Brain tumours as an unrecognized etiology of infantile epileptic spasms syndrome (IESS): The role of resective epilepsy surgery

Highlights

• •Tumours are a rare treatable cause of Infantile Epileptic Spasm Syndrome (IESS).
• •Tumour-associated IESS can be focal or symmetric.
• •IESS-associated tumours are most frequently temporal or frontal.
• •Here, symmetric IESS resolved with resection of a left temporal ganglioglioma.
• •Absence of focality should not delay surgical resection in refractory IESS.

Abstract

Brain tumours are a rare cause of infantile epileptic spasm syndrome (IESS). Epilepsy surgery may be utilized in refractory IESS due to tumours, although reports are limited. We report an illustrative case of IESS which resolved after tumour resection and performed a literature review of all reported cases of brain tumours causing IESS with hypsarrhythmia. We present a 10-month-old girl with drug-resistant symmetric epileptic spasms (ES). Initial EEG showed symmetric hypsarrhythmia and brain MRI demonstrated a left temporal lobe tumour. Despite non-focal clinical and EEG findings, she underwent tumour resection at 21 months (ganglioglioma). She subsequently had resolution of ES and hypsarrhythmia and developmental gains. Twenty-seven published cases of brain tumours and IESS were reviewed. ES were refractory to medical management in 74 % of reported cases (17/23). ES were symmetric in 17/21 (81 %) and asymmetric in 4/21 (19 %) specified cases. Hypsarrhythmia pattern was asymmetric in 10/20 (50 %) specified cases. Of 21 surviving patients who underwent surgery, 13 (62 %) were seizure free at follow-up and only 1 (5 %) had no improvement. Developmental outcomes were available in 19/21 (90 %) surviving surgical patients and 10/19 had improvements (52 %); 6/19 had normalized development (32 %). Various tumour types were reported and a temporal or frontal location was most common. Brain tumours can cause IESS with or without clinical or EEG focality and are amenable to surgical treatment. We emphasize that the absence of focal signs should not delay surgical assessment for patients with refractory IESS and brain tumours.

1. Introduction

Infantile epileptic spasm syndrome (IESS) is a severe developmental and epileptic encephalopathy (DEE) and includes both West syndrome (characterized by the electroencephalogram (EEG) pattern of hypsarrhythmia, flexor and/or extensor epileptic spasms (ES), and developmental regression or stagnation) as well as infants who do not meet all criteria for West syndrome[1,2]. While timely and effective medical therapy may have ongoing benefits[3], long-term outcomes remain poor, with approximately three quarters of individuals having unfavorable cognitive outcomes as well as ongoing epilepsy [4].

Multiple studies have shown that etiology remains the most important factor in the long-term prognosis of IESS [4]. Both advances in genetic testing and neuroimaging have been important in determining the cause of IESS. In approximately 60–80 % of cases, an underlying etiology can be determined [[5], [6], [7]] and these include acquired structural anomalies, developmental structural anomalies, and genetic diagnoses [8]. Both cortical and subcortical interactions are believed to be important in the generation of IESS and ES are hypothesized to arise subcortically [9]. While focal lesions such as focal cortical dysplasias, cortical tubers, and developmental structural abnormalities associated with IESS have been widely reported [10,11], IESS associated with tumours is poorly understood [[10], [11], [12]]. In a recent cohort study of IESS, tumours accounted for 1/377 (0.2 %) cases [5].

Given that many children with IESS of lesional etiology do not have sustained electroclinical remission with medical therapy, identifying causes of IESS remediable by surgical intervention is of the utmost importance. Brain tumours are a rare cause of IESS, but one that can be amenable to surgical treatment. Herein, we present a case of IESS with classical hypsarrhythmia and semiology that resolved with resection of a temporal ganglioglioma. We supplement this with a literature review of all reported cases of brain tumours causing IESS and review their seizure outcomes.

2. Methods

A retrospective chart review of the index case was performed. Informed written consent was obtained from caregivers in accordance with the research ethics board at McMaster University, Hamilton, Ontario, Canada. Abstracted data included patient age, sex, age at onset of ES, seizure types, developmental trajectory, initial medical treatment with anti-seizure medications (ASMs), EEG findings, details of surgery, follow up time, seizure and developmental outcomes, and tumour histopathology and location.

In addition, we performed a narrative literature review and identified cases of brain tumours associated with IESS and the EEG finding of hypsarrhythmia. A data search was performed using the terms “IESS OR Infantile Spasms OR Epileptic Spasms OR West syndrome OR hypsarrhythmia” and “brain or intracranial” and “tumour OR tumor OR neoplasm” in PubMed, Google Scholar, and Scopus until April 25, 2025. Citations were screened to identify additional cases. Cases with established genetic syndromes (Tuberous Sclerosis (TSC), Neurofibromatosis 1 (NF1)) or confounding pathology which could have caused IESS (i.e., severe meningitis) were excluded. Cases that did not report a pre-surgical EEG were excluded, as were cases with clinical ES but no hypsarrhythmia. We note that these restrictive inclusion criteria may exclude some cases of IESS, as hypsarrhythmia is not required for the diagnosis of IESS, and we acknowledge this as a limitation of the study. Three cases of cavernous hemangiomas associated with ES were excluded due to the unique features of these growths. Case series or cohort studies where individual data were not presented were not included (e.g. [12]), in addition cases that had missing data were not included.

3. Results

3.1. Case description

A 10-month-old female, previously healthy and developmentally normal, presented with a 2-month history of clusters of flexor ES – unknown whether focal or generalized occurring multiple times per day. She had concomitant developmental regression, reduced social smile and eye contact, and less interactivity with her environment. She was non-dysmorphic and had a non-localizing neurological exam. Her family history was non-contributory.

An awake and asleep EEG at 10 months demonstrated classical hypsarrhythmia (Fig. 1). Brain MRI demonstrated a 7 mm cystic lesion within the left lateral occipitotemporal gyrus associated with increased T2/FLAIR signal and surrounding T2 hyperintensity. The amygdala and hippocampus were involved. Differential diagnosis included ganglioglioma, dysembryoplastic neuroepilepthial tumour (DNET) and focal cortical dysplasia (Fig. 2, A and B). A metabolic, infectious, and genetic workup (epilepsy panel, chromosomal microarray, and whole exome sequencing) was negative.

Fig. 1.

A) Initial EEG at presentation, 10 months old. AP bipolar, 21 scalp electrodes. LFF = 1 Hz, HFF = 70 Hz. Sensitivity 10 uV/mm, time base 30 mm/second. EEG demonstrates hypsarrhythmia. Periods of elecrodecrement were seen.

Fig. 2.

A) Axial T2 and B) Coronal T2) An MRI at 10 months demonstrated a cystic lesion suspicious for tumour. A 7 mm cystic lesion is shown in the lateral occipitotemporal gyrus of the left temporal lobe associated with surrounding white matter signal changes. There was no contrast enhancement, mass effect, or hydrocephalus. An incidentally discovered L temporal arachnoid cyst is also demonstrated at the anterior temporal pole. C) Axial T2 and D) Coronal T2; MRI at 29 months demonstrated expected changes without concern for recurrence. There is a large encephalomalacia cavity at the previous resection site at the left anterior temporal lobe. There is minimal gliosis adjacent to the resection cavity without any abnormal enhancement or diffusion restriction.

Despite the use of vigabatrin, high-dose prednisolone, classic ketogenic diet and topiramate, the ES persisted. An epilepsy surgery workup was initiated at 12 months. A 3 T MRI demonstrated stability of the focal cystic appearing lesion in the left inferomedial temporal lobe, while Positron Emission Tomography (PET) demonstrated hypometabolism of the lesion, and multiple prolonged video EEGs demonstrated hypsarrhythmia with no evident single focus interictally or ictal (Fig. 1).

Despite the absence of clinical and EEG focality, a multidisciplinary decision was made to proceed with lesionectomy as a first step in the context of ongoing ES/hypsarrhythmia and worsening developmental regression. It was hypothesized by the multidisciplinary team that the lesion was responsible for the ES, and this was also supported by PET hypometabolism. Developmental regression was believed to be secondary to her underlying DEE. At 21 months she underwent a left anterior temporal lobectomy with resection of the left parahippocampal gyrus, amygdala, and hippocampus. Pathology was finalized as ganglioglioma (WHO Grade 1, BRAF V600E positive). After surgery, the patient had complete resolution of ES and some developmental gains, including a return of smiling and laughing, though she remains behind development for age. Her post-surgical MRI has consistently demonstrated post-surgical changes without concern for tumour recurrence (Fig. 2, C and D).

EEG 4 weeks after surgery at 22 months demonstrated a posterior dominant rhythm (7 Hz) with normal sleep features (Fig. 2). However, while no hypsarrhythmia was observed, there was at times diffuse delta slowing and slowing seen maximal in the bifrontal head regions. She developed a new seizure semiology at 22 months, described as bilateral tonic-clonic seizures- unknown whether focal or generalized. Valproate was added to her topiramate, with some success. Vigabatrin was discontinued. Per her most recent assessment at 29 months, she continues to have short (10–20 s) bilateral tonic-clonic seizures- unknown whether focal or generalized, occurring in clusters one day per week. She continues to make developmental gains despite seizure frequency, including a new ability to sit and to take some steps with a walker, though she remains significantly delayed. Formal developmental quotient was not available. Her most recent prolonged EEG at 28 months demonstrated a normal posterior dominant rhythm, intermittent focal slowing and interictal discharges from the left posterior temporal occipital region (Fig. 3), and two clinical seizures of indeterminant onset (Supplemental Fig. S1).

Fig. 3.

EEG at 2 years of age after tumour resection, AP bipolar, 33 scalp electrodes, LFF = 1 Hz, HFF = 70 Hz, sensitivity 15 uV/mm, time base 30 mm/second, demonstrates spike and after going slow waves at T5, T9, P9, F1-C1, FP1, F3-C3.

4. Literature review

We initially identified 28 cases of IESS associated with intracranial tumours from the literature, ranging from 1979 to 2024 [[13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31]]. However, after review, 1 case with late onset ES at 110 months and asymmetric hypsarrhythmia was excluded as the presentation was not felt to be in keeping with IESS [12]. Twenty-seven cases were included in the final review. The characteristics of this group are summarized in Table 1, with our case shown for ease of comparison but not included in the following statistics. Eighteen of 24 cases (75 %) that reported sex were male. In the 26 cases that reported age of ES onset, the median age was 7 months and the range was 3–33 months. All cases had clinical ES. Spasm semiology (symmetric versus asymmetric) was reported in 21/27 cases (78 %) and in 6/27 (22 %) it was not specified. In those where ES semiology was specified 17/21 (81 %) were reported as symmetric, 4/21 (19 %) were reported as asymmetric or focal. Six of 27 (22 %) patients had other seizure types onset before surgery.

Table 1.

Characteristics of previously reported cases of Infantile Epileptic Spasm Syndrome (IESS) associated with brain tumours. * = The current manuscript. ACTH = Adrenocorticotropic hormone. ASM = anti-seizure medications. B6 = Vitamin B6, pyridoxine. CNZ = clonazepam. Dev = Developmental. DN-1417 = thyrotropin-releasing hormone analog (experimental therapy). ES = epileptic spasms (sES = symmetric epileptic spasms, aES = asymmetric epileptic spasms). ES Age = age at presentation with ES. F = female. FS = focal seizures. GS = gelastic seizures. GTC = generalized tonic clonic seizures. HA = hypsarrhythmia (mHA = modified hypsarrhythmia). IED = inter-ictal epileptiform discharges. IVMP = intravenous methylprednisolone. KETO = ketogenic diet therapy. L = left. LAM = lamotrigine. LEV = levetiracetam. M = male. mo = months. NR = not reported. OXC = oxcarbazepine. PHB = phenobarbital. PHT = phenytoin. POD = post-operative day. PRED = prednisone. R = right. Ref = Reference. TOP = topiramate. VIG = vigabatrin. VNS = vagal nerve stimulation. VPA = valproate. WHO = World Health Organization. ZON = zonisamide.

In 18/20 cases (90 %) that reported information on developmental status at initial assessment developmental delays were evident. All patients for whom information was available (n = 25) were initially treated with medical therapy, including 14/25 (56 %) with adrenocorticotropic hormone/corticotrophins and 5/25 (20 %) with vigabatrin. Response to medical management was available in 23 cases. In 17/23 (74 %) cases, ES were refractory to medical management. Hypsarrhythmia or modified hypsarrhythmia was observed in each patient on the initial reported EEG; symmetry of the pattern was available in 20/27 cases (74 %). In 10/20 (50 %) cases the EEG pattern was described as symmetric and 10/20 (50 %) cases were described as focal or asymmetric (i.e., either regarding the location of the hypsarrhythmia or superimposed electrographic abnormalities).

Table 1 provides information on tumour types and locations. Briefly, common tumour types included 5/27 (19 %) gangliogliomas, 5/27 (19 %) hypothalamic hamartomas, and 3/27 (11 %) high-grade gliomas. Common locations included 13/27 (48 %) that involved the frontal and/or temporal lobes, as well as 5/27 (19 %) hypothalamic hamartomas.

Only 3/27 (11 %) cases were treated non-surgically: one case of hypothalamic harmatoma (HH) that responded entirely to ACTH treatment, one case where surgery was declined and the child was managed with chemotherapy and radiation and died at 18 months (high grade glioma), and a third case of a basal ganglia lesion that was not thought to be amenable to surgery. The median age of surgery was 15.5 months (IQR: Q1 8.5, Q3: 29). There were two outliers who had surgery at 109 and 256 months respectively (cases 12 and 14); with these two cases excluded, the average latency to surgery was 4 months. Lesionectomies and lobectomies predominated except in cases of HH (five total), where disconnection was typical. Two patients died of complications in the immediate post-surgical period and a third died 54 months after surgery from undisclosed causes.

The mean follow-up was 36 months. Of the surviving surgical patients, 13/21 (62 %) were seizure free at their most recent evaluation, 1/21 (5 %) had resolution of ES but ongoing other seizures types, 5/21 (24 %) had a reduction in spasm burden, 1/21 (5 %) was entirely refractory, and 1/21 (5 %) was initially refractory but improved to an Engel 3a outcome after a third revision of the resection for recurrent tumour. An EEG outcome was only reported in 9 cases; of these, all had resolution of hypsarrhythmia, as summarized in Table 1.

Developmental outcomes were available in 19/21 (90 %) surviving surgical patients. Of these, 10/19 patients had improvements (52 %) after surgery but remained delayed, 6/19 (32 %) had a normal developmental status at time of last assessment, and only 3/19 (16 %) did not show an improvement, all of whom had malignant brain tumours.

5. Discussion

Tumours are a rare cause of IESS, accounting for 1/377 (0.2 %) [5] and 2/541 (0.4 %) [32] cases of IESS in two recent studies. However, they are an important cause of IESS, as they can be amenable to surgical intervention. In IESS surgical cohorts, 12/80 (15 %) [33] and 3/60 (5 %) [11] patients had tumours. Our review of the literature supports the role of surgery in this context, as of the surviving surgical patients, 13/21 (62 %) were spasm free at their most recent evaluation and only a single patient (1/21) (5 %) continued to have refractory ES with no improvement. However, these excellent results may represent a degree of reporting bias. The benefits of epilepsy surgery are not limited to seizure control, as developmental outcomes were also improved in half and normal in a third of surviving patients in the literature.

Despite originating from a focal lesion, spasm semiology was symmetric when specified in 17/21 (81 %) cases and EEG was symmetric hypsarrhythmia in 10/20 (50 %) cases. Our case further demonstrates that an apparent mismatch between EEG/semiology and neuroimaging, should not prevent surgical consideration, especially in refractory cases. Resolution of hypsarrhythmia in our case unmasked focal IEDs from another region, which may indicate an underlying dual pathology such as focal cortical dysplasia (focal cortical dysplasia type III with developmental tumor) as a contributor to the patient’s ongoing epilepsy. An underlying genetic etiology is also possible, although work-up to date was negative, further genetic testing may be completed in the future (i.e., re-analysis of whole exome sequencing, or whole genome sequencing if becomes clinically available). Furthermore, given the recurrence of seizures, additional surgical evaluation is anticipated and re-discussion in multidisciplinary seizure conference.

While the pathophysiology of IESS remains to be fully elucidated, it is hypothesized that disrupted cortical and subcortical networks provoke multifocal local epileptogenic activity and impact normal cognitive functions [9]. In our series, both cortical tumours (15/27, 56 %) and tumours within the subcortical regions such the basal ganglia, thalamus, hypothalamus, and brainstem (12/27, 44 %) were associated with IESS. However, additional studies are needed to better understand the interaction of cortical and subcortical mechanisms in the development of IESS [9].

Multiple different tumour types were associated with IESS in the series we identified, with a fronto-temporal tumour location seemingly conferring more risk more than any specific histopathology. The high representation of ganglioglioma and hypothalamic hamartomas may reflect a propensity of these tumours and/or their locations to cause IESS, but may also reflect the publication of case series (e.g. [18]). For comparison, [33] reported 10 glioneuronal tumours, one oligoastrocytoma, and one pilocytic astrocytoma in a series of brain tumours causing IESS. Analyzing the impact of tumor location and histology on IESS outcome in future studies would be of benefit. In addition, given that we only included cases with ES with hypsarrhythmia, it is possible that we missed cases.

Finally, our review demonstrates that medically refractory cases and IESS cases with both focal and non-focal findings should prompt special concern for tumour/structural causes as the etiology and prompt referral for surgical assessment. Time to surgery varied widely in the cases we described here, from cases that presented and were treated emergently to cases of tumours that were resected years after the onset of ES. Treatment delays have been associated with worse outcomes in IESS [3], and we therefore advocate for an expedited surgical assessment in any patient with refractory IESS and an intracranial tumour, regardless of the focality of their semiology or EEG.

6. Conclusions

We report a case in which non-focal ES/hypsarrhythmia resolved with the resection of a left temporal ganglioglioma. Other cases from the literature support the association of focal lesions with ES and hypsarrhythmia, with or without focal semiologies or EEGs. Frontal and temporal tumour locations were most common. We propose that the absence of focality on hypsarrhythmia and seizure semiology should not delay surgical assessment for patients with refractory IESS and brain tumour.

7. Declaration of generative AI and AI-assisted technologies in the writing process: Statement

No generative AI or AI-assisted technologies were used in the writing process.

Funding sources

This research did not receive specific funding via grants or other sources from public, commercial, or not-for-profit agencies.

CRediT authorship contribution statement

Rowan Pentz: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Data curation. Kevin Jones: Writing – review & editing, Investigation, Data curation. Ronit Mesterman: Writing – review & editing, Data curation. Rajesh RamachandranNair: Writing – review & editing, Data curation. Hiroshi Otsubo: Writing – review & editing, Data curation. Ayako Ochi: Writing – review & editing, Data curation. Ivanna Yau: Writing – review & editing, Data curation. Elizabeth J Donner: Writing – review & editing, Data curation. George M Ibrahim: Writing – review & editing, Data curation. Puneet Jain: Writing – review & editing, Methodology, Data curation. Robyn Whitney: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: