Functional hemispherectomy for seizure control in encephalocraniocutaneous lipomatosis: illustrative case

Abstract

BACKGROUND

Encephalocraniocutaneous lipomatosis (ECCL) is a rare neurocutaneous syndrome composed of a spectrum of congenital cutaneous, ocular, and brain anomalies. Principal anomalies include nevus psiloliparus lesions of the scalp, ocular choristomas consisting of dermolipomas, and intracranial lipomas and arachnoid cysts. The hypothesized pathogenic basis in ECCL is a developmental mesenchymal defect involving neural crest cell derivatives. Management is typically focused on treating symptoms, directed at the organ system and supportive neurorehabilitation. Reports to date have only focused on pharmacotherapeutic treatment of drug-refractory epilepsy (DRE), a principal neurological phenotype. Surgical management of medication-refractory epilepsy, which can occur in up to 70% of ECCL subjects, has not been described to date.

OBSERVATIONS

The authors present an illustrative case report of a 4-year-old girl with ECCL who developed DRE and underwent a functional hemispherectomy for seizure relief. The clinical course, diagnostic evaluation, and surgical treatment are described, with emphasis on surgical-pathological observations.

LESSONS

To the authors’ knowledge, surgical treatment for hemispheric onset DRE in the context of ECCL has not been reported. Systemic and clinical features on presentation are reviewed to highlight aberrant mesenchymal differentiation as the developmental basis for the syndrome. The authors also underscore the role of functional hemispherectomy in the treatment of DRE in the context of a genetic/developmental syndrome.

https://thejns.org/doi/10.3171/CASE2578

Encephalocraniocutaneous lipomatosis (ECCL), or Haberland syndrome, is a rare neurocutaneous syndrome that consists of a spectrum of congenital cutaneous, ocular, and brain anomalies. Since the first description in 1970,¹ diagnostic, clinical, and pathological aspects of the syndrome have been identified.^2–5 The principal (cutaneous) finding in ECCL is a nevus psiloliparus lesion consisting of a smooth, hairless fatty nevus of the scalp associated with nonscarring alopecia. Additional cutaneous findings include subcutaneous fat collections, most commonly in the zygomatic or frontal-temporal regions, and ocular, cutaneous skin tags, composed of fibrous or lipomatous tissue. The most common ocular anomalies are choristomas, benign tumors consisting of dermolipomas and lipodermoids, which are derived from dermoid and adipose tissue, respectively. Additional ocular findings include corneal and scleral anomalies, colobomas, aniridia, and microphthalmia. Bilateral skin and ocular changes may be seen in up to 40% of affected individuals.⁵ Intracranial findings in EECL most commonly include cranial and spinal lipomas and arachnoid cysts.⁶ Additional changes observed include unilateral or asymmetric hemispheric volumetric loss, calcification, and hydrocephalus, hypothesized to likely result from underlying vascular defects including leptomeningeal angiomatosis.⁶

Phenotypic expression is variable since ECCL is associated with genetic mosaicism involving gain-of-function variants in FGFR1 (fibroblastic growth factor receptor 1) or KRAS (Ki-ras2 Kirsten rat sarcoma viral oncogene homolog) genes.⁵ Neurocognitive impairment and early-onset epilepsy are the most common neurological phenotypes in ECCL. The spectrum of neurological features and CNS anomalies is hypothesized to arise secondary to aberrant neural crest development, rather than as a consequence of the primary brain malformation. Hence, the neurocognitive and clinical status of affected individuals does not correlate with the extent of neuroimaging anomalies.

Up to 50%–75% of ECCL subjects develop epilepsy, which often follows a medication-refractory course, with antiepileptic medications described as the mainstay treatment to date.⁴,⁵ Candidacy for surgical treatment of epilepsy and epilepsy surgery outcomes in pediatric/adult subjects with ECCL have not been described despite the high prevalence of drug-refractory epilepsy (DRE) within the syndrome. We describe the diagnostic evaluation of epilepsy in a 4-year-old subject with ECCL and surgical treatment with a functional hemispherectomy. Complete seizure freedom was achieved at the 2.5-year follow-up. Our results demonstrate the efficacy of surgical treatment of DRE in ECCL and within the broader context of a genetic syndromic etiology of epilepsy.

Illustrative Case

Clinical Presentation

Our subject was a 4-year-old left-handed girl who was born at 41 weeks of gestation. She was initially evaluated for ocular and skin anomalies observed following birth. Dermatological examination indicated a left frontal nevus psiloparis, alopecia, and subcutaneous lipoma in the frontal-temporal region (Fig. 1A and B). Ophthalmic evaluation indicated a unilateral, left epibulbar dermoid, a fibrous hamartoma of the eyelid, optic nerve hypoplasia, and retinal hamartomas (Fig. 1C). She subsequently developed new-onset seizures at 10 months of age with motor seizures involving the right leg and hand. Her seizures persisted despite trialing levetiracetam, zonisamide, valproate, and clobazam. Brain MRI revealed left lateralized findings that included 1) intracranial lipoma within Meckel’s cave extending to the left skull base, 2) hemispheric atrophy that was most pronounced within the left temporal lobe, 3) ex vacuo prominence of the left lateral ventricle and extra-axial subarachnoid space, and 4) dysplastic leptomeninges extending throughout the entire subarachnoid space, most pronounced over the posterior left cerebral hemisphere and consisting of serpiginous bridging vessels, proteinaceous fluid content, and coarse fatty deposits along meninges (Fig. 1D–I).

FIG. 1.

Dermatological examination indicated a left frontal nevus psiloparis lesion consisting of a lipomatous nevus with nonscarring alopecia (major criterion) (A; white asterisks) and a frontal-temporal subcutaneous fibrous-lipomatous lesion (B; black asterisk). Ophthalmic examination indicated the presence of a left epibulbar dermoid (black asterisk), a fibrous hamartoma of the eyelid, optic nerve hypoplasia, and retinal hamartomas (C). Brain MRI revealed dysplastic leptomeningeal angiomatosis as evident on axial gradient echo (D; black arrows) and postcontrast (E; white arrows) images. Axial T2-weighted image (F) showing an expanded anterior middle cranial fossa (black arrows) with temporal and occipital lobe atrophy, ex vacuo dilated left atrium and occipital horn, and left to right torcular shift. Axial and coronal T1-weighted images showing parafalcine (G) and convexity (H) lipomatous deposits (white arrows), along with an intracranial lipoma (I; white arrow) in the left middle cranial fossa, Meckel’s cave.

ECCL was diagnosed, based on diagnostic criteria that include identification of major and minor clinical features within the skin, CNS, and eye.^5,6 In our index subject, diagnosis was established by involvement of the skin;nevus psiloparis (major criterion), alopecia, and subcutaneous lipoma in the frontal temporal region (minor criterion); intracranial lipoma (major criterion), angiomatosis, hemispheric atrophy, and asymmetric ventricle (minor criterion) in the CNS; and choristoma (major criterion) in the eye (Table 1).

TABLE 1.

Summary of diagnostic clinical and radiographic features by organ system

CNS	Ocular	Cutaneous
Middle cranial fossa lipoma*	Epibulbar & limbal lipodermoid*	Nevus psiloliparus*
Ipsilateral ventriculomegaly†	Lt optic nerve hypoplasia	Alopecia†
Atrophy of lt hippocampal, mesial temporal, & occipital lobe†	Fibrous hamartoma of the lids†	Subcutaneous lipoma†
Leptomeningeal angiomatosis†	Retinal hamartomas	Nodular skin tag†
Lt occipital cortical mantle atrophy	Choroidal hamartoma
Callosal dysgenesis & thinning of the splenium
Colpocephaly
Overgrowth of lt hemicranium & polymicrogyria

* Major diagnostic criteria for clinical diagnoses of ECCL.⁵

† Minor diagnostic criteria for clinical diagnoses of ECCL.⁵

Phase I evaluation

A phase I evaluation was undertaken due to the patient’s progressive course of DRE. Video-EEG monitoring indicated a combination of clinical and electrographic seizures with unilateral left hemispheric onset involving the left occipital (most active focus) and temporal regions (Fig. 2). On physical examination, she had spastic weakness in her left hand. Functional MRI (fMRI) involved acquisition of blood oxygen level–dependent contrast images at 3T using the following paradigms: for motor, passive (assisted) right hand finger tapping alternating with rest; and for auditory, passive listening to music and story in the parents’ voices alternating with rest and 5 minutes of resting fMRI scan. Passive right finger tapping and resting-state analysis revealed a right dominant motor network with mild activation along the left paracentral gyri in the vicinity of the left hand knob region. Passive auditory testing and resting-state analysis revealed right hemispheric lateralization for language/music processing. Diffusion tensor imaging showed markedly reduced fractional anisotropy throughout the posterior half of the left cerebral hemisphere, involving the major white matter tracts, including the corticospinal tracts, optic radiations, and left superior and inferior longitudinal fasciculi. Neuropsychological testing revealed broadly impaired intellectual functioning across all domains. At 4 years of age, she was assessed to function at the developmental level of a 2.5-year-old. Her Wechsler Preschool & Primary Scale of Intelligence, Fourth Edition (WPPSI-IV) scores were 53 for verbal comprehension index, 54 for visual-spatial index, and 53 for full-scale IQ.

FIG. 2.

Preoperative electroencephalogram showing seizure foci in the left hemisphere: the O1 electrode (left occipital, the most active seizure focus; red arrow) with a clinical correlate of subtle head turn to the right; the O2 electrode (right occipital) with a clinical correlate of subtle head turn to the right (of note, in this child, the O2 electrode actually overlies a portion of the left occipital lobe, as this lobe crosses the midline into the right hemicranium; blue arrow); and the T3 and T5 electrodes (left temporal lobe), which were associated with subtle behavior change and sometimes an ictal cough (black arrow).

Given her seizure burden, early age at onset, underlying unilateral hemispheric anomaly, concordant phase I evaluation indicating unilateral left hemispheric onset for all her seizures, and fMRI indicating reorganization of language and motor function to the right hemisphere, a left hemispheric disconnection was recommended for seizure relief and to reduce risk for epileptic propagation to the normal right hemisphere.

Surgery

A left peri-insular hemispherectomy was undertaken at 5 years of age. Surgical anatomy was notable for a thickened, dysplastic pial-arachnoid layer with superficial lipomatous deposits throughout the hemisphere (Fig. 3A and B). Abnormal leptomeningeal blood vessels were observed (Fig. 3C). The disconnection involved an anterior temporal lobectomy, intraventricular parasagittal callosotomy, and disconnections in the transcircular, anterior frontal-basal, and posterior planes, respectively, with insular cortex resection. She tolerated the procedure well and underwent in-hospital rehabilitation for her expected right hemiparesis.

FIG. 3.

A:Surgical anatomical view of the affected left hemisphere corroborating the radiographic features with dysplastic leptomeninges and lipomatous deposits through the left hemisphere. B: Higher-magnification view of abnormal leptomeninges with thickened pial-arachnoid layers, infiltrated with lipomatous deposits. C: Higher-magnification view of the aberrant pial-arachnoid leptomeningeal blood vessels. A = anterior; F = frontal; P = posterior; T = temporal.

Histopathology

Histopathological examination of the leptomeninges (Fig. 4A–C) demonstrated aberrant thickening, with abnormal vessels, mature adipose tissue, and extensive abnormal cells of apparent fibroblastic differentiation. The underlying cortex demonstrated subpial gliosis and scattered calcifications. The left hippocampus (Fig. 4D and E) showed moderate to focally severe hippocampal sclerosis, with pyramidal neuron loss most notable within CA1 as well as within granule cells of the dentate gyrus, with reduplication of the granule cell layer.

FIG. 4.

Histopathological examination of the leptomeninges and hippocampus. A: Hematoxylin and eosin (H&E)–stained section acquired at lower power (magnification ×40) showing thickened leptomeninges (Lep) with numerous abnormal vessels (arrowheads) as well as underlying cortex (Cx) with scattered calcifications (arrows). B:H&E-stained section at high power (magnification ×400) showing leptomeninges with abnormal cells of apparent fibroblastic differentiation. C: H&E-stained section (magnification ×200) showing significant portions of leptomeninges containing mature adipose tissue (arrows) with some abnormal, fibroblastic cells demonstrating multinucleated florets (circle). The inset (magnification ×400) shows diffusely positive CD34 immunohistochemical staining within abnormal cells, further suggesting fibroblastic lineage. D: H&E-stained section (magnification ×40) of the hippocampus showing moderate to focally severe sclerosis, most notably characterized by pyramidal neuron dropout, apparent within CA1, as well as granule cells of the dentate gyrus (DG). E:NeuN immunohistochemical staining (magnification ×40) best elucidates neuronal dropout (arrows point to neuronal loss within CA1), with the inset (magnification ×100) further demonstrating reduplication of the granule cell layer in DG. The hippocampus also demonstrated diffuse gliosis, best seen by glial fibrillary acidic protein immunohistochemical staining (not shown).

Follow-Up

She was ambulatory with ankle orthotic assistance by 3 months following surgery. Her language skills were unaffected immediately following surgery. Follow-up neuropsychological evaluation at 1.5 years after surgery showed improvement in her WPPSI-IV scores: 73 for verbal comprehension index, 58 for visual-spatial index, and 63 for full-scale IQ. By 2 years after surgery, her language exceeded her preoperative baseline, and she was working on early reading skills. As of her most recent follow-up, she has been seizure free (Engel class IA) for 2.5 years.

Informed Consent

The necessary informed consent was obtained in this study.

Discussion

Observations

Our subject demonstrated major and minor clinical and radiographic diagnostic features for clinical diagnosis of ECCL based on dermatological, ophthalmic, and CNS evaluation. Neurological aspects of the ECCL syndrome include developmental and psychomotor delays and epilepsy, which can variably manifest across a wide severity spectrum. Following diagnosis, a multisystemic evaluation is recommended that involves ophthalmic, dermatological, neurological, and musculoskeletal assessments and imaging, together with assessment of neurodevelopmental status to identify supportive measures for neurocognitive and rehabilitative needs. ECCL has been designated as a tumor predisposition syndrome.² In addition to lipomas and jaw tumors, CNS (low- and high-grade tumors)⁷,⁸ and renal, Wilms tumors have been reported in ECCL.²,⁵,⁷,⁸

The hypothesized pathogenic basis for the diverse cutaneous, ocular, and CNS anomalies observed in this syndrome is a developmental mesenchymal defect involving neural crest cell derivatives.⁴ This hypothesis is further supported by the common mesenchymal basis of the various tumors observed in ECCL, including lipomas, odontomas, osteomas, and bone cysts. CNS lipomas and leptomeningeal angiomatosis, as seen in histological assessment in our subject, result from incorrect differentiation of the meninx primitiva, which is derived from neural crest cells. These observation therefore support the hypothesis that the brain anomalies in ECCL are not primary brain malformations, but arise secondary to aberrant neural crest development. Hence, it has been inferred that the neurocognitive status of affected individuals does not correlate with the presence and/or extent of neuroimaging anomalies.

ECCL is associated with genetic mosaicism, and molecular criteria for ECCL diagnosis include identification of (mosaic-activating) pathogenic gain-of-function variants in either FGFR1 or KRAS.⁵ These two gene defects converge on the RAS/MAPK pathway that regulates cell growth and cell differentiation. It is therefore hypothesized that FGFR1, which is expressed during mesoderm induction and regulates the transformation of epiblast cells into mesoderm, may be linked to the pathogenesis of ECCL.³

Surgical candidacy in genetic/syndromic epilepsy, although well established for certain etiologies like Sturge-Weber syndrome, is not well elucidated for many other genetic etiologies.⁹ Seizure outcomes following surgery for DRE secondary to genetic causes also vary significantly by etiology. Patients with epilepsy related to mTOR pathway mutations have favorable odds for seizure freedom as compared with patients with genetic mutations associated with channelopathies.¹⁰ The relationship between timing of surgical intervention and epilepsy control outcomes is confounded by underlying etiology, age at epilepsy onset, and the severity and duration of epilepsy.¹¹ Surgery at younger age often reflects developmental anomalies like cortical malformation syndromes that are associated with lower rates of long-term seizure control. Younger age at surgery is also associated with lower improvement in language outcomes, possibly reflecting underlying disease severity.¹² Additional studies have demonstrated improved long-term developmental scores with early surgery and shorter duration of epilepsy. Studies examining the impact of surgical disconnection techniques on seizure control outcome have not consistently demonstrated whether particular surgical approaches are associated with better outcomes.¹³,¹⁴

Despite observations of somatic mosaicism in ECCL, the lateralization of cerebral anomalies likely contributed to improved surgical candidacy in our subject. The hemispheric developmental anomalies in ECCL likely predispose one toward development of early-onset epilepsy. In our subject, following epilepsy progression to a drug-refractory state, diagnostic studies showed concordant left hemispheric seizure localization. High-density EEG indicated left posterior epileptiform discharges. Video-EEG captured prototypical clinical seizures with electrographic onset in the left temporal and occipital regions. A PET-MRI study showed marked left temporal and parietal-occipital hypometabolism, although this interpretation was limited by significant cortical atrophy, as noted on structural neuroimaging. In addition, passive motor and language task localization along with resting-state analysis revealed right hemispheric lateralization, suggesting a reduced risk for motor-language impairment following hemispheric disconnection. A hemispheric functional disconnection was therefore recommended, in line with known prognostication factors for seizure control and cognitive outcomes in pediatric subjects with (unilateral) hemispheric onset DRE.^15,16 Seizure freedom, as observed in our subject, demonstrates the efficacy of surgical intervention, based on appropriate patient candidacy in subjects with syndromic genetic epilepsy.¹⁷

Lessons

Our case report is the first description of surgical treatment for DRE in a pediatric subject with ECCL, a rare neurocutaneous syndrome. We highlight the phenotypic spectrum of cerebral anomalies in ECCL, through description of surgical anatomy, radiographic features, and histopathological findings observed in our subject. The developmental etiology of the syndrome along with its systemic features, clinical evaluation, and management is reviewed through a literature-based review. In light of the reported incidence of epilepsy in up to 70% of subjects with ECCL,⁴ our case report demonstrates the feasibility of effective surgical treatment for DRE in the context of a multisystemic and CNS developmental disorder due to aberrant mesenchymal differentiation.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: Ahmed, Koueik. Acquisition of data: all authors. Analysis and interpretation of data: Ahmed, Koueik, Hsu. Drafting the article: Ahmed, Koueik, Helgager. Critically revising the article: Ahmed, Koueik. Reviewed submitted version of manuscript: Ahmed, Helgager. Approved the final version of the manuscript on behalf of all authors: Ahmed. Administrative/technical/material support: Ahmed. Study supervision: Ahmed.

Correspondence

Raheel Ahmed: University of Wisconsin, Madison, WI. raheel.ahmed@neurosurgery.wisc.edu.