In this retrospective study, we reviewed a large surgical cohort to investigate the demographic and clinical characteristics, EEG and neuroimaging findings, and post-operative seizure outcome in patients with drug-resistant PLE. This may shed more light on presentations and surgical outcome in this uncommon syndrome of focal epilepsies.
Materials and methods
In this retrospective study, all patients with a diagnosis of drug-resistant PLE, who underwent epilepsy surgery at Jefferson Comprehensive Epilepsy Center, were recruited. Patients were prospectively registered in a database from 1986 to 2015. The diagnosis of PLE was made by the epileptologists working at this institution based on ictal semiology, MRI lesion, and EEG findings. There was no age limitation to enter the study. All patients underwent a comprehensive presurgical evaluation including brain MRI and prolonged video-EEG monitoring.
All patients underwent resective brain surgery. Office visits, telephone contacts, and letters were used to monitor seizure outcome periodically. Postsurgical seizure outcome was classified as Class 1 (seizure-free for one year or more after surgery), Class 2 (less than three seizures per year during the last year of follow-up or nocturnal seizures only), and failure. Aura was not considered as a relapse; only postoperative tonic-clonic seizures and focal seizures with impaired awareness were considered as relapse.
Age, gender, race, epilepsy risk factors (e.g. history of febrile seizures in childhood, any family history of epilepsy, etc.), age at seizure onset, seizure type(s) and history, date of surgery, date of the first relapse (if any), date of the last contact with all patients, and EEG and MRI findings were registered routinely.
No informed consent was required as this was a retrospective study. This study was conducted with the approval of the Thomas Jefferson University Institutional Review Board.
Among 1,225 patients with epilepsy who were investigated for surgery between 1986 and 2015, we identified 19 patients (1.6%) with drug-resistant PLE. Sufficient data were available for 18 patients (11 males and seven females). Mean age (±standard deviation) at the epilepsy onset was 16.2 (±11.5) years and at the time of surgery was 29.7 (±14.2) years. The mean (±standard deviation) preoperative epilepsy duration was 13.13 (±10.22) years. None of our patients had a history of febrile convulsion at childhood. Three patients (16.6%) had a positive family history of epilepsy. Only one patient (5.5%) had a history of status epilepticus before surgery. The mean (±standard deviation) full scale IQ of the patients was 93.8±9.6.
Sixteen patients (88%) had a history of tonic-clonic seizure(s) before surgery. Twelve patients (66%) described focal seizures with impaired awareness and 13 patients (72%) described auras; seven (38.8%) patients had sensory auras contralateral to the resected side, two (11%) had contralateral simple motor signs (one hand tonic posture and one hand clonic jerks), two (11%) patients reported dizziness, one (5.5%) had auditory aura, one (5.5%) had olfactory aura, and one (5.5%) reported epigastric aura. All patients who had motor auras underwent intracranial EEG recordings to determine the epileptogenic zone.
Brain MRI was available for 15 patients. MRI showed parietal tumours in five (33%) patients. Four (26.6%) patients had gliosis or encephalomalacia and one (6.6%) had dysplasia, one (6.6%) had multiple cavernous hemangioma, one patient (6.6%) had right medial parietal cavernous hemangioma, one patient (6.6%) had parietal lobe infarct, one patient (6.6%) had right parietal superficial sidrosis, and one patient (6.6%) had normal MRI.
Interictal scalp EEG recordings were available for 15 patients (83%). Two patients (13%) had right parietal, one (6.6%) had left parietal, five (33.3%) had right extra-parietal (T4, T6, F4, C4), and five (33.3%) had left extra-parietal (F7, T3, T5, F3) interictal spikes. All the unilateral interictal spikes were ipsilateral to the resected side. Two patients (13%) did not have any interictal spikes on surface EEG recordings. Ictal surface EEG recordings were available for 12 patients (66.6%). Of these, two (16.6%) had right parietal ictal onset, one (8%) had left parietal ictal onset, three (25%) had right extra-parietal (one right mid-temporal, one right posterior temporal, and one right frontocentral), four (33.3%) had left extra-parietal (two had left frontocentral, one left mid-temporal, and one left posterior temporal), and two (16.6%) had bilateral mid-temporal ictal onsets. Unilateral ictal EEG findings were ipsilateral to the resected side. Intracranial EEG recordings were available for four (22%) patients. Of these, one had left parieto-occipital origin, two left parietal, and one right fronto-parietal ictal onset.
Seven patients (38%) underwent left parietal resection, 10 (55%) right parietal resection, and one patient (5%) underwent right parietal subpial transection. Pathological findings were available for 13 patients (72%). Four patients (30%) had low-grade tumours (astrocytoma, oligodendroglioma, ganglioglioma, and desmoplastic neuroepithelial tumour), one patient (7%) had brain infarction, three patients (23%) had gliosis, one patient (7%) had cortical dysplasia, and two patients (15%) had cavernous hemangioma. Two patients (15%) had normal pathological findings.
The mean (±standard deviation) follow-up duration after epilepsy surgery was 8.6 (±5.9) years (minimum of one year and maximum of 21 years). Fourteen (77.7%) patients had favourable seizure outcome. Eleven (61.1%) patients had Class 1 (seizure-free) and three (16.6%) patients had Class 2 (rare seizures) seizure outcome. Favourable surgical outcomes were not relevant to specific pathological findings.
In conclusion, PLEs are rare causes of focal epilepsy syndromes. Clinical manifestations and EEG findings in patients with PLE may be unusual or even misleading, but brain MRI is the most valuable tool to localize the epileptogenic zone in these patients. In patients with drug-resistant PLE, surgery is a valuable therapeutic option.
This was a small and clinic-based series and may not represent the full spectrum of PLEs.
Acknowledgement and disclosures
This was a non-funded study.
Marjan Asadollahi, M.D, Ali A. Asadi-Pooya, M.D. and Amin H. Rabiei, M.D. report no disclosures. Michael R. Sperling, M.D. has contracts with Thomas Jefferson University and Eisai, UCB Pharma, Sunovion, SK Life Sciences, Marinus, Lundbeck, Medtronics, Accorda, Upsher-Smith, Brain Sentinel, Pfizer, and Glaxo.