Genetic Testing in Adult Epilepsy Surgery: Low Yield, but Not No Yield

Abstract

Low diagnostic yield of presurgical genetic testing in adult patients with epilepsy

Jünemann C, Stuart A, Kaur N, Wiebe S, Jette N, Singh S, Borlot F, Knake S; Calgary Comprehensive Epilepsy Program Collaborators; Billie Au PY, Klein KM. Epilepsia. 2026 May 20. doi: 10.1002/epi.70291. Epub ahead of print. PMID: 42157695.

Objective: To determine the diagnostic yield of genetic testing in patients undergoing presurgical evaluation for epilepsy. Methods: We conducted a cohort study including 115 adult patients who underwent presurgical evaluation in the Calgary Epilepsy Program between 2019 and 2023 and who had undergone research exome sequencing. A curated epilepsy gene panel comprising 765 Online Mendelian Inheritance in Man (OMIM)-listed epilepsy-associated genes was applied. Variants were classified according to American College of Medical Genetics and Genomics guidelines and assessed for clinical relevance and association with postsurgical outcomes. Results: Pathogenic or likely pathogenic variants in DEPDC5, NPRL2, KCNT2, and PRRT2 were identified, respectively, in 4 individuals (3.5%, 4/115). All variants met stringent quality criteria with high pathogenicity scores (Combined Annotation Dependent Depletion (CADD) 34–37) and absent or extremely low population frequencies in gnomAD v4.1. None of these patients had intellectual disability, and only 1 patient (PRRT2) had a positive family history. The patient with the KCNT2 variant underwent epilepsy surgery with good outcome (Engel class ID). Significance: This presurgical cohort demonstrates a low diagnostic yield of genetic testing in adult epilepsy surgery candidates. However, 3 of 4 patients with (likely) pathogenic variants did not have features that would have prompted clinical genetic testing, indicating that their genetic diagnosis would have been missed based on typical clinical genetic testing criteria in many jurisdictions.

Commentary

In individuals with drug-resistant epilepsy (DRE), consideration of nonpharmacologic therapies, ie surgery or dietary therapies, for epilepsy is considered standard of care. ¹ For individuals who may be candidates for a definitive or palliative epilepsy surgery, undergoing a presurgical evaluation is part of that process. Increasingly during that process, the question of whether there is an underlying genetic etiology for the individual's epilepsy comes up. This question often comes up in pediatric epilepsy, whether there is an already identified structural abnormality or not; can there be a genetic variant that is driving the epilepsy? And if so, how does this affect counseling, prognosis, and surgical decision making? In children, there is some evidence that incorporating genetic testing into the presurgical evaluation can be helpful and may alter management.^2,3 In adults with DRE, the process is less clear, with retrospective studies often identifying genetic variants during or after the presurgical evaluation. ⁴

Jünemann and colleagues aimed to address this knowledge gap by prospectively evaluating a cohort of adult patients undergoing presurgical evaluation at the Calgary Epilepsy Program between 2019 and 2023. ⁵ Patients undergoing evaluation for a definitive surgical procedure or neuromodulation were recruited, with 115 patients (out of 151, 76%) undergoing research exome sequencing, with the yield for pathogenic or likely pathogenic variants being 3.5% (4/115). An additional 3 missense variants of uncertain significance were identified. In this cohort, 92% (106/115) had focal epilepsy with the remainder identified during monitoring as having either generalized or a combination of generalized and focal epilepsy. The cohort was split between nonlesional (55%, 63/115) and structural/lesional epilepsy (45%, 52/115), with structural abnormalities including hippocampal sclerosis (18%), focal cortical dysplasia (6%), other malformations of cortical development (5%), acquired, vascular, or immune mediated lesion (9%), or other (9%).

The authors provide details on those patients who had positive genetic testing with pathogenic or likely pathogenic variants seen in DEPDC5, NPRL2, PRRT2, and KCNT2. Variants of uncertain significance were seen in RELN, RORB, and NPRL2. In 36 remaining patients who did not undergo research exome sequencing, 2 had identified pathogenic findings, including a patient with trisomy X, with mild intellectual disability, and a patient with a variant in KCNT1, with intractable focal epilepsy and a family history of epilepsy. Only 1 patient with positive genetic findings underwent epilepsy surgery. The patient with the KCNT2 variant underwent right temporal lobectomy and right posterior insular resection with an Engel class ID outcome. In the other patients, surgery was not pursued because seizures remitted with further antiseizure medication trials, an epileptogenic focus was not identified, or because risk outweighed potential benefit.

The authors determined that the overall yield of genetic testing is low in adults undergoing presurgical evaluation based on their prospective cohort. They note that in most patients with positive genetic results, phenotypic features which may otherwise trigger genetic testing were absent or rare, namely cognitive/intellectual disabilities and a family history of epilepsy. In this cohort, only 28 patients (24%) would have met the authors’ criteria for genetic testing based on clinical grounds. While the overall yield was 3.5%, this cohort included all those undergoing presurgical evaluation and was not limited to those who had confirmed focal epilepsy or when accounting for an acquired etiology, which would raise the yield of identifying a pathogenic variant to 4.2% in this study (4/96). In contrast, prior studies on adults undergoing surgical evaluation have had higher yields, including a Dutch study which identified pathogenic or likely pathogenic variants in 12% (40/325 who underwent testing, 2% of the total cohort), ⁴ and a U.S. study which had an overall yield of 13% (31/236). ⁶ Jünemann et al note that these studies utilized genetic testing on more clinical grounds, a point to highlight given that the present study looked at all comers, meaning that clinical selection criteria may miss some genetic diagnoses. Pediatric patients, particularly those with coexisting developmental and cognitive impairments and/or features of developmental and epileptic encephalopathies, would expectedly undergo more routine genetic testing as part of their evaluation, which is likely why the yield in pediatric cohorts is higher, such as 21% in the study by Alsubhi et al ³ In another prospective cohort including adults and children with focal epilepsy, the yield was 12.5%, with those patients with pathogenic/likely pathogenic findings more likely to have an earlier median age of seizure onset, a finding which was not necessarily seen in the Jünemann cohort. ⁷

While the authors found an overall low yield of genetic testing as part of the presurgical evaluation, it's important to recognize some key points. While finding a pathogenic/likely pathogenic variant did not necessarily alter the management of these patients, most of these patients were ultimately deemed not ideal surgical candidates. We also do not have a clear denominator in terms of the number of patients who did undergo surgery in this cohort with negative genetic testing. Another point would be that the selection criteria for patients undergoing genetic testing were quite broad and included patients who, under typical circumstances, would not undergo clinical genetic testing, including those with an acquired etiology and those who ultimately did not have focal epilepsy or had a combination of generalized and focal epilepsy, making it challenging to truly call this the ideal cohort to pursue genetic testing in. As the authors note, preexisting cognitive or developmental impairment as well as family history was not common in those with positive genetic findings, highlighting that thinking beyond these factors may be important in adult patients with refractory focal epilepsy. One could argue that in clinical practice, applying a one-size-fits-all approach is not ideal.

Of the variants identified, DEPDC5 and NPRL2 are part of the mTORC1-GATOR pathway and can demonstrate significant phenotypic variability due to variable penetrance and expressivity, including patients with both lesional and nonlesional epilepsy. Individuals may not have features seen with more severe phenotypes, which is in keeping with the study findings. Beyond potential treatment implications, such as candidacy for surgery or targeted/precision therapy, for many patients and families, the why matters. From a family planning perspective, they may want to know if there is a variant that may be inherited or passed down. For some, finding out why they have intractable epilepsy might be information that provides some additional acceptance of the epilepsy journey, which many patients may never get. So while the yield of genetic testing may be low in adults undergoing epilepsy surgery evaluation, low yield does not equal no yield. Balancing resources and availability of testing with the potential benefits, both from a surgical planning and outcomes perspective as well as from the patient perspective, are important.