Magnetic resonance–guided laser interstitial thermal therapy for pediatric drug-resistant epilepsy: a pooled analysis and systematic review of the literature

OBJECTIVE Magnetic resonance–guided laser interstitial thermal therapy (MRgLITT) is a minimally invasive alternative to open resection for pediatric drug-resistant epilepsy (DRE). This systematic review and individual participant data meta-analysis aimed to identify independent predictors of seizure outcomes and operative and neurological complications following MRgLITT. METHODS Uni- and multivariable mixed-effects Cox proportional-hazards regressions models were used to identify independent predictors of time to seizure recurrence following MRgLITT. Among patients with at least 12 months of follow-up, uni- and multivariable mixed-effects logistic regression analyses were conducted to ascertain the independent risk factors associated with seizure recurrence at last follow-up, operative complications, and postoperative neurological complications. RESULTS A literature review identified 354 pediatric patients with a mean epilepsy duration of 7.5 (SD 5.3) years prior to MRgLITT. The mean age at seizure onset was 4.52 (SD 4.69) years, and focal seizures were more common (85.5%) than generalized seizures (14.5%). Lesions were detected on MRI in 82.1% of cases. The most common epilepsy etiologies were hypothalamic hamartoma (HH; 23.7%) and malformations of cortical development (23.7%). The mean follow-up duration after MRgLITT was 16.02 (SD 11.63) months. Engel class I outcomes were achieved in 57% of patients. In 205 cases where information was available regarding postoperative neurological complications, 35 patients (17.1%) experienced postoperative neurological complications, with hemiparesis as the most frequent complication (n = 16 patients). Of the 354 total patients who underwent MRgLITT, 8.2% underwent revision epilepsy surgery. No operative or clinical characteristics were associated with seizure recurrence. Seizure freedom probability was significantly higher among patients with HH compared to those with nonlesional MRI (p = 0.012). Patients with mesial temporal sclerosis experienced earlier seizure recurrence (p = 0.023), and an extratemporal surgical location was associated with longer seizure freedom probability (p = 0.034). Lesional MRI was associated with reduced odds of postoperative neurological complications (p = 0.031). CONCLUSIONS MRgLITT may be a safe and effective alternative option for pediatric DRE. Further prospective studies are warranted to elucidate MRgLITT strategies in pediatric DRE.