Seizure, Motor, and Cognitive Outcomes After Epilepsy Surgery for Patients With Sturge-Weber Syndrome: Results From a Multicenter Study

Abstract

Background and objectives: Surgery is widely performed for refractory epilepsy in patients with Sturge-Weber syndrome (SWS), but reports on its effectiveness are limited. This study aimed to analyze seizure, motor, and cognitive outcomes of surgery in these patients and to identify factors associated with the outcomes.

Methods: This was a multicenter retrospective observational study using data from patients with SWS and refractory epilepsy who underwent epilepsy surgery between 2000 and 2020 at 16 centers throughout China. Longitudinal postoperative seizures were classified by Engel class, and Engel class I was regarded as seizure-free outcome. Functional (motor and cognitive) outcomes were evaluated using the SWS neurologic score, and improved or unchanged scores between baseline and follow-up were considered to have stable outcomes. Outcomes were analyzed using Kaplan-Meier analyses. Multivariate Cox regression was used to identify factors associated with outcomes.

Results: A total of 214 patients with a median age of 2.0 (interquartile range 1.2-4.6) years underwent surgery (focal resection, FR [n = 87]; hemisphere surgery, HS [n = 127]) and completed a median of 3.5 (1.7-5.0) years of follow-up. The overall estimated probability for being seizure-free postoperatively at 1, 2, and 5 years was 86.9% (95% CI 82.5-91.6), 81.4% (95% CI 76.1-87.1), and 70.7% (95% CI 63.3-79.0), respectively. The overall estimated probability of being motor stable at the same time post operatively was 65.4% (95% CI 58.4-71.2), 80.2% (95% CI 73.8-85.0), and 85.7% (95% CI 79.5-90.1), respectively. The overall probability for being cognition stable at 1, 2, and 5 years was 80.8% (95% CI 74.8-85.5), 85.1% (95% CI 79.3-89.2), and 89.5% (95% CI 83.8-93.2), respectively. Both FR and HS were effective at ensuring seizure control. For different HS techniques, modified hemispherotomy had comparable outcomes but improved safety compared with anatomical hemispherectomy. Regarding FR, partial resection (adjusted hazard ratio [aHR] 11.50, 95% CI 4.44-29.76), acute postoperative seizure (APOS, within 30 days of surgery; aHR 10.33, 95% CI 3.94-27.12), and generalized seizure (aHR 3.09, 95% CI 1.37-6.94) were associated with seizure persistence. For HS, seizure persistence was associated with APOS (aHR 27.61, 9.92-76.89), generalized seizure (aHR 7.95, 2.74-23.05), seizure frequency ≥30 times/month (aHR 4.76, 1.27-17.87), and surgical age ≥2 years (aHR 3.78, 1.51-9.47); motor stability was associated with severe motor defects (aHR 5.23, 2.27-12.05) and postoperative seizure-free status (aHR 3.09, 1.49-6.45); and cognition stability was associated with postoperative seizure-free status (aHR 2.84, 1.39-5.78) and surgical age <2 years (aHR 1.76, 1.13-2.75).

Discussion: FR is a valid option for refractory epilepsy in patients with SWS and has similar outcomes to those of HS, with less morbidity associated with refractory epilepsy. Early surgical treatment (under the age of 2 years) leads to better outcomes after HS, but there is insufficient evidence that surgical age affects FR outcomes. These findings warrant future prospective multicenter cohorts with international cooperation and prolonged follow-up in better exploring more precise outcomes and developing prognostic predictive models.

Classification of evidence: This study provides Class IV evidence that in children with SWS and refractory seizures, surgical resection-focal, hemispherectomy, or modified hemispherotomy-leads to improved outcomes.

Figure 1. Graphical Representation of Study Design, Participant Inclusion Process, and Surgery Types

(A) Survey (n = 36) and inclusion (n = 16) centers; (B) study design and participant inclusion process; (C) surgical types and areas of all included patients with SWS (N = 214); (D–E) pre- and postsurgical neuroimaging of 2 patients with SWS who underwent focal resection and hemisphere surgery, respectively. LAH = left anatomical hemispherectomies (n = 15); LF = left frontal resection (n = 1); LMH = left modified hemispherotomy (n = 43); LO = left occipital resection (n = 5); LPO = left parieto-occipital resection (n = 1); LT = left temporal resection (n = 1); LTO = left temporo-occipital resection (n = 5); LTPO = left temporo-parieto-occipital resection (n = 29); RAH = right anatomical hemispherectomy (n = 18); RF = right frontal resection (n = 2); RFO = right fronto-occipital resection (n = 1); RFT = right fronto-temporal resection (n = 3); RFTPO = right fronto-temporo-parieto-occipital resection (n = 2); RMH = right modified hemispherotomy (n = 51); RPO = right parieto-occipital resection (n = 5); RT = right temporal resection (n = 4); RTO = right temporo-occipital resection (n = 9); RTPO = right temporo-parieto-occipital resection (n = 19); SWS = Sturge-Weber syndrome. (D) Preoperative and postoperative CT and MR images of a patient with SWS who underwent focal resection (LTO) and E, Preoperative and postoperative CT and MR images of a patient with SWS who underwent hemisphere surgery (RMH). (D-i and E-i) preoperative CT (axial image); (D-ii and E-ii) preoperative MRI (enhanced T1WI, axial image); (D-iii and E-iii) preoperative MRI (enhanced T1WI, sagittal image); (D-iv and E-iv) preoperative MRI (enhanced T1WI, coronal image); (D-v and E-v) postoperative CT (axial image); (D-vi and E-vi) postoperative MRI (enhanced T1WI, axial image); (D-vii and E-vii) postoperative MRI (enhanced T1WI, sagittal image); (D-viii and E-viii) postoperative MRI (enhanced T1WI, coronal image).

Figure 2. Estimated Seizure-Free, Motor-Stable, and Cognition-Stable Probabilities for the Study Groups and Subgroups

Study groups: FR, n = 87; HS, n = 127; subgroups: AH, n = 33 vs MH, n = 94. All the statistical comparisons were performed by log-rank tests. AH = anatomical hemispherectomy; FR = focal resection; HS = hemisphere surgery; MH = modified hemispherotomy.

Figure 3. Kaplan-Meier Curves of All Significant Factors Associated With Seizure-Free, Motor-Stable, and Cognition-Stable Statuses in the Multivariable Analyses in the FR Group (A) and in the HS Group (B and C)

All the statistical comparisons were performed by log-rank tests; for A-iv, because seizure-free status was the only factor in the univariate analysis for cognitive stability in the group that had FR, it was not included in the multivariate analysis. APOS = acute postoperative seizure; FR = focal resection; HP = hemiparesis; HS = hemisphere surgery; Sz = seizure. All the statistical comparisons were performed by log-rank tests.