The clinical spectrum of nodular heterotopias in children: Report of 31 patients

Authors


Address correspondence to Lionel Carmant, Epilepsy Clinic and Epilepsy Research Group, Centre Hospitalier Universitaire Sainte-Justine, Université de Montréal, 3175 Côte-Sainte-Catherine, Montreal, QC H3T 1C5, Canada. E-mail: lionel.carmant@umontreal.ca

Summary

Purpose: The phenotypic and etiologic spectrum in adults with nodular heterotopias (NHs) has been well characterized. However, there are no large pediatric case series. We, therefore, wanted to review the clinical features of NHs in our population.

Methods: Hospital records of 31 patients with pathology or imaging-confirmed NHs were reviewed. Two-sided Fisher's exact t-test was used to assess associations between distribution of NHs and specific clinical features.

Key Findings: NHs were distributed as follows: 8 (26%) unilateral focal subependymal, 3 (10%) unilateral diffuse subependymal, 5 (16%) bilateral focal subependymal, 12 (39%) bilateral diffuse subependymal, and 3 (10%) isolated subcortical. The phenotypic spectrum in our population differs from that described in adults. Significant morbidity and mortality are associated with presentation in childhood. Twenty-two of 31 patients (71%) died in the neonatal period or in childhood. Additional cerebral malformations were found in 80% and systemic malformations in 74%. The majority of patients had developmental delay, intellectual deficit, and intractable epilepsy. Patients with unilateral focal NHs were more likely to have ventriculomegaly (p = 0.027), and those with bilateral diffuse NHs more likely to have cerebellar abnormalities (p = 0.007). Isolated subcortical NHs were associated with multiple malformations (p = 0.049) and cardiac abnormalities (p = 0.027). Underlying etiology was heterogeneous and determined in only six cases (19%): del chr 1p36, del chr 15q11, pyruvate dehydrogenase deficiency, sialic acidosis type 1, Aicardi syndrome, and FLNA mutation.

Significance: NHs are present in childhood as part of multiple cerebral and systemic malformations; developmental delay and refractory seizures are the rule rather than the exception. Milder forms go unrecognized until seizure onset in adulthood.

Heterotopias are malformations of cortical development characterized by the presence of normal brain cells in abnormal positions. Nodular heterotopias (NHs) are round masses of normal neurons and glial cells lacking laminar organization (Guerrini & Carrozzo, 2001). They can be classified into three groups, based on their location: periventricular, subcortical, and leptomeningeal. Periventricular nodular heterotopias (PNHs) are by far the most common (Guerrini & Parrini, 2010). They can range from isolated single nodules to bilateral confluent nodules. Although their prevalence in the general population is unknown, it is reported that NHs are found in 13–20% of epileptic patients with cortical malformations (Dubeau et al., 1995; Raymond et al., 1995; Battaglia et al., 2006). NHs have a number of identified genetic causes. Defects in the filamin A gene (FLNA; chromosomal locus Xq28), which encodes the F-actin–binding cytoplasmic cross-linking phosphoprotein filamin A, result in an X-linked form of bilateral, symmetrical periventricular heterotopias (PNHs; OMIM 300049) (Fox et al., 1998), seen almost exclusively in female patients and accounting for 100% of familial cases and 26% of isolated female cases (Parrini et al., 2006). Mutations in male patients are associated with high rates of fetal lethality. Mutations in the ARFGEF2 gene have been identified as responsible for an autosomal recessive form of bilateral nodular heterotopias (OMIM *605371) in two consanguineous families with microcephaly, severe delay, and early onset seizures (Sheen et al., 2003a,b; Sheen et al., 2004; de Wit et al., 2009). Bilateral PNHs are also associated with several chromosomal copy number variations such as 1p36 deletion (Neal et al., 2006; Saito et al., 2008), 7q11.23 deletion (Ferland et al., 2006) and 5pter duplication (Sheen et al., 2003b; Cardoso et al., 2009).

There have been several large case series of patients with PNHs (Dubeau et al., 1995; Parrini et al., 2006). However, this is the first large pediatric case series of PNHs that also includes patients who did not survive infancy. It remains unclear whether the imaging characteristics, clinical features, associated systemic malformations, and genetic etiology parallel the findings in adults. The aim of this study is to better delineate the clinical spectrum of NHs in a large pediatric population.

Methods

Patients were ascertained from the St. Justine Hospital, a tertiary care pediatric hospital (n = 31). We retrieved from medical records the charts of all children seen between 1999 and 2005 with either pathologically or radiologically confirmed subcortical or periventricular heterotopias (with or without other associated abnormalities). For a diagnosis of nodular heterotopia, the lesions had to have the signal intensity equal to that of gray matter on all pulse sequences. Patients with NHs surrounding a porencephalic cyst associated with polymicrogyria were excluded (n = 5) because the underlying pathophysiology was believed to be nongenetically based. Clinical histories were reviewed and the following information was systematically collected from the medical charts: prenatal history, perinatal complication, associated cerebral and systemic malformations, head circumference, family history of epilepsy or cerebral malformation, and developmental outcome. In addition, presence of seizures was recorded including seizure type, age of onset, location of epileptic discharges on electroencephalography (EEG) and complete seizure control by medication. Specifically, presence of a correlation between EEG pattern and location of the NH was sought.

Patients were subdivided into five categories according to the distribution of the heterotopias. The first four categories classified PNHs as: (1) unilateral diffuse (presence of nodular heterotopias all along the lateral ventricles); (2) unilateral focal (presence of single or multiple nodular heterotopias in a restricted distribution along the lateral ventricular walls); (3) bilateral diffuse; and (4) bilateral focal. The fifth category consisted of patients with subcortical heterotopias in the absence of periventricular heterotopias. We did not retrieve a single patient with leptomeningeal NHs.

Two-sided Fisher’s exact test was used to determine whether the presence of clinical features including gender, prenatal factors, neonatal and childhood death, developmental delay, seizure characteristics, and presence of cerebral and systemic malformations were correlated with the type and distribution of the heterotopias. A p ≤ 0.05 was used for establishing statistical significance.

Results

Clinical features

A total of 31 patients with cerebral heterotopias were included in our series: 19 were identified through imaging, and the remaining 12 were identified through pathology reports. Clinical characteristics of individual patients are presented in Table 1. A summary of the clinical characteristics based on distribution of the NHs are presented in Table 1. There was a clear female predominance, with 24 female and seven male patients affected (3.4:1). The ages at the time of last follow-up ranged from newborn to 25 years. Mortality was high: 14 patients died in the neonatal period (10 at <1 week, 4 between 1 and 4 weeks). Eight other patients died in childhood, and only 9 patients (29%) are currently living. The main presenting symptoms in this cohort were central nervous system (CNS) malformation (e.g., myelomeningocele, macrocephaly) or multiple congenital malformations in 18 of 31 (56%), seizures in 6 of 31 (19%), and global developmental delay in 4 of 31 (13%). Approximately one-half of the deaths were CNS related (11/23, 48%) and occurred in the neonatal period. Other causes of death included cardiorespiratory (5 of 23, 22%), metabolic decompensation related to underlying diagnosis (2 of 23, 9%), and accidental drowning (1 of 23, 4%) and sudden death (1/23, 4%). Thirteen of the 16 patients (81%) who survived the neonatal period had moderate to severe global developmental delay, and two others had learning difficulties at school and were in special classes (two with borderline-mild intellectual deficit on Wechsler Intelligence Scale for Children [WISC-III]). The three normal or mildly delayed patients had a varied distribution of their heterotopias: one patient had unilateral diffuse PNHs with extension from the subcortical region to the overlying cortex; another had bilateral occipital nodules and one large subcortical temporal heterotopia; and the third had bilateral occipital heterotopias without subcortical nodules.

Table 1.   Clinical features of patients with nodular heterotopias
Patient/sex/age (age at death)Heterotopia categoryCerebral findingsSystemic malformationsDevelopmental
delay or MR
Genetic findingsAge/Symptom at presentationCause of death
  1. ASD, atrial septal defect; B, bilateral; CC, corpus callosum; CNS, central nervous system; diff, diffuse, F, female; het, heterotopia; L, left; M, male; MR, mental retardation; n, nucleus; N/A, not applicable; ND, not done; PDA, patent ductus arteriosus; PDH, pyruvate dehydrogenase; PNH, periventricular nodular heterotopia; R, right; se, subependymal; U, unilateral; VSD, ventricular septal defect; d, day; mo, months; y, years.

1/F/(d 23)U, focalUnilateral, single, L frontal PNH
Ventriculomegaly, cerebellar abnormalities
MyelomeningoceleN/ANDBirth/myelomeningocele and ventriculomegalyCNS
2/M/(d 7)U, focalUnilateral, single R occipital PNH
Ventriculomegaly, cerebellar cortex dysplasia, Chiari II
MyelomeningoceleN/ANDBirth/myelomeningocele and ventriculomegalyCNS
3/F/(stillbirth)U, focalUnilateral, R frontal PNH
Ventriculomegaly, Chiari II
MyelomeningoceleN/ANDBirth/myelomeningocele and ventriculomegalyCNS
4/F/(stillbirth)U, focalUnilateral, single PNH
Ventriculomegaly, iniencephaly
Diaphragmatic hernia, pulmonary and renal hypoplasiaN/ANDBirth/multiple congenital ventriculomegalyCNS
5/F/(d 2)U, focalUnilateral, R occipital PNH
Frontal encephalocele, ventriculomegaly
NoneN/ANDBirth/encephalocele +ventriculomegalyCNS
6/F/(10 mo)U, focalUnilateral R parietal PNH
CC hypoplasia
ASD, pulmonary abnormalityYesKaryotype normal19 mo/respiratory distress and global developmental delayCardiorespiratory
7/F/(13 mo)U, focal
subcortical het
Unilateral, R frontal PNH and white matter cerebellar heterotopia
CC hypoplasia
Sialidosis type IYesSialidosis type I
Karyotype normal
3 mo/global developmental delay and hepatosplenomegalySialidosis type I
8/F/9 yU, focalUnilateral, single, R frontal PNHNoneYesKaryotype and FLNA
testing normal
2 y/seizuresNot applicable
9/F/(15 y)U, diff
subcortical het
Unilateral, diffuse (sparing frontal) L PNH
Contiguous subcortical heterotopias to cortex, overlying polymicrogyria, hippocampal dysplasia, bilateral mesiotemporal sclerosis
VSD, ASDNo- but learning difficultiesND15 y/sudden deathUnknown
10/F/17 yU, diff
subcortical het
Unilateral, diffuse L PNH
Contiguous subcortical heterotopias to L parietal cortex, L hemisphere atrophy
NoneYesND2 mo/seizuresNot applicable
11/F/14 yU, diffUnilateral, diffuse L PNH
Agenesis of corpus callosum
Craniosynostosis, trigonocephaly, vertebral agenesis, bilateral calcaneus valgus feet, dysmorphicYesNDBirth/multiple congenital malformationsNot applicable
12/F/(d 6)B, focalBilateral, occipital PNH
Ventriculomegaly, Chiari II
MyelomeningoceleN/ANDBirth/myelomeningocele and ventriculomegalyCNS
13/F/(d 1)B, focal
subcortical het
R temporal PNH
Bilateral subcortical perisylvian nodules, abnormal claustrum, polymicrogyria
Arthrogryposis, intestinal atresia, ASD, PDA, dysmorphismsN/ADeletion chr15q11
on karyotype and FISH
Birth/multiple congenital malformationsCardiorespiratory
14/F/25 yB, focal
subcortical het
Bilateral occipital heterotopias
Large R subcortical temporal heterotopia, partial agenesis of CC
NoneYesND4 y/seizuresNot applicable
15/M/18 yB, focalBilateral, single L frontal and single R frontal heterotopiasCryptorchidismYesKaryotype, subtelomeres and
FLNA testing normal
17 mo/global developmental delayNot applicable
16/F/(5 y)Subcortical onlyBilateral subcortical scattered nodules (in cerebrum and cerebellum), dysplasia of inf. Olivary n and dentate nASD, scoliosis, hip subluxation, clubfeet, turicephaly, dysmorphicYesKaryotype normalBirth/multiple congenital malformationsCardiorespiratory
17/M/14 yB, focalBilateral, occipital PNHNoneNo, but borderline IQKaryotype and FLNA
testing normal
8 y/seizuresNot applicable
18/M/(neonatal)B, diffBilateral, diffuse PNH
Ventriculomegaly, Chiari II
MyelomeningoceleN/ANDBirth/myelomeningocele and ventriculomegalyCNS
19/F/(d 3)B, diffBilateral, diffuse PNH
Ventriculomegaly, Chiari II
NoneN/AKaryotype normalBirth/macrocephalyCNS
20/F/(d 4)B, diffBilateral diffuse PNH on autopsy, fusion of the thalamus, ventriculomegaly, cerebral atrophyNoneN/ANDBirth/macrocephalyCNS
21/M/(d 14)B, diffBilateral, diffuse PNH
Ventriculomegaly, polymicrogyria, Chiari II, cerebellar cortical dysplasia
Myelomeningocele, club feetN/ANDBirth/myelomeningocele and ventriculomegalyCNS
22/F/(d 15)B, diffBilateral, temporooccipital PNH
Polymicrogyria, hypoplasia of CC, atrophy of cerebellar vermis, cavum septum pellucidum
Microophthalmia, aniridia, macrocrania, dysmorphismsN/AKaryotype normalBirth/multiple congenital malformationsNot available
23/F/(12 y)B, diffBilateral, diffuse, parietooccipital PNH
Mega cisterna magna
NoneYesND2 y/global developmental delayAccidental drowning
24/F/(2 mo)B, diff
subcortical het
Bilateral diffuse PNH (unilateral on imaging)
Agenesis of CC, polymicrogyria, cerebellar vermis hypoplagia, cerebellar white matter heterotopias and periventricular leukomalacia
Microophthalmia, scoliosis, chorioretinal lacunaeN/AAicardi syndrome
Karyotype normal
Birth/neonatal seizuresAspiration pneumonia
25/F/(5 y)B, diffBilateral diffuse PNH
Pachygyria, hippocampal, and cerebellar dysplasia
DysmorphismsYesPDH deficiencyBirth/lactic acidosisLactic acidosis
26/F/10 yB, diffBilateral, diffuse heterotopias
Hypoplasia of corpus callosum
Macrocephaly, ASD, VSD, diaphragmatic hernia, R renal ectasiaYesFLNA c41.415G>C (p.G139R)
Karyotype and CGH normal
Birth/cardiomyopathyNot applicable
27/F/(11 y)B, diffBilateral diffuse PNH,
lobar holoprosencephaly, R frontotemporal polymicrogyria, Dandy-Walker malformation, dysgenesis of cerebellar vermis, absent septum pellucidum, interhemispheric cyst, hypoplasia of optic chiasm, ventriculomegaly
ASD, bilateral microophthalmiaYesNDBirth/dysmorphismsNot available
28/F/4 yB, diffBilateral, diffuse PNHASD, VSD, PDAYesDeletion chr1p36
on karyotype
Birth/muliple congenital malformationsNot applicable
29/M/12 yB, diff
subcortical het
Bilateral, diffuse temporooccipital PNH
Bilateral parietal subcortical heterotopias
Megalencephaly, megalocorneasYesKaryotype and CGH normal4 mo/seizuresNot applicable
30/F/(d 16)Subcortical onlyBilateral subcortical heterotopias
Agenesis of splenium of corpus callosum, pachygyria
Choanal atresia, midface hypoplasia with exophthalmos, duplication R renal collecting system, hepatomegaly, PDAN/AKaryotype normalBirth/neonatal seizures and upper airway obstructionCNS
31/M/(d 1)Subcortical onlyBilateral subcortical perisylvian nodules,
Agenesis of corpus callosum, microgyria, pachygyria
Abnormal ureters, transposition of great arteries, ASD, VSD, coarctation of aorta, cryptorchidism, dysmorphic traitsN/ANDBirth/multiple congenital malformationsCardiorespiratory

Abnormal pregnancy was reported in 10 patients and included polyhydramnios (in 4), first trimester vaginal bleeding (in 2), third trimester vaginal bleeding, intrauterine growth retardation (IUGR), maternal cocaine use, twin gestation, and insulin-dependent gestational diabetes in 1 each. Three children were born prematurely: one each at 29 weeks, at 32 weeks, and at 35 weeks.

Imaging findings

Nineteen patients had computed tomography (CT) scans and 10 had magnetic resonance imaging (MRI) studies. In three cases, CT did not detect the heterotopias, and these were only identified on autopsy (cases 16, 20, and 30). There were 12 patients (39%) with bilateral and diffuse PNHs; 5 patients (16%) with bilateral, focal PNH; 3 (10%) with unilateral and diffuse PNHs; 8 (26%) unilateral and focal PNHs; and 3 (10%) with subcortical heterotopias only. Eight (29%) of 28 patients with PNHs also had subcortical heterotopias (Fig. 1).

Figure 1.


(A) Coronal T2 MRI image showing focal PNH along the right occipital horn in Patient 17 and (B) axial T2 MRI image in same patient showing focal PNH along right occipital horn. (C) Axial CT showing bilateral diffuse PNH and mild hydrocephalus in Patient 26; (D) Axial T2 MRI image showing diffuse unilateral left PNH with extension of heterotopias through white matter to parietal cortex in Patient 10. Left hemisphere is slighter smaller than right. (E) Coronal pathology specimen showing extensive left PNH with extension to parietal cortex in Patient 9.

Associated cerebral malformations

More than 80% of the patients (26 of 31) had other associated brain malformations, and these are summarized in Table 2. These included: ventriculomegaly (29%), cortical abnormalities (26%, mostly polymicrogyria), cerebellar and posterior fossa abnormalities (23%), and abnormalities of the corpus callosum (19%). Six patients had Chiari II malformation with myelomeningocele and hydrocephalus. All six of these patients died in the neonatal period. Nodular heterotopias were discovered at autopsy. The nodular heterotopias had a unilateral focal distribution in three cases (one frontal, one frontal with associated polymicrogyria, and one occipital), had a diffuse bilateral distribution in two cases, and a bilateral occipital distribution in one case. None of the patients had a karyotype or further genetic testing. None of the patients had a family history of myelomeningocele or heterotopias.

Table 2.   Association between distribution of nodular heterotopias and clinical features, cerebral and systemic malformations
 Unilateral focal
(n = 8)
Unilateral diffuse
(n = 3)
Bilateral focal
(n = 5)
Bilateral diffuse
(n = 12)
Subcortical only
(n = 3)
Total
(n = 31)
  1. aStatistically significant value.

Clinical features
 Male gender225%p = 1.000133%p = 0.550120%p = 1.000217%p = 0.676133%p = 0.550723%
 Perinatal death563%p = 0.41200%p = 0.232240%p = 1.000542%p = 1.000267%p = 0.5761445%
 Childhood death788%p = 0.379133%p = 0.195240%p = 0.131975%p = 1.0003100%p = 0.5332271%
 Presence of seizure1/425%p = 0.1303100%p = 0.266 3/475%p = 1.000 7/1070%p = 0.678133%p = 1.00015/2463%
 Presence of developmental delay3/3100%p = 1.000267%p = 0.176 3/3100%p = 1.000 7/7100%p = 0.485133%p = 1.00016/1794%
 Presence of identified genetic abnormality00%p = 0.55000%p = 1.000120%p = 0.525325%p = 0.27200%p = 1.000413%
 Presence of abnormal prenatal factors338%p = 1.000267%p = 0.543120%p = 0.624542%p = 1.000133%p = 1.0001239%
Cerebral malformations
 Subcortical heterotopias113%p = 0.222267%p = 0.237240%p = 1.000217%p = 0.2403100%NA1032%
 Hydrocephalus563%p = 0.027a00%p = 0.537120%p = 1.000325%p = 1.00000%p = 0.537929%
 Cortical abnormalities00%p = 0.146133%p = 0.550120%p = 1.000325%p = 1.000267%p = 0.120723%
 Corpus callosal abnormalities225%p = 1.000133%p = 1.000120%p = 1.000325%p = 1.000267%p = 0.195929%
 Hippocampal abnormalities00%p = 0.550133%p = 0.27100%p = 1.00018%p = 1.000133%p = 0.271310%
 Cerebellar abnormalities113%p = 0.64200%p = 1.00000%p = 0.562650%p = 0.007a00%p = 1.000723%
Systemic malformations
 Cardiac113%p = 0.222133%p = 1.000240%p = 1.000325%p = 0.6973100%p = 0.027a1032%
 Pulmonary225%p = 0.58300%p = 1.000120%p = 1.00018%p = 0.624133%p = 0.422516%
 Genitourinary113%p = 0.642133%p = 0.550240%p = 0.56218%p = 0.201267%p = 0.120723%
 Ophthalmologic00%p = 0.29100%p = 1.00000%p = 0.56433%p = 0.600133%p = 0.422516%
 Myelomeningocele338%p = 0.16100%p = 1.000120%p = 1.000217%p = 1.00000%p = 1.000619%
 Spinal cord338%p = 0.335133%p = 0.550120%p = 1.000217%p = 0.67600%p = 1.000723%
 Orthopedic00%p = 0.291133%p = 0.422120%p = 1.000217%p = 1.000133%p = 0.422516%
 Gastrointestinal00%p = 0.550267%p = 1.000240%p = 0.11218%p = 1.000133%p = 0.349619%
 Multiple malformations225%p = 0.433267%p = 0.543240%p = 1.000325%p = 0.2743100%p = 0.049a1239%

Association with systemic malformations

Systemic malformations were found in 23 (74%) of 31 patients and were multiple in 12 (39%) of 31. These are outlined in Table 2. Cardiac malformations were the most common associated malformations and were observed in 10 patients (32%). These were varied and included atrial septal defects, ventral septal defects, bicuspid aortic valve, and patent ductus arteriosus. Other systemic malformations included myelomeningoceles (in 6), genitourinary (in 7), orthopaedic (in 5), ophthalmologic (in 5), pulmonary (in 5), and gastrointestinal (in 4). One patient (Patient 30) had frontonasal dysplasia, patent ductus arteriosus, hepatomegaly, and renal malformation with bilateral subcortical heterotopias, pachygyria, agenesis of splenium of corpus callosum, and cerebral hypoplasia, reminiscent of the PNHs with frontonasal dysplasia described by Guerrini (Guerrini & Dobyns, 1998).

EEG and seizure outcome

Fourteen (45%) of 31 patients were diagnosed with a seizure disorder, but 13 (76%) of the 17 patients who survived the neonatal period developed epilepsy, and all currently living patients have seizures. Details are presented in Table 3. The age of onset of the seizures ranged from the first day of life to 13 years. Five children (all with bilateral lesions) developed infantile spasms. The majority of patients had partial seizures. EEG abnormalities did not necessarily correlate with the location of unilateral or focal heterotopias.

Table 3.   EEG findings and seizure characteristics of patients with nodular heterotopias and epilepsy
Patient numberHeterotopia subtypeMain seizure type,
age at onset
Seizure frequency
no of meds
EEG
  1. B, bilateral; Diff, diffuse; het, heterotopia; L, left; R, right; sz, seizure; U, unilateral; VNS, vagus nerve stimulator.

 8U, focal (R frontal)Partial sz
2 years
Monthly
on 3 meds
Bitemporal epileptic activity
10U, diff (L)
subcortical het (L parietal)
Partial sz
2 months
Daily
on 2 meds
Left frontotemporal epileptic activity
11U, diff (L)Partial sz
1 year
Monthly
on 3 meds
Bilateral, independent epileptic activity
14B, focal (occipital)
subcortical het (R temporal)
Partial sz
3 years
Daily
on 2 meds
Right temporal epileptic activity
15B, focal (frontal)Partial sz
13 years
Daily
on 5 meds
Left frontotemporal epileptic activity
17B, focal (occipital)Partial sz
8 years
weekly
on 3 meds + VNS
Bioccipital epileptic activity
23B, diffGeneralized sz
6 years
Yearly
on 2 meds
Normal
24B, diff
subcortical het (parietal)
Infantile spasms
10 days
Weekly
on 3 meds
Multifocal epileptic activity
25B, diffInfantile spasms
2 months
Daily
on 2 meds
Multifocal epileptic activity
26B, diffPartial sz
2 months
No sz,
on 1 med
No epileptic activity
27B, diffGeneralized sz
2 days
UnknownMultifocal epileptic activity
28B, diffInfantile spasms
5 months to age 2
No sz
on 1 med
Multifocal epileptic activity
29B, diff
subcortical het
Infantile spasms
4 months
Daily
on 4 meds
Multifocal epileptic activity
30Subcortical onlyGeneralized sz
24 h
Daily
on 3 meds
Multifocal epileptic activity

In general, the seizure disorders were refractory to treatment, and all but two patients were on polytherapy. Five patients had daily seizures, two patients had at least weekly seizures, and two patients had at least monthly seizures. In only two children was complete seizure control achieved with monotherapy: the patient with chromosome 1p36.2 deletion as well as the patient with the FLNA mutation. Both these children had bilateral diffuse PNHs without subcortical heterotopias, the phenotype most often encountered in the adult population.

Genetic testing and underlying etiology

Twenty-two (71%) of 31 patients had genetic and or metabolic testing. Karyotype analysis was carried out in 14 patients, comparative genomic hybridization in 2 patients, and sequencing and deletion analysis of the FLNA gene in 4 patients. The underlying genetic cause was identified in only five patients: one FLNA gene mutation [c.415G>C (p.G139R), previously unreported but highly conserved amino acid], one chromosome 1p36 deletion, one chromosome 15q11 deletion, one sialic acidosis type I, and one pyruvate dehydrogenase deficiency (mutation in E1α subunit).

Other underlying etiologies could be defined in only two other patients. One patient had the clinical diagnosis of Aicardi syndrome as she had the triad of infantile spasms, agenesis of the corpus callosum, and chorioretinal lacunae. Contribution of cocaine exposure is suspected in one case, as cocaine exposure has been shown to disturb neuronal migration and cortical organization in mice and monkeys (Gressens et al., 1992).

Statistical analysis

Results of the statistical analysis using Fisher’s exact test are presented in Table 2. Unilateral focal heterotopias were significantly associated with presence of ventriculomegaly (5 of 8 vs. 4 of 23, p = 0.0027), bilateral diffuse heterotopias were associated with presence of cerebellar abnormalities (6 of 12 vs. 1 of 19, p = 0.007), and presence of subcortical heterotopias without periventricular heterotopias were associated with cardiac malformations (3 of 3 vs. 9 of 28, p = 0.027) and multiple malformations (i.e., presence of two or more systemic malformations) (3 of 3 vs. 7 of 28, p = 0.049).

Discussion

This series of 31 patients is the largest pediatric series of nodular heterotopias published to date. Our study reveals that the clinical spectrum and characteristics of the patients diagnosed in childhood differs significantly from the adult patients previously reported. In general, the patients who present early in life have NHs as part of a more complex clinical picture with multiple cerebral and systemic malformations; developmental delay is the rule rather than the exception and most have refractory seizures and there is an elevated perinatal and childhood mortality rate.

Most striking is the elevated morbidity and mortality observed in our cohort. Systemic malformations were seen in 74% and cerebral malformations were present in 80%. All had developmental delay or mental retardation. The severity of the disease phenotype of pediatric heterotopias is in contrast to that of older patients. In the series by Parrini et al. (2006), 60% of patients had “classic” PNH consisting of bilateral symmetrical involvement, lining the lateral ventricles with limited extension into the occipital horns. These patients had normal intelligence or mild mental retardation, and no systemic or cerebral malformations other than infrequent cardiovascular abnormalities and cerebellar vermis hypoplasia (Parrini et al., 2006). In another study of 33 patients with periventricular and subcortical nodular heterotopias (Dubeau et al., 1995), most patients again had normal intellectual function, with only a minority having mild mental retardation. Again, systemic and cerebral malformations were not frequent. Because a large proportion of our patients did not survive into adulthood [indeed, only 5 (22%) of 23 patients from our series with systemic malformations are currently alive], the association between NH and systemic malformations and the high level of mortality may not be evident in adult-based series.

We found a common association between heterotopias and cardiac malformations. The association between NHs and cardiac abnormalities, especially bicuspid aortic valve and patent ductus arteriosus, is well established in patients with FLNA mutations (Parrini et al., 2006; Kyndt et al., 2007; de Wit et al., 2009). Our study shows that this association can be broadened to include heterotopias with varied appearances and underlying causes, and suggests a potential shared mechanism of cardiac development and neuronal migration. Our study also demonstrates an important association between nodular heterotopias and Chiari II malformations/myelomeningocele. Although there have been previous scattered reports of nodular heterotopias with Chiari II malformations (Kawamura et al., 2001; Humphreys et al., 2007), there is no clearly recognized association between abnormalities of neural tube closure and neuronal migration. Our results may provide support for McLone and Knepper’s (1989) theory of Chiari II pathogenesis, which stipulates that the neural tube defect is the primary event, leading to lack of maintenance of ventricular cerebrospinal fluid CSF pressure, which is required for induction of both neural and calvarial development.

The underlying causes of the heterotopias in our cohort were heterogenous and were identified in 5 (16%) of 31 patients: two chromosomal abnormalities (chr 1p36 del, chr 15q11 del), pyruvate dehydrogenase deficiency, sialidosis type I, and FLNA mutation in one patient. Although FLNA was tested in only a minority of patients, the yield was likely to remain low given the lack of classical bilateral PNH and the absence of family history of heterotopias. Therefore, mutations in FLNA are likely to represent only a minority of patients diagnosed early, in contrast to older patients where FLNA mutations are found in up to one third (Parrini et al., 2006). This again highlights the heterogenous nature of the causes of heterotopias in children as compared to adults.

In our case series, all surviving patients have epilepsy, which is generally severe and difficult to control. This is similar to observations in previous series (Raymond et al., 1994; Dubeau et al., 1995; Battaglia et al., 1997, 2006). There was no correlation between the distribution of heterotopias and the refractory nature of the seizures, but infantile spasms were observed only in children with bilateral disease. The lack of correlation between the location of these heterotopias and the EEG abnormalities further exemplifies the diffuse impact of this migration disorder on cortical function.

There were several limitations to our study. Only 10 of 31 patients had a cerebral MRI. This was mostly due to premature death or clinical instability of the patients. In addition, genetic testing was not performed systematically in our cohort (only performed in 15 of 31); therefore, potentially underestimating the contribution of genetic causes including chromosomal aberrations in the etiology of NH. Because FLNA was tested in only a minority of patients, it is difficult to comment on whether the lower yield as compared to adult series represents a real disparity between the two populations.

In summary, our study reveals that when NHs are symptomatic in children, they are often found in association with multiple cerebral and systemic malformations, and that a high proportion of these patients do not survive into adulthood. Cardiac malformations are particularly common. Bilateral PNHs due to FLNA mutations do not commonly present in children, whereas other chromosomal abnormalities and metabolic disorders are important potential causes that should be adequately investigated. Furthermore, in contrast to adult series, developmental delay and mental retardation are the general rule.

Disclosure

None of the authors has any conflict of interest to disclose. We also confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Ancillary