Electroclinical and Multimodality Neuroimaging Characteristics and Predictors of Post-Surgical Outcome in Focal Cortical Dysplasia Type IIIa

Abstract

Focal cortical dysplasia (FCD) type IIIa is an easily ignored cause of intractable temporal lobe epilepsy. This study aimed to analyze the clinical, electrophysiological, and imaging characteristics in FCD type IIIa and to search for predictors associated with postoperative outcome in order to identify potential candidates for epilepsy surgery. We performed a retrospective review including sixty-six patients with FCD type IIIa who underwent resection for drug-resistant epilepsy. We evaluated the clinical, electrophysiological, and neuroimaging features for potential association with seizure outcome. Univariate and multivariate analyses were conducted to explore their predictive role on the seizure outcome. We demonstrated that thirty-nine (59.1%) patients had seizure freedom outcomes (Engel class Ia) with a median postsurgical follow-up lasting 29.5 months. By univariate analysis, duration of epilepsy (less than 12 years) (p = 0.044), absence of contralateral insular lobe hypometabolism on PET/MRI (p _Log-rank = 0.025), and complete resection of epileptogenic area (p _Log-rank = 0.004) were associated with seizure outcome. The incomplete resection of the epileptogenic area (hazard ratio = 2.977, 95% CI 1.218-7.277, p = 0.017) was the only independent predictor for seizure recurrence after surgery by multivariate analysis. The results of past history, semiology, electrophysiological, and MRI were not associated with seizure outcomes. Carefully included patients with FCD type IIIa through a comprehensive evaluation of their clinical, electrophysiological, and neuroimaging characteristics can be good candidates for resection. Several preoperative factors appear to be predictive of the postoperative outcome and may help in optimizing the selection of ideal candidates to benefit from epilepsy surgery.

Keywords: characteristics; drug-resistant epilepsy; focal cortical dysplasia type IIIa; postoperative outcome; predictors.

FIGURE 1

Flowchart of patient selection. FCD, focal cortical dysplasia; TLE, temporal lobe epilepsy.

FIGURE 2

Engel class Ia seizure freedom. Kaplan–Meier curve demonstrated the seizure freedom outcome after resection in the overall population (66 patients). The cumulative probability of continuous seizure freedom was 69.7% (95% CI, 58.6–80.8%), 62.3% (95% CI, 50.3–74.4%), 56.5% (95% CI, 43.0–70.0%), and 48.4% (95% CI, 29.8–67.1%) at 12, 24, 36, and 60 months or more after surgery, respectively.

FIGURE 3

Engel classification outcome during postoperative follow-up. Engel classification outcomes at postoperative 12, 18, 24, 36, and 48 months and last follow-up. Total numbers of investigated cohorts and the rates of cohorts were illustrated for each Engel class outcome. The Engel class Ia outcome worsened with time.

FIGURE 4

Kaplan–Meier survival curves of Engel class Ia seizure freedom depending on different outcome predictors. (A) Duration of epilepsy was not significantly associated with seizure freedom. (B) Absence or presence of FC did not significantly influence seizure outcome. (C) Absence or presence of FBTCS did not impact seizure outcome. (D) Complete resection of the epileptogenic area was significantly more likely to obtain seizure freedom. (E) ATLR was also not significantly associated with seizure outcome. **p < 0.01. FC, febrile convulsion; FBTCS, focal to bilateral tonic–clonic seizure; ATLR, anterior temporal lobe resection.

FIGURE 5

Kaplan–Meier survival curves of Engel class Ia seizure freedom since surgery. (A) Absence or presence of gray–white blurring on MRI was not significantly associated with seizure outcome. (B) Concordance or discordance of preoperative MRI with ictal scalp VEEG did not significantly influence seizure outcome. (C) Absence or presence of contralateral frontal lobe hypometabolism on PET/MRI was also not significantly correlated with seizure outcome. (D) Patients without contralateral insular lobe hypometabolism on PET/MRI were significantly more likely to obtain seizure freedom. *p < 0.05. VEEG, video electroencephalogram.