Abstracts

Rationale: MR-guided Laser Interstitial Thermal Therapy (MRgLITT) is a minimally invasive technique used in epilepsy surgery. It has the potential to produce comparable seizure outcomes to traditional approaches while reducing risk to the patient. We report the feasibility, safety and clinical outcomes of an exploratory study of MRgLITT as a minimally invasive surgical procedure for the ablation of epileptogenic foci in children with drug-resistant, lesional epilepsy. Methods: A retrospective chart review of all MRgLITT procedures was conducted at a single tertiary care center. All procedures were performed using a FDA-approved surgical laser ablation system (Visualase Thermal Therapy System; Visualase, Inc., Houston, TX). Pre-defined clinical and surgical variables were extracted from archived medical records. Results: Seventeen patients underwent 19 MRgLITT procedures from May 2011 to January 2014. Mean age at seizure onset, 7.1 years (range: 0.1 - 14.8 years). Mean age at surgery, 15.3 years (range: 5.9 - 20.6 years). Surgical substrates included focal cortical dysplasia (n=11), tuberous sclerosis complex (n=5), hypothalamic hamartoma (n=1), mesial temporal sclerosis (n=1), developmental tumor (n=1) and one patient with Rasmussen encephalitis who chose to undergo MRgLITT and refused hemispherectomy. Unexpected post-ablation edema prolonged hospital stay in one patient. Another patient experienced an anticipated inferior visual field deficit. Three technical complications occurred: inaccurate fiber placement (n=2) resulting in one of the patients being converted to open resection and failure of the laser cooling apparatus (n=1) with early termination MRgLITT. No major or permanent, unexpected complications occurred. Average length of hospitalization post-surgery, 1.56 days. Mean follow-up, 16.1 months (n = 16; range: 3.5 - 35.9 months;). Engel Class I outcome was achieved in 41% (7/17) of patients, Engel Class II in 6% (1/17), Engel Class III in 18% (3/17) and Engel Class IV in 35% (6/17). Of the 8 patients with Class I and II outcomes, 38% (3/8) had at least one prior resection and, of the 9 patients who resulted in Class III and IV outcomes, 56% (5/9) had at least one prior resection with one patient from this latter group having undergone 3 prior resections. Fisher exact test was not statistically significant for the association between Engel Class outcome and previous resection (p = 0.64). Conclusions: Our cohort demonstrates that MRgLITT is a viable option for surgically complex situations where risk of reoperation is high. In comparison to published groups of similar patients, MRgLITT was less effective than conventional surgery and was associated with technical complications. These findings were expected in this cohort that comprises the first patients at our institution to undergo this novel procedure. We anticipate improved outcomes and a reduced number of complications in the future as our experience develops. Further studies are required to delineate optimal candidates for this procedure and larger cohorts are needed to more accurately define outcome and complication rates.