (another nine had stopped driving), 13 were in open employment (five of them wheelchair users), and seven women and two men had had a total of 13 children, none with visible spina bifida; only one-third of survivors were still followed in hospital, with most under general practitioner care [127].

The article’s figure stratifies the cohort by sensory level in infancy and shows that mortality and disability increased with the height of the deficit (χ² test for trend, excluding the five cases with an asymmetric level) [127]. The percentages cited by Oxford (71% with CSF shunts, 88% with IQ at least 80, 67% able to walk 50 m with aids if necessary, 33% continent, 58% living independently and driving, and 38% in open employment) correspond to the subgroup of 24 survivors with sensory level below L3 (17, 21, 16, 8, 14 and 14, and 9 of 24), rather than all 54 survivors, among whom the proportions are lower (for example, 16 of 54 ambulant and 11 of 54 continent) [10, 127].

Table source: figure 1 of Hunt and Oakeshott, 2003 (https://pmc.ncbi.nlm.nih.gov/articles/PMC162127/figure/fig1/), absolute numbers; survivor rows use survivors at each level as the denominator [127].

At assessment at 40 ± 3 years in the same cohort, 46 (39%) survived; of the 38 children with absent sensation only below the knee (level below L3), 61% survived and, of these, 61% were community ambulators and 22% needed daily care, whereas of the 42 with absent sensation up to the umbilicus (above T11) only 17% survived, none walked and 71% needed daily care; survivors without shunt revisions more frequently walked, lived independently and drove [128]. The authors attribute the breadth of outcomes, from apparent normality to severe disability, both to initial neurological deficit and shunt adverse events, and emphasise that more than one-third of survivors continued to need daily care [127]. These figures describe the course of a British cohort from the era of non-selective treatment, with neonatal closure and ventriculoatrial shunting, before fetal surgery and fortification; they serve for counselling as long-term data from a complete cohort, but are not an estimate of prognosis for a child treated today.

Following MOMS, children operated on in utero maintained, at 5–10 years, greater self-care competence (90.8% versus 85.5% of the maximum FRESNO scale score), a higher proportion of community ambulation (51.3% versus 23.1%) and a lower likelihood of a motor level worse than the anatomical level (adjusted RR 0.44; 0.25–0.77) [89]. The benefit of fetal surgery for bladder and bowel continence, sexual function and mental capacity required prolonged follow-up according to the MOMS authors [26].

11.3. Multidisciplinary follow-up

Follow-up should be coordinated by a physician trained in the care of children with spina bifida and include a coordinator responsible for follow-up, with communication between the team and the attending physician [8]. For the child with open dysraphism, the sources consulted do not set numerical intervals by specialty beyond the neonatal investigations, and frequency follows the protocols of multidisciplinary centres; for filum terminale lipoma and for the first months of life, the 2021 French PNDS (French National Diagnosis and Care Protocols) propose the investigations in Table 11.3a; for the adult, the French urology PNDS proposes the intervals in Table 11.4, and the bowel PNDS recommends screening for constipation and faecal incontinence at every visit [17, 27]. Hunt and Oakeshott point out that headache, neck pain, drowsiness, deteriorating vision or new ocular signs may indicate shunt failure and require prompt intervention to prevent serious long-term consequences, and that, in their cohort, most adults were under the care of their general practitioner, who managed incontinence, pressure ulcers, sepsis, epilepsy, urinary and respiratory infections, hypertension and obesity [127].

Table 11.3a. Surveillance in the child according to the 2015 and 2021 French PNDS.

The filum lipoma PNDS stresses that there are no specific surveillance recommendations and that urological schedules are based mainly on expert opinion [28]. The PRM PNDS points out that, in adulthood, premature locomotor ageing must be distinguished from the neurological complications specific to the disease (tethered cord, syringomyelia, hydrocephalus, Chiari) [32].

11.4. Urological and bowel surveillance in the adult

The French PNDS proposes the following surveillance algorithm, adapted from the 2007 GENULF, 2016 EAU and 2018 SIU-ICUD recommendations; surveillance should be regular, multidisciplinary and preferably provided by a specialised centre [17].

11.5. Risk of bladder cancer

The incidence of bladder cancer has traditionally been considered higher in patients with neurogenic bladder, with late diagnosis and locally advanced tumours, but recent series show an incidence close to that of the general population, while increasing life expectancy raises all cancers linked to ageing [17]. Indwelling catheterisation is a recognised risk factor and should only be used long term in the absence of an alternative (grade B), clean intermittent catheterisation being the reference voiding method (grade A) [17]. Small-bowel enterocystoplasty appears to increase the risk (estimated at 0–5.5%), and in 90% of cases the cancer arises more than 10 years after surgery; for gastrocystoplasty, the PNDS reports malignant transformation in 17–20% of cases in the medium and long term, but the series it cites (Marseille, 11 paediatric gastrocystoplasties performed between 1992 and 2000, four for myelomeningocele, with a median follow-up of 17 years) describes adenocarcinoma of the gastric patch in 3 of the 11 patients (27%), after a median of 20 years (11–22), two of them metastatic, and excision of the patch in 7 of the 11 [17, 129]. The small denominator does not allow a precise risk estimate, but it justifies endoscopic surveillance of all gastrocystoplasties. Any macroscopic haematuria requires complete evaluation, with upper tract imaging, cytology, cystoscopy and biopsy of lesions (grade C) [17]. Annual cystoscopy with cytology, proposed by the GENULF in 2007 for age over 50 years, neurogenic bladder of more than 15 years' duration or cystoplasty of more than 10 years, has no proof of efficacy (grade C), and the SIU no longer recommends systematic cystoscopy, even with an indwelling catheter (grade B), because the risk is lower than that of smokers (who are not screened), mucosal changes limit the reliability of the examination, there is no evidence of benefit in morbidity or mortality, and CT allows the bladder wall to be studied [17]. The PNDS recommends cystoscopy for haematuria, change in symptoms or recurrent symptomatic infections (grade D), from 10–15 years after an enterocystoplasty, and close surveillance after gastrocystoplasty, with excision of the gastric patch in the presence of atrophic gastritis [17].

12. Summary and Final Considerations

Neural tube defects are multifactorial malformations, with estimated heritability of 60–70% and an environmental contribution, in which fewer than 10% of cases are syndromic or chromosomal [8]. Open defects result from failure of closure during primary neurulation, between the third and fourth weeks after fertilisation, and closed defects from secondary neurulation abnormalities, with different mechanisms, clinical signs and responses to folic acid [8, 10]. In open spina bifida, neurological damage results from two insults, failure of closure and intrauterine neurodegeneration, and is associated with hydrocephalus and Chiari II malformation [8, 10].

Susceptibility genetics comprises rare low-penetrance variants in planar cell polarity and one-carbon metabolism genes and small-effect polymorphisms, such as MTHFR c.665C>T (historical 677C>T), whose association depends on population and folate status and whose genotyping has no clinical utility [11, 43, 44]. Counselling continues to rely on empirical risks: 2–5% after one affected child, approximately 10% after two and 4% for children of an affected parent, values that decrease with supplementation [8, 10].

Primary prevention with folic acid is the intervention with the greatest impact: 0.4–0.8 mg/day (United States) or 400 µg/day (WHO, Portugal) from before conception, and 4–5 mg/day in women with a previous affected pregnancy or other defined risk factors, with a 72% reduction in recurrence in the MRC trial [10, 19–21, 25]. Mandatory fortification reduced prevalence by 28% in the United States and approximately 30% in Brazil, whereas Europe, without mandatory fortification until recently, recorded no decline between 1991 and 2011; England introduces mandatory fortification on 13 December 2026 [15, 22–24]. Valproate, pregestational diabetes, obesity and hyperthermia are modifiable factors requiring preconception planning [10, 57, 95].

Prenatal diagnosis relies on ultrasound, with detection of 90–98% of myelomeningoceles in routine programmes, and serum AFP as an adjunct [8, 10]. Open fetal repair before 26 weeks, in experienced centres and eligible pregnant women, reduces shunt requirement and improves motor function into school age, at the cost of prematurity and maternal uterine risk [26, 88, 89]. Fetoscopic techniques and cell therapy are active research areas, with safety and feasibility results but no demonstration of superiority in randomised trials [96, 97]. Lifelong follow-up is multidisciplinary, and prognosis depends mainly on lesion level, hydrocephalus and shunt complications [8, 128].

In the adult, the French PNDS place the neurogenic bladder and bowel at the centre: urological diseases remain the leading cause of death, clean intermittent catheterisation with anticholinergics and, as second line, botulinum toxin precede reconstructive surgery, asymptomatic bacteriuria is not treated, and surveillance combines clinic review and ultrasound every 1–2 years, annual renal function and urodynamics on average every 2 years [17]. In the perinatal period, the 2021 French PNDS bases diagnosis on reference ultrasound and the CPDPN (multidisciplinary prenatal diagnosis centre), with amniocentesis for karyotype, microarray and acetylcholinesterase, without AFP or non-invasive screening, distinguishes LDM and Mie-LDM from myelomeningocele, does not recommend caesarean section without fetal surgery and proposes folic acid 5 mg/day after any dysraphism [18]. In filum terminale lipoma, section of the filum is indicated for symptoms of tethered cord, and prophylactic surgery remains without consensus [28]. Rehabilitation continues throughout life, with orthoses adjusted to the level, early education in catheterisation, planned transition and prevention of pressure ulcers [32]. For the bowel, PEG (grade A), abdominal massage, transanal irrigation (grade B), antegrade irrigation and sacral neuromodulation (grade C) are organised into an algorithm by predominant symptom, and correcting an isolated deficiency is not sufficient to improve quality of life [27].

13. Bibliographic References

1. Orphanet. Neural tube defect (ORPHA:3388) [Internet]. Paris: INSERM; version dated 2026 Jul 2; accessed 2026 Oct 6. Available at: https://www.orpha.net/en/disease/detail/3388
2. Orphanet. Isolated anencephaly/exencephaly (ORPHA:1048) [Internet]. Paris: INSERM; version dated 2026 Jul 2; accessed 2026 Oct 6. Available at: https://www.orpha.net/en/disease/detail/1048
3. Orphanet. Open spinal dysraphism with a myelomeningocele (ORPHA:93969) [Internet]. Paris: INSERM; version dated 2026 Jul 2; accessed 2026 Oct 6. Available at: https://www.orpha.net/en/disease/detail/93969
4. Orphanet. Spinal dysraphism with a posterior meningocele (ORPHA:268744) [Internet]. Paris: INSERM; version dated 2026 Jul 2; accessed 2026 Oct 6. Available at: https://www.orpha.net/en/disease/detail/268744
5. Orphanet. Lipomyelomeningocele (ORPHA:268835) [Internet]. Paris: INSERM; version dated 2026 Jul 2; accessed 2026 Oct 6. Available at: https://www.orpha.net/en/disease/detail/268835
6. GeneGuidance. Clinical curation of genes MTHFR, VANGL1, VANGL2, CELSR1, FZD6, PRICKLE1, GLDC, AMT, MTHFD1 and DACT1 and condition entries OMIM 182940 and 601634 [database]. Version dated 2026 Sep 26; accessed 2026 Oct 6. Primary references listed: Kibar 2009 (PMID 19319979), Merello 2015 (PMID 25208524), Lei 2010 (PMID 20558380), Shi 2012 (PMID 22610794), Bashford 2020 (PMID 32533132).
7. Banhos MEH, Barbosa MA, Ribeiro EM. Defeitos do fechamento do tubo neural (DFTN). In: Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 160–2.
8. Copp AJ, Adzick NS, Chitty LS, Fletcher JM, Holmbeck GN, Shaw GM. Spina bifida. Nat Rev Dis Primers. 2015;1:15007. doi:10.1038/nrdp.2015.7. PMID: 27189655.
9. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Multifactorial inheritance; Genetic counseling for defects that are a result of multifactorial inheritance; p. 947–9.
10. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Neural tube defects; p. 500–5.
11. Juriloff DM, Harris MJ. A consideration of the evidence that genetic defects in planar cell polarity contribute to the etiology of human neural tube defects. Birth Defects Res A Clin Mol Teratol. 2012;94(10):824–40. doi:10.1002/bdra.23079. PMID: 23024041.
12. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Maternal diabetes mellitus and diabetic embryopathy; p. 756–7.
13. Rasmussen SA, Chu SY, Kim SY, Schmid CH, Lau J. Maternal obesity and risk of neural tube defects: a metaanalysis. Am J Obstet Gynecol. 2008;198(6):611–9. doi:10.1016/j.ajog.2008.04.021. PMID: 18538144.
14. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Hyperthermia-induced spectrum of defects; p. 808–11.
15. Khoshnood B, Loane M, de Walle H, Arriola L, Addor MC, Barisic I, et al. Long term trends in prevalence of neural tube defects in Europe: population based study. BMJ. 2015;351:h5949. doi:10.1136/bmj.h5949. PMID: 26601850.
16. Stallings EB, Isenburg JL, Rutkowski RE, Kirby RS, Nembhard WN, Sandidge T, et al. National population-based estimates for major birth defects, 2016–2020. Birth Defects Res. 2024;116(1):e2301. doi:10.1002/bdr2.2301. PMID: 38277408.
17. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Dysraphisme spinal (Spina Bifida) – prise en charge urologique à l'âge adulte. Saint-Denis La Plaine: HAS; 2020 Mar 24 (text of March 2020, Centre de référence Spina Bifida – Dysraphismes, CHU de Rennes, constituent site of C-MAVEM). Available at: https://www.has-sante.fr/jcms/p_3166008
18. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Prise en charge des dysraphismes en période périnatale. Saint-Denis La Plaine: HAS; 2021 Dec 27 (text and scientific argument of September 2021, Centre de référence des malformations vertébrales et médullaires C-MAVEM, Armand-Trousseau Hospital, Paris). Available at: https://www.has-sante.fr/jcms/p_3293075
19. US Preventive Services Task Force; Barry MJ, Nicholson WK, Silverstein M, et al. Folic acid supplementation to prevent neural tube defects: US Preventive Services Task Force reaffirmation recommendation statement. JAMA. 2023;330(5):454–9. doi:10.1001/jama.2023.12876. PMID: 37526713.
20. Direção-Geral da Saúde. Programa Nacional para a Vigilância da Gravidez de Baixo Risco. Lisboa: DGS; 2015 Nov. Available at: https://www.dgs.pt/em-destaque/programa-nacional-para-a-vigilancia-da-gravidez-de-baixo-risco-pdf11.aspx
21. World Health Organization. Guideline: optimal serum and red blood cell folate concentrations in women of reproductive age for prevention of neural tube defects. Geneva: WHO; 2015. ISBN 9789241549042. Available at: https://www.who.int/publications/i/item/9789241549042
22. Williams J, Mai CT, Mulinare J, Isenburg J, Flood TJ, Ethen M, et al. Updated estimates of neural tube defects prevented by mandatory folic acid fortification — United States, 1995–2011. MMWR Morb Mortal Wkly Rep. 2015;64(1):1–5. PMID: 25590678. Available at: https://pmc.ncbi.nlm.nih.gov/articles/PMC4584791/
23. Santos LM, Lecca RC, Cortez-Escalante JJ, Sanchez MN, Rodrigues HG. Prevention of neural tube defects by the fortification of flour with folic acid: a population-based retrospective study in Brazil. Bull World Health Organ. 2016;94(1):22–9. doi:10.2471/BLT.14.151365. PMID: 26769993.
24. The Bread and Flour (Amendment) (England) Regulations 2024. UK Statutory Instruments 2024 No. 1162. Made 2024 Nov 13. Available at: https://www.legislation.gov.uk/uksi/2024/1162/made
25. MRC Vitamin Study Research Group. Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. Lancet. 1991;338(8760):131–7. doi:10.1016/0140-6736(91)90133-A. PMID: 1677062.
26. Adzick NS, Thom EA, Spong CY, Brock JW 3rd, Burrows PK, Johnson MP, et al. A randomized trial of prenatal versus postnatal repair of myelomeningocele. N Engl J Med. 2011;364(11):993–1004. doi:10.1056/NEJMoa1014379. PMID: 21306277.
27. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Spina Bifida – Dysraphismes – Gestion du handicap intestinal. Saint-Denis La Plaine: HAS; 2021 Dec 2 (text of June 2021, Centre de référence Spina Bifida – Dysraphismes, CHU de Rennes, constituent site of C-MAVEM). Available at: https://www.has-sante.fr/jcms/p_3299635
28. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Prise en charge des lipomes du filum terminal. Saint-Denis La Plaine: HAS; 2021 Dec 1 (text of November 2021, Centre de référence C-MAVEM, paediatric neurosurgery department, Necker-Enfants Malades Hospital, Paris). Available at: https://www.has-sante.fr/jcms/p_3301913
29. ERN ITHACA — European Reference Network on Rare Malformation Syndromes, Intellectual and Other Neurodevelopmental Disorders [Internet]. Accessed 2026 Oct 6. Available at: https://ern-ithaca.eu
30. International Federation for Spina Bifida and Hydrocephalus (IFSBH) [Internet]. Accessed 2026 Oct 6. Available at: https://www.ifglobal.org
31. Associação Spina Bífida e Hidrocefalia de Portugal (ASBIHP) [Internet]. Accessed 2026 Oct 6. Available at: https://www.asbihp.pt
32. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Prise en charge en médecine physique et de réadaptation du patient atteint de Spina Bifida. Saint-Denis La Plaine: HAS; 2015 May 7 (text of April 2015, Centre de référence maladies rares Spina Bifida, CHU de Rennes). Available at: https://www.has-sante.fr/jcms/c_2032137
33. Czeizel AE, Dudás I. Prevention of the first occurrence of neural-tube defects by periconceptional vitamin supplementation. N Engl J Med. 1992;327(26):1832–5. doi:10.1056/NEJM199212243272602. PMID: 1307234.
34. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Fetal valproate syndrome; p. 794–5.
35. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Meningomyelocele, anencephaly, iniencephaly sequences; p. 874–7.
36. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Occult spinal dysraphism sequence (tethered cord malformation sequence); p. 878–9.
37. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Complex inheritance; p. 14–5.
38. VariantValidator [Internet]. Version 4.0.1 (vvta_2025_02); GRCh38 query, MANE Select transcripts NM_005957.5:c.665C>T and NM_005957.5:c.1286A>C, on 2026 Oct 6, and NM_138959.3:c.715G>A, c.821G>A and c.983T>C and NM_005956.4:c.1958G>A, on 2026 Nov 1. Available at: https://variantvalidator.org
39. National Center for Biotechnology Information. ClinVar: VCV000003520 (NM_005957.5(MTHFR):c.665C>T) and VCV000003521 (NM_005957.5(MTHFR):c.1286A>C), accessed 2026 Oct 6; VCV000001346, VCV000001347 and VCV000001348 (NM_138959.3(VANGL1):c.715G>A, c.821G>A and c.983T>C), accessed 2026 Nov 1 [Internet]. Available at: https://www.ncbi.nlm.nih.gov/clinvar/variation/3520/
40. Kibar Z, Torban E, McDearmid JR, Reynolds A, Berghout J, Mathieu M, et al. Mutations in VANGL1 associated with neural-tube defects. N Engl J Med. 2007;356(14):1432–7. doi:10.1056/NEJMoa060651. PMID: 17409324.
41. Kibar Z, Salem S, Bosoi CM, Pauwels E, De Marco P, Merello E, et al. Contribution of VANGL2 mutations to isolated neural tube defects. Clin Genet. 2011;80(1):76–82. doi:10.1111/j.1399-0004.2010.01515.x. PMID: 20738329.
42. Umair M, Alfadhel M. Homocystinuria due to deficiency of N(5,10)-methylenetetrahydrofolate reductase activity. 2025 May 29 [revised 2025 Dec 4]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 40440437. Available at: https://www.ncbi.nlm.nih.gov/books/NBK615089/
43. Amorim MR, Lima MA, Castilla EE, Orioli IM. Non-Latin European descent could be a requirement for association of NTDs and MTHFR variant 677C>T: a meta-analysis. Am J Med Genet A. 2007;143A(15):1726–32. doi:10.1002/ajmg.a.31812. PMID: 17618486.
44. Hickey SE, Curry CJ, Toriello HV. ACMG Practice Guideline: lack of evidence for MTHFR polymorphism testing. Genet Med. 2013;15(2):153–6. doi:10.1038/gim.2012.165. PMID: 23288205.
45. Van Hove JLK, Coughlin C II, Swanson M, Hennermann JB. Nonketotic hyperglycinemia. 2002 Nov 14 [updated 2019 May 23]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301531. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1357/
46. Manoli I, Sloan JL, Venditti CP. Isolated methylmalonic acidemia. 2005 Aug 16 [updated 2022 Sep 8]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301409. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1231/
47. Morris AAM, Spiekerkoetter U. Disorders of mitochondrial fatty acid oxidation & riboflavin metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 12, section 12.2. doi:10.1007/978-3-662-63123-2_12.
48. Lemay P, Guyot MC, Tremblay É, Dionne-Laporte A, Spiegelman D, Henrion É, et al. Loss-of-function de novo mutations play an important role in severe human neural tube defects. J Med Genet. 2015;52(7):493–7. doi:10.1136/jmedgenet-2015-103027. PMID: 25805808.
49. Schutte BC, Saal HM, Goudy S, Leslie EJ. IRF6-related disorders. 2003 Oct 30 [updated 2021 Mar 4]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301581. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1407/
50. Milunsky JM. Waardenburg syndrome type I. 2001 Jul 30 [updated 2022 Oct 20]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301703. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1531/
51. Turnpenny PD, Sloman M, Dunwoodie S. Spondylocostal dysostosis, autosomal recessive. 2009 Aug 25 [updated 2023 Aug 17]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301771. Available at: https://www.ncbi.nlm.nih.gov/books/NBK8828/
52. Tekendo-Ngongang C, Muenke M, Kruszka P. Holoprosencephaly overview. 2000 Dec 27 [updated 2020 Mar 5; revised 2026 Jun 9]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301702. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1530/
53. van der Crabben SN, de Koning TJ. Serine deficiency disorders. 2023 Jun 22. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 37347880. Available at: https://www.ncbi.nlm.nih.gov/books/NBK592681/
54. Chen CP. Chromosomal abnormalities associated with neural tube defects (I): full aneuploidy. Taiwan J Obstet Gynecol. 2007;46(4):325–35. doi:10.1016/S1028-4559(08)60002-9. PMID: 18182338.
55. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Edwards syndrome (trisomy 18); p. 670–1.
56. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Meckel-Gruber syndrome; p. 246–7.
57. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Fetal anticonvulsant syndrome (FACS); p. 732–5.
58. AbbVie Inc. DEPAKOTE (divalproex sodium) delayed-release tablets — prescribing information. NDA018723. Revision effective 2026 Mar 31. U.S. Food and Drug Administration; accessed 2026 Oct 6. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=018723
59. Bell R, Glinianaia SV, Tennant PW, Bilous RW, Rankin J. Peri-conception hyperglycaemia and nephropathy are associated with risk of congenital anomaly in women with pre-existing diabetes: a population-based cohort study. Diabetologia. 2012;55(4):936–47. doi:10.1007/s00125-012-2455-y. PMID: 22314812.
60. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Drugs in pregnancy; p. 718–21.
61. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Minor congenital anomalies; p. 240–1.
62. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Hypertrichosis; p. 178–81.
63. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Ventriculomegaly; p. 802–4.
64. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Imaging in prenatal diagnosis; p. 742–3.
65. Boyd PA, Wellesley DG, De Walle HE, Tenconi R, Garcia-Minaur S, Zandwijken GR, et al. Evaluation of the prenatal diagnosis of neural tube defects by fetal ultrasonographic examination in different centres across Europe. J Med Screen. 2000;7(4):169–74. doi:10.1136/jms.7.4.169. PMID: 11202581.
66. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Prenatal diagnosis; p. 950–4.
67. Viana LBOP, Ribeiro EM. Malformações do SNC. In: Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 289–90.
68. Bashford MT, Hickey SE, Curry CJ, Toriello HV; ACMG Professional Practice and Guidelines Committee. Addendum: ACMG Practice Guideline: lack of evidence for MTHFR polymorphism testing. Genet Med. 2020;22(12):2125. doi:10.1038/s41436-020-0843-0. PMID: 32533132.
69. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Caudal dysplasia sequence (caudal regression syndrome); p. 896–7.
70. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Exstrophy of cloaca sequence (OEIS complex); p. 888–9.
71. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Jarcho-Levin syndrome; p. 848–51.
72. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Acrocallosal syndrome; p. 308–9.
73. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Waardenburg syndrome; p. 358–61.
74. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Hydrocephalus; p. 172–5.
75. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Monozygotic twinning and structural defects; p. 914–21.
76. World Health Organization. e-Library of Evidence for Nutrition Actions (eLENA). Effects and safety of periconceptional oral folate supplementation for preventing birth defects: systematic review summary (De-Regil LM et al., Cochrane 2015) [Internet]. Accessed 2026 Oct 6. Available at: https://www.who.int/tools/elena/review-summaries/folate-periconceptional--effects-and-safety-of-periconceptional-oral-folate-supplementation-for-preventing-birth-defects
77. Viswanathan M, Urrutia RP, Hudson KN, Middleton JC, Kahwati LC. Folic acid supplementation to prevent neural tube defects: updated evidence report and systematic review for the US Preventive Services Task Force. JAMA. 2023;330(5):460–6. doi:10.1001/jama.2023.9864. PMID: 37526714.
78. Greene ND, Leung KY, Gay V, Burren K, Mills K, Chitty LS, et al. Inositol for the prevention of neural tube defects: a pilot randomised controlled trial. Br J Nutr. 2016;115(6):974–83. doi:10.1017/S0007114515005322. PMID: 26847388.
79. Wald NJ, Vale SH, Bestwick JP, Morris JK. Blood folate level needed for fully effective fortification in the prevention of neural tube defects. Arch Dis Child. 2025;110(8):651–6. doi:10.1136/archdischild-2024-328115. PMID: 39939144.
80. Rodrigues VB, Silva END, Dos Santos AM, Santos LMP. Prevented cases of neural tube defects and cost savings after folic acid fortification of flour in Brazil. PLoS One. 2023;18(2):e0281077. doi:10.1371/journal.pone.0281077. PMID: 36812183.
81. Kirolos A, Vieira R, Calvert C, Griffiths E, Wood R. Trends in neural tube defects in Scotland in 2000–2021 prior to the introduction of mandatory folic acid fortification of non-wholemeal wheat flour: a population-based study. Arch Dis Child. 2025;110(11):878–84. doi:10.1136/archdischild-2024-328295. PMID: 40097151.
82. Anghelina L, Şerbănescu MS, Gheonea C, Anghelina AM, Siminel MA, Camen AR, et al. Severe myelomeningocele in the fourth pregnancy of a 29-year-old woman: a case report. Rom J Morphol Embryol. 2025;66(3):587–92. doi:10.47162/RJME.66.3.16. PMID: 41382955.
83. U.S. Food and Drug Administration. Oxybutynin chloride tablets USP, 5 mg — prescribing information. ANDA209025. Revision effective 2025 Nov 18; accessed 2026 Nov 6. Available at: https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=oxybutynin+chloride+tablets
84. Lee JH, Kim KR, Lee YS, Han SW, Kim KS, Song SH, et al. Efficacy, tolerability, and safety of oxybutynin chloride in pediatric neurogenic bladder with spinal dysraphism: a retrospective, multicenter, observational study. Korean J Urol. 2014;55(12):828–33. doi:10.4111/kju.2014.55.12.828. PMID: 25512818.
85. Tabbers MM, DiLorenzo C, Berger MY, Faure C, Langendam MW, Nurko S, et al. Evaluation and treatment of functional constipation in infants and children: evidence-based recommendations from ESPGHAN and NASPGHAN. J Pediatr Gastroenterol Nutr. 2014;58(2):258-274. doi:10.1097/MPG.0000000000000266. PMID: 24345831.
86. INFARMED. Casenlax 4 g pó para solução oral em saqueta (macrogol 4000): resumo das características do medicamento. Casen Recordati. Approved 2018 Apr 13. Infomed, med_id 53814 [accessed 2026 Nov 7]. Available at: https://extranet.infarmed.pt/INFOMED-fo/download-ficheiro.xhtml?med_id=53814&tipo_doc=rcm
87. INFARMED. Duphalac 667 mg/ml xarope (lactulose): resumo das características do medicamento. Cooper Consumer Health. Approved 2023 Aug 7. Infomed, med_id 2877 [accessed 2026 Nov 7]. Available at: https://extranet.infarmed.pt/INFOMED-fo/download-ficheiro.xhtml?med_id=2877&tipo_doc=rcm
88. Farmer DL, Thom EA, Brock JW 3rd, Burrows PK, Johnson MP, Howell LJ, et al. The Management of Myelomeningocele Study: full cohort 30-month pediatric outcomes. Am J Obstet Gynecol. 2018;218(2):256.e1–256.e13. doi:10.1016/j.ajog.2017.12.001. PMID: 29246577.
89. Houtrow AJ, MacPherson C, Jackson-Coty J, Rivera M, Flynn L, Burrows PK, et al. Prenatal repair and physical functioning among children with myelomeningocele: a secondary analysis of a randomized clinical trial. JAMA Pediatr. 2021;175(4):e205674. doi:10.1001/jamapediatrics.2020.5674. PMID: 33555337.
90. Houtrow AJ, Thom EA, Fletcher JM, Burrows PK, Adzick NS, Thomas NH, et al. Prenatal repair of myelomeningocele and school-age functional outcomes. Pediatrics. 2020;145(2):e20191544. doi:10.1542/peds.2019-1544. PMID: 31980545.
91. Goodnight WH, Bahtiyar O, Bennett KA, Emery SP, Lillegard JB, Fisher A, et al. Subsequent pregnancy outcomes after open maternal-fetal surgery for myelomeningocele. Am J Obstet Gynecol. 2019;220(5):494.e1–494.e7. doi:10.1016/j.ajog.2019.03.008. PMID: 30885769.
92. Belfort MA, Whitehead WE, Shamshirsaz AA, Espinoza J, Nassr AA, Lee TC, et al. Comparison of two fetoscopic open neural tube defect repair techniques: single- vs three-layer closure. Ultrasound Obstet Gynecol. 2020;56(4):532–40. doi:10.1002/uog.21915. PMID: 31709658.
93. Diehl D, Belke F, Kohl T, Axt-Fliedner R, Degenhardt J, Khaleeva A, et al. Fully percutaneous fetoscopic repair of myelomeningocele: 30-month follow-up data. Ultrasound Obstet Gynecol. 2021;57(1):113–8. doi:10.1002/uog.22116. PMID: 32510722.
94. Kabagambe SK, Jensen GW, Chen YJ, Vanover MA, Farmer DL. Fetal surgery for myelomeningocele: a systematic review and meta-analysis of outcomes in fetoscopic versus open repair. Fetal Diagn Ther. 2018;43(3):161–74. doi:10.1159/000479505. PMID: 28910784.
95. European Medicines Agency. Valproate and related substances — referral (Article 31): CMDh position and European Commission final decision, 2018 [Internet]. Accessed 2026 Oct 6. Available at: https://www.ema.europa.eu/en/medicines/human/referrals/valproate-related-substances-0
96. Papanna R, Fletcher SA, Mann LK, Bergh EP, Austin MT, Zhu S, et al. Feasibility of fetoscopic spina bifida repair using a human umbilical cord meningeal patch: a nonrandomized clinical trial. Am J Obstet Gynecol. 2026. doi:10.1016/j.ajog.2026.08.011. PMID: 42600674.
97. Farmer DL, Kumar P, Reynolds E, Lee SY, Powne AB, Pivetti CD, et al. Feasibility and safety of cellular therapy for in-utero repair of myelomeningocele (CuRe Trial): a first-in-human, phase 1, single-arm study. Lancet. 2026;407(10531):867–75. doi:10.1016/S0140-6736(25)02466-3. PMID: 41763744.
98. ClinicalTrials.gov. NCT04652908 (Cellular Therapy for In Utero Repair of Myelomeningocele — The CuRe Trial), NCT04243889 (Fetoscopic NEOX Cord 1K Spina Bifida Repair) and NCT00060606 (Management of Myelomeningocele Study) [Internet]. Bethesda (MD): National Library of Medicine; accessed 2026 Oct 6. Available at: https://clinicaltrials.gov/study/NCT04652908
99. Gielchinsky Y, Alter R, Idelson A, Tenenbaum-Gavish K, Duvdevani NR, Jaffe A, et al. Totally percutaneous fetoscopic repair of open spina bifida: a single-center experience with perinatal, obstetric, and early functional outcomes. J Clin Med. 2026;15(11):4088. doi:10.3390/jcm15114088. PMID: 42278950.
100. Kunpalin Y, Sahakyan Y, Sander B, Snelgrove JW, Raghuram K, Kulkarni AV, et al. Comparison of open fetal, fetoscopic and postnatal surgical repair for open spina bifida: decision analysis. Ultrasound Obstet Gynecol. 2025;66(4):433–42. doi:10.1002/uog.29289. PMID: 40741643.
101. Miranda J, Parra-Saavedra MA, Contreras-Lopez WO, Abello C, Parra G, Hernandez J, et al. Implementation of in utero laparotomy-assisted fetoscopic spina bifida repair in two centers in Latin America: rationale for this approach in this region. AJOG Glob Rep. 2025;5(1):100442. doi:10.1016/j.xagr.2025.100442. PMID: 40027474.
102. U.S. Food and Drug Administration. Trospium chloride tablets, 20 mg — prescribing information. ANDA215781. Revision effective 2024 Dec 19; accessed 2026 Nov 6. Available at: https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=trospium+chloride
103. Pfizer. TOVIAZ (fesoterodine fumarate) extended-release tablets — prescribing information. NDA022030. Revision effective 2026 Apr 7. U.S. Food and Drug Administration; accessed 2026 Nov 6. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=022030
104. Allergan. BOTOX (onabotulinumtoxinA) for injection — prescribing information. BLA103000. Revision effective 2023 Nov 18. U.S. Food and Drug Administration; accessed 2026 Nov 6. Available at: https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=botox
105. U.S. Food and Drug Administration. Fosfomycin tromethamine granules for oral solution — prescribing information. ANDA214554. Revision effective 2025 Sep 25; accessed 2026 Nov 6. Available at: https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=fosfomycin+tromethamine
106. Dinh A, Hallouin-Bernard MC, Davido B, Lemaignen A, Bouchand F, Duran C, et al. Weekly sequential antibioprophylaxis for recurrent urinary tract infections among patients with neurogenic bladder: a randomized controlled trial. Clin Infect Dis. 2020;71(12):3128–35. doi:10.1093/cid/ciz1207. PMID: 31867616.
107. Pfizer. VIAGRA (sildenafil citrate) tablets — prescribing information. NDA020895. Revision effective 2017 Dec 15. U.S. Food and Drug Administration; accessed 2026 Nov 6. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=020895
108. Pfizer. CAVERJECT (alprostadil) injection — prescribing information. NDA020379. Revision effective 2024 May 13. U.S. Food and Drug Administration; accessed 2026 Nov 6. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=020379
109. U.S. Food and Drug Administration. Midodrine hydrochloride tablets — prescribing information. ANDA077746. Revision effective 2026 Jun 24; accessed 2026 Nov 6. Available at: https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=midodrine+hydrochloride
110. Rendeli C, Ausili E, Tabacco F, Focarelli B, Pantanella A, Di Rocco C, et al. Polyethylene glycol 4000 vs. lactulose for the treatment of neurogenic constipation in myelomeningocele children: a randomized-controlled clinical trial. Aliment Pharmacol Ther. 2006;23(8):1259–65. doi:10.1111/j.1365-2036.2006.02872.x. PMID: 16611288.
111. Krogh K, Jensen MB, Gandrup P, Laurberg S, Nilsson J, Kerstens R, et al. Efficacy and tolerability of prucalopride in patients with constipation due to spinal cord injury. Scand J Gastroenterol. 2002;37(4):431–6. doi:10.1080/003655202317316060. PMID: 11989834.
112. U.S. Food and Drug Administration. MiraLAX (polyethylene glycol 3350) powder for oral solution — OTC drug facts label. ANDA090685. Revision effective 2025 Dec 4; accessed 2026 Nov 6. Available at: https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=miralax
113. Takeda. MOTEGRITY (prucalopride) tablets — prescribing information. NDA210166. Revision effective 2025 Jul 17. U.S. Food and Drug Administration; accessed 2026 Nov 6. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=210166
114. U.S. Food and Drug Administration. MOVANTIK (naloxegol) tablets — prescribing information. NDA204760. Revision effective 2026 Sep 1; accessed 2026 Nov 6. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=204760
115. U.S. Food and Drug Administration. FDA Drug Safety Communication: FDA restricts use of prescription codeine pain and cough medicines and tramadol pain medicines in children; recommends against use in breastfeeding women. 2017 Apr 20 [accessed 7 Nov 2026, archived copy]. Available at: https://web.archive.org/web/2023/https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-restricts-use-prescription-codeine-pain-and-cough-medicines-and
116. Unichem Pharmaceuticals (USA). Tramadol hydrochloride tablets — prescribing information, section 4 (Contraindications) (DailyMed, setid 007bf37f-0e46-426a-ac8c-be63d4b7414c). Revision effective 10 April 2026. Available at: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=007bf37f-0e46-426a-ac8c-be63d4b7414c
117. Agence nationale de sécurité du médicament et des produits de santé (ANSM). CONTRAMAL 100 mg/ml, solution buvable en gouttes (chlorhydrate de tramadol): résumé des caractéristiques du produit. Base de données publique des médicaments, CIS 60477650; mise à jour le 30/10/2025 [accessed 2026 Nov 7]. Available at: https://base-donnees-publique.medicaments.gouv.fr/medicament/60477650/extrait
118. Pfizer. LYRICA (pregabalin) capsules and oral solution — prescribing information. NDA021446. Revision effective 2020 Jun 15. U.S. Food and Drug Administration; accessed 2026 Nov 7 (openFDA). Available at: https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=lyrica
119. Pfizer. NEURONTIN (gabapentin) capsules, tablets and oral solution — prescribing information. NDA020235. Revision effective 2026 May 12. U.S. Food and Drug Administration; accessed 2026 Nov 7 (openFDA). Available at: https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=neurontin
120. Eli Lilly and Company. CYMBALTA (duloxetine delayed-release capsules), for oral use: prescribing information. NDA021427. Revision 2023 Aug [accessed 2026 Nov 7]. Available at: https://pi.lilly.com/us/cymbalta-pi.pdf
121. European Medicines Agency. Cymbalta (duloxetine) 30 mg and 60 mg hard gastro-resistant capsules: summary of product characteristics. EMEA/H/C/000572. Amsterdam: EMA; initial authorisation 2004 Dec 17 [accessed 2026 Nov 7]. Available at: https://www.ema.europa.eu/en/documents/product-information/cymbalta-epar-product-information_en.pdf
122. Centers for Disease Control and Prevention. Spina bifida data and statistics [Internet]. Atlanta (GA): CDC; updated 2026 Sep 4, revised 2026 Sep 10; accessed 2026 Oct 6. Available at: https://www.cdc.gov/spina-bifida/data/index.html
123. López-Camelo JS, Castilla EE, Orioli IM; INAGEMP; ECLAMC. Folic acid flour fortification: impact on the frequencies of 52 congenital anomaly types in three South American countries. Am J Med Genet A. 2010;152A(10):2444–58. doi:10.1002/ajmg.a.33479. PMID: 20814949.
124. Poletta FA, Rittler M, Saleme C, Campaña H, Gili JA, Pawluk MS, et al. Neural tube defects: sex ratio changes after fortification with folic acid. PLoS One. 2018;13(3):e0193127. doi:10.1371/journal.pone.0193127. PMID: 29538416.
125. Rosenthal J, Casas J, Taren D, Alverson CJ, Flores A, Frias J. Neural tube defects in Latin America and the impact of fortification: a literature review. Public Health Nutr. 2014;17(3):537–50. doi:10.1017/S1368980013000256. PMID: 23464652.
126. Schuler-Faccini L, Sanseverino MT, de Rocha Azevedo LM, Moorthie S, Alberg C, Chowdhury S, et al. Health needs assessment for congenital anomalies in middle-income countries: examining the case for neural tube defects in Brazil. J Community Genet. 2014;5(2):147–55. doi:10.1007/s12687-013-0165-x. PMID: 23990401.
127. Hunt GM, Oakeshott P. Outcome in people with open spina bifida at age 35: prospective community based cohort study. BMJ. 2003;326(7403):1365–6. doi:10.1136/bmj.326.7403.1365. PMID: 12816823.
128. Oakeshott P, Hunt GM, Poulton A, Reid F. Open spina bifida: birth findings predict long-term outcome. Arch Dis Child. 2012;97(5):474–6. doi:10.1136/archdischild-2011-300624. PMID: 22121146.
129. Boissier R, Di Crocco E, Faure A, Hery G, Delaporte V, Lechevallier E, et al. What is the outcome of paediatric gastrocystoplasty when the patients reach adulthood? BJU Int. 2016;118(6):980–6. doi:10.1111/bju.13558. PMID: 27322857.