Focal cortical dysplasia is a common cortical malformation and an important cause of epilepsy. There is evidence for shared molecular mechanisms underlying cortical dysplasia, ganglioglioma, hemimegalencephaly, and dysembryoplastic neuroepithelial tumor. However, there are no familial reports of typical cortical dysplasia or co-occurrence of cortical dysplasia and related lesions within the same pedigree. We report the clinical, imaging, and histologic features of six pedigrees with familial cortical dysplasia and related lesions. Twelve patients from six pedigrees were ascertained from pediatric and adult epilepsy centers, eleven of whom underwent epilepsy surgery. Pedigree data, clinical information, neuroimaging findings, and histopathologic features are presented. The families comprise brothers with focal cortical dysplasia, a male and his sister with focal cortical dysplasia, a female with focal cortical dysplasia and her brother with hemimegalencephaly, a female with focal cortical dysplasia and her female first cousin with ganglioglioma, a female with focal cortical dysplasia and her male cousin with dysembryoplastic neuroepithelial tumor, and a female and her nephew with focal cortical dysplasia. This series shows that focal cortical dysplasia can be familial and provides clinical evidence suggesting that cortical dysplasia, hemimegalencephaly, ganglioglioma, and dysembryoplastic neuroepithelial tumors may share common genetic determinants.
A PowerPoint slide summarizing this article is available for download in the Supporting Information section here.
Dr Richard Leventer is a consultant pediatric neurologist at the Royal Children's Hospital.
Focal cortical dysplasia (FCD) encompasses a spectrum of cortical abnormalities characterized by dyslamination with or without abnormal cell types. FCDs are among the most common malformations of cortical development (MCDs) and are frequently associated with intractable epilepsy. FCD is often the only congenital abnormality present in the brain or elsewhere, with seizures usually being the sole clinical manifestation. The classification of FCD is based on the presence or absence of features in addition to cortical dyslamination: type I (no abnormal cell types), type II (dysmorphic neurons with or without balloon cells), and type III (type I with another lesion). FCD may accompany hippocampal sclerosis and the “developmental” glioneuronal tumors, ganglioglioma and dysembryoplastic neuroepithelial tumors (DNETs), refining the classification to type IIIa and type IIIb, respectively. A broad developmental classification of MCDs classifies the tubers of tuberous sclerosis complex (TSC), FCD type II, ganglioglioma, hemimegalencephaly, and DNET as all being malformations due to abnormal neuronal and glial proliferation with abnormal cell types.
The etiology of FCD is largely unknown. Although the lesions of FCD (particularly FCD type II) share many imaging and histologic features with cortical tubers of TSC, FCD is usually sporadic, without clinical evidence to support a simple genetic etiology. Mutations in CNTNAP2 have been reported in children with FCD from Old Order Amish pedigrees. These children differed from most patients with FCD because of mental retardation and macrocephaly, and the imaging findings and histology were not typical of either FCD type I or FCD type II. There have been no other reports of familial FCD. Apart from a single report of a father-son pair with DNET, clinical evidence is lacking to suggest a familial basis for ganglioglioma, DNET, or nonsyndromic hemimegalencephaly, or familial co-occurrence of FCD with hemimegalencephaly, ganglioglioma, or DNET. A role for human papilloma virus (HPV)16 infection in the etiology of FCD type IIb has recently been suggested from studies of resected tissue and a mouse model.
Recent studies of resected tissue from patients with hemimegalencephaly identified de novo somatic mosaic mutations in the genes of the PI3K-AKT3-mTOR (mammalian target of rapamycin) pathway in 9 of 28 patients.[6, 7] Evidence from molecular analysis of FCD specimens has led to the hypothesis that these malformations may have a shared pathogenesis, due to abnormalities of mTOR or other cellular pathways that regulate neuronal and glial proliferation. However, until now, there has been no evidence to suggest a link between these disorders from family studies. Here, we report six families, each with two individuals with FCD, ganglioglioma, hemimegalencephaly, or DNET, adding clinical evidence that suggests a shared genetic susceptibility underlying these disorders, and showing for the first time that typical FCD can be familial.
Families were ascertained by referral to pediatric and adult epilepsy services. Clinical details were obtained from patient interview and medical records. Each institution's human research ethics committee approved the study. Informed consent was obtained from patients or their parents in the case of minors.
Brain magnetic resonance imaging (MRI) was obtained using age-specific epilepsy protocols on 1.5 T or 3 T scanners. Neurologists and neuroradiologists skilled in the detection of MCDs reviewed the MRI scans. Resected tissue was classified by a neuropathologist, according to the proposed system of the International League Against Epilepsy (ILAE) Diagnostic Methods Commission. TSC1 and TSC2 mutation screening was performed using DNA extracted from resected brain tissue or whole blood using polymerase chain reaction (PCR) with exon-specific primers for TSC1 and TSC2 or by whole exome sequencing techniques.
Six families, each with two affected individuals with seizures and at least one having FCD, were ascertained (pedigrees and selected MRI and neuropathology images: Fig. 1, clinical and imaging details: Table 1, additional MRI and neuropathology images: Fig. S1). All but one patient has undergone epilepsy surgery. None of the patients had clinical features suggestive of TSC, and all patients except family 5 were screened for TSC1 and TSC2 mutations. No mutations were identified.
Aphasia, R facial sensorimotor with generalization
L middle frontal gyrus
>50% Sz reduction
Visual aura then L arm dystonia with generalization
Sz – free
Focal dyscognitive with vomiting
L posterior parietal
Sz – free
Arousal and bipedal hyperkinetic movements
L medial superior frontal gyrus
Sz – free
Focal dyscognitive with bilateral hand automatisms +/− generalization
L hippocampus and parahippocampal gyrus
Lesionectomy and anterior temporal lobectomy
Sz – free
Epileptic spasms and focal motor
L posterior quadrant
8 m and 5 y
Focal corticectomy then L posterior quadrantectomy
LTG, CBZ, PHT
Epileptic spasms and multifocal clonic
Sz – free
Family 1 comprises brothers with neonatal seizures secondary to right hemisphere FCD type IIa, multifocal in patient 1A and restricted to the right posterior quadrant in patient 1B. The father and paternal uncle of these brothers have each had rare nocturnal seizures without focal features on interictal electroencephalography (EEG). Review of their recent brain MRI studies performed at 3 T revealed no abnormalities. Family 2 includes female 2A with FCD type Ia at the depth of an abnormal branch of the left central sulcus. Her female first cousin 2B had a ganglioglioma in the left superior temporal gyrus. Family 3 includes a female 3A with FCD type IIb in the right anterior temporal pole. Her male cousin 3B had a left middle frontal gyrus DNET. Family 4 includes female 4A with right occipital lobe FCD type IIb. Her nephew 4B had well-controlled focal seizures with a left posterior parietal region lesion on MRI, highly suggestive of FCD that was not removed (see Fig. S1). Family 5 includes a male 5A with an area of FCD type IIb at the depth of an abnormally deep left medial frontal lobe sulcus. His sister 5B had left temporal lobe imaging consistent with both hippocampal sclerosis and FCD in the posteromedial left temporal region, with cortical thickening, blurring of the gray-white matter junction, and high signal on T2 and fluid-attenuated inversion recovery (FLAIR) images. Histopathology showed hippocampal sclerosis. It was not surprising that no histologic features of FCD were found as there was limited surgical tissue obtained from the abnormal posterior region. Family 6 includes a female 6A with a large area of FCD type IIa in the left posterior quadrant. Her brother 6B had left hemimegalencephaly, containing areas of severe dyslamination and dysmorphic neurons as seen in FCD type IIa.
Despite its prevalence, and histopathologic similarity to the tubers in TSC, the molecular causes of FCD remain unknown, and there are no familial cases of typical FCD reported. This suggests that either FCD does not have a genetic basis, or that it does have a genetic basis but occurs sporadically, possibly due to somatic mutations in affected tissue. Evidence to support a relationship between FCD, DNET, hemimegalencephaly, and ganglioglioma has come from a number of sources. First, there are numerous reports of both FCD type I and FCD type II occurring with DNET and ganglioglioma,[9, 10] and a report of DNET, ganglioglioma, and FCD coexisting in a “composite lesion” in one patient. Second, an association of FCD, hemimegalencephaly, and ganglioglioma has been suggested from molecular and genetic studies of surgical specimens. These studies show that cytomegaly, seen not only in tubers of TSC but also in hemimegalencephaly, FCD, and ganglioglioma, may reflect aberrant activation of the mTOR and β-catenin signaling cascades, known regulators of cell growth, consequently causing defective control of neuronal and glial proliferation.[12-14] As found in our series, attempts to detect germ line or somatic mutations in TSC1 and TSC2 in patients with FCD and related lesions have largely been unsuccessful, suggesting that abnormalities in other genes in the mTOR cascade or related pathways of neuronal proliferation and differentiation may play a role. If we assume that some forms of FCD, DNET, hemimegalencephaly, and ganglioglioma may have a shared etiologic mechanism and timing, then it is reasonable to consider that the occurrence of these lesions in closely related family members may be more than a coincidence and reflective of a shared mechanism and genetic etiology.
If the lesions in our families are caused by genetic mutations giving rise to increased susceptibility, then the question remains as to how different family members may have different lesions, especially in families 2, 3, 4, and 6 in which the proband has FCD but their relative has a developmental tumor or hemimegalencephaly. In TSC, there is significant phenotypic pleiotropy with variation of lesions in family members with the same mutation. More so, within an individual patient with TSC, a mutation can result in both a dysplastic cortical lesion (cortical tubers) and a benign neoplastic lesion (giant cell astrocytoma). Alternatively, lesions in these families could result from “two hits,” one causing a nonpathogenic germ line mutation in a cortical development gene carried within the family, and the other causing a somatic mutation of the other allele of that gene within the FCD or other MCD. This hypothesis may explain the sparing of other family members. One may speculate that the nature of the “second hit” may explain the differences in phenotypes within families, with the second mutation either affecting different cellular precursors (i.e., glial vs. neuronal) or affecting precursors at different developmental time points (i.e., early vs. late progenitors).
Identifying the etiology of FCD and related lesions has remained elusive, despite FCD being a relatively common entity. The six families presented herein provide suggestive clinical evidence of a genetic link between FCD, ganglioglioma, hemimegalencephaly, and DNET. It must be acknowledged that FCD and related lesions may occur within the same pedigree as a chance association. Although accurate data on the prevalence of FCD are lacking, we estimate from our own experience that FCD and related lesions are the cause of epilepsy in at most 1 in 100 patients, so the occurrence in multiple families by chance alone would be quite unlikely. Families 1 and 6 in our study, each with siblings with pathologic features of FCD type IIa, are therefore the most convincing pedigrees supporting our hypothesis of a shared genetic susceptibility. Whether this susceptibility extends to other pedigrees that include a family member with FCD and others with MRI-negative epilepsy requires further study, but it is important that in such families the MRI studies of patients with MRI-negative epilepsy be closely scrutinized for subtle lesions. Exploring these relationships further will require molecular exploration of germline mutations in families of interest, in combination with the study of tissue resected at epilepsy surgery for somatic mutations.
We would like to thank the patients and their families for participating in this study, Jacinta McMahon for the preparation of pedigree data in Figure 1, and Kate Pope and Rosie Burgess for helping to obtain patient data and DNA samples. This work has been supported by the Victorian Government's Operational Infrastructure Support Program. Funding was provided by the National Health and Medical Research Council of Australia and the Murdoch Childrens Research Institute.
Disclosures or Conflicts of Interest
None of the authors have any conflicts of interest to disclose. We confirm that we have read the Journal's position in issues involved in ethical publication and affirm that this report is consistent with those guidelines.