We depended in our pre-operative evaluation on the cumulative data gathered through proper history taking, seizure semiology, clinical examination and lesional location on cerebral imagining. Surgery was done even if the EEG was non-contributory without further investigations such as invasive recording, electrocorticography or functional mapping due to paucity of resources as well as the presence of an obvious lesion on brain MRI that was concordant with seizure semiology,. The convergence of clinical and radiological evidence pointed to the lesion in all cases as being the most probable epileptogenic focus.
Our decision is reinforced by the study done by Roper et.al (2009) [5] that concluded that each test can be helpful in localizing the Epileptogenic zone in extra-temporal epilepsy, but no single method is completely sensitive or specific and that’s why the final decision making always depends on assessing concordant results from several different diagnostic modalities.[5] Similarly, Sandok and Cascino (1998)[6] clarified that the use of EEG before surgery depends on the planned surgical procedure and has variable results.[6, 7] Peri-operative and intraoperative functional mapping are less important in patients who are undergoing lesionectomy based on MRI.[6]
Anatomical location
In our series, most patients were found to have frontal lesions (12 patients, 60%), followed by parietal lesions (Six patients, 30%), and one patient with occipital lesion and another with parieto-occipital lesion (5% each). This goes in line with literature as seizures are predominantly associated with lesions in the frontal or temporal lobe, insular regions as well as lesions in close relation to the cortex.[3, 8]
Within the follow up period (Mean 23.31 months), 4 patients were lost (3 frontal, 1 parietal). Of the remaining 16 patients, only 11 were seizure free with Engel’s class 1 (68.75 %). Eight of nine patients with frontal lesions were seizure free (88.9%), while 3 of 5 patients with parietal lesions were seizure free (60%), one patient was not controlled, and one patient with convexity meningioma had three times recurrence and deteriorated and eventually died. The remaining two patients had occipital and Parieto-occipital lesions, and both were not seizure free.
Spencer and Huh (2008) [9] found that seizure freedom after extratemporal cortical resections reached up to 66% in children ( with higher proportion of lesional resections), and up to 76 % in adults. The follow up period for studies included in their review was at least 1 year. Outcome of resective surgery to treat epilepsy in children differs from that in adults primarily because the syndromes amenable to surgery are much more heterogeneous.[9]
Similarly, Téllez-Zenteno and colleagues did a systematic review and meta-analysis about surgical outcome in lesional and non-lesional epilepsy [10]. The outcomes were reported after at least 1 year of follow-up. They found that seizure freedom in extratemporal lesional (LE) Epilepsy (MRI positive) is about 60 %. These results are close to our results where 68.75 % were seizure free within our follow up period (mean 23.31 months).
Relation to cerebral parenchyma
In addition to the anatomical location of the lesions, we found it of special interest to compare the intra-axial with the extra-axial lesions to evaluate the impact of the relation between the lesion and the cerebral parenchyma on seizure outcome after lesionectomy. It’s known that both intra-axial and extra-axial brain tumors may be associated with seizures (Rajneesh and Binder 2009) .[11] Although in cases with intra-axial lesions the epileptogenic zone might be included within, around, or far from the lesional zone, the simple lesionectomy still have a chance of resecting the epileptogenic zone if it was in close relation to the lesional zone. In extra-axial lesions, however, the lesions are excised without dealing with the previously compressed, edematous or irritated cortex, or in other words, the epileptogenic zone is not resected. The pathophysiology of seizures with extra-axial lesions ,such as meningiomas, is assumed to be caused by the combined effect of the of tumor mass effect on epileptogenic cortex, acid/base derangements from cerebral edema, and/or disturbances of neurotransmitter pathways (Chaichana et al. 2013).[12] Therefore, seizures with extra-axial lesions might need further management in addition to simple lesionectomy. During our follow up period (mean 23.31 months), eight out of 11 patients with extra-axial lesions (72.7 %) were seizure free, while three out of five patients with intra-axial lesions (60%) were seizure free. Although we couldn’t record a significant statistical difference in seizure outcome between intra-axial and extra-axial lesions (P value 0.611), we still highly recommend a further study on a larger scale and longer follow up period to get a strong evidence based conclusion regarding the difference in epilepsy outcome between intra-axial and extra-axial lesions.
Histopathology
The most common pathology encountered in our series was low grade glioma for intra-axial lesions and meningioma for extra-axial lesions. Among nine intra-axial lesions, seven (78%) were found to be low grade gliomas, of which three were astrocytoma WHO Grade 2, two oligodendroglioma, one ependymoma and one sub-ependymoma. The other two intra-axial lesions (22%) were infarction and cavernoma.
During the long follow up period, 4 patients with low grade gliomas were lost, leaving 5 patients with intra-axial lesions including three patients with low grade gliomas of which two were seizure free with Engel’s Class 1 (oligodendroglioma grade 2, and subependymoma grade 2), and one was not free with Engel’s class 4 (ependymoma grade 2). The remaining two patients were the patient with infarction that was seizure free with Engel’s class 1, and the cavernoma patient that was not seizure free with Engel’s Class 3.
Although we operated on the infarction patient erroneously thinking it was a tumor, yet there is evidence in literature to suggest that a focal ischemic or hemorrhagic event may result in the formation of gliotic tissue within the atrophic brain lesion (encephalomalacia), that present with intractable seizures.[13] Ghatan et al 2014 [14] evaluated surgical management of medically refractory epilepsy due to early childhood stroke and concluded that epilepsy surgery is effective in controlling medically intractable seizures after perinatal vascular insults.[14]
In addition to the intra-axial lesions, we had 11 patients with extra-axial lesions including nine meningiomas (82%) of which eight were seizure free Engel’s Class 1 (88.9%), the other two patients (18%) had AVM and plasma cell myeloma of the occipital bone and both were not seizure free. These results matches what was found in previous studies about the most common pathologies to present with seizures, but due to the few number of each pathology in our series, we found that it would be inaccurate to compare our results for each individual pathology with the studies that focus on one pathological type and its seizure outcome. We could, however, show the number of meningiomas, and low grade gliomas (grouped together) to allow us to do such comparison.
Lynam et al (2007) [15] showed that the patients with primary CNS tumor more commonly presenting with seizures were found to have low grade Gliomas. Rajneesh and Binder (2009)[11] had similar results. Likewise, Tandon and Esquenazi (2013)[3] studied a series of tumor related epilepsy and the most common tumors in their group included low-grade gliomas and GNTs (glioneuronal tumors) including both temporal and extratemporal locations.
Chaichana et al. 2013[12] evaluated seizure control for patients undergoing meningioma surgery and found that among patients who presented with seizures, the majority remained seizure free. They related poor seizure control to factors such as uncontrolled preoperative seizures, parasagittal/ parafalcine tumors, and sphenoid wing tumors. [12] We had nine meningiomas, eight of them were seizure free with Engel’s Class 1, while one patient had three times recurrence until he deteriorated and died after 20 months.
Extent of surgical resection:
Despite the variability of the lesions included in our study regarding their exact anatomical location, their relation to the cerebral parenchyma or even their histopathological nature, lesionectomy was successful to guarantee a satisfactory outcome with Engel’s Class 1 in 68.75% of cases during the follow up period (Mean 23.31 months). Only two patients had subtotal resection (AVM and infarction). The patient with infarction is seizure free with Engel’s Class 1, while the patient with AVM is not seizure free with Engel’s Class 2. Sixteen patients underwent total resection of which three were lost in follow up, nine are seizure free with Engel’s Class 1, and four are not seizure free. Two cases were operated upon by total resection with safety margin (both were oligodendrioglioma), one of them was lost in follow up, while the other is seizure free with Engel’s class 1.
Although we achieved seizure freedom in 68.75% of our cases, it should be emphasized that most of the lesions in our series were tumors (13 cases out of 16 (81.25%)) and the fact that we didn’t have any case of focal cortical dysplasia or gliosis or other non tumoral lesions apart from three cases with AVM, Cavernoma and infarction in our series. It was found significant to mention the results of previous studies regarding focal cortical dysplasia as it has the poorest outcome of lesional epilepsy [5], and since we didn’t have such cases with known poor outcome, it is necessary to be clarified.
Our results are strongly aligned with the results of previous studies that evaluated seizure control after lesionectomy for extratemporal tumors. Tandon and Esquenazi (2013)[3] concluded that initial aggressive tumor resection yielded Engel’s Class 1 in all of the patients in their study except for the cases with tumor residual or tumor recurrence [3]. Sandok and Cascino (1998) [6] evaluated the surgical treatment for perirolandic lesional epilepsy and Engel’s Class 1 was found in 13 of 14 cases (93 %).[6] Elsharkawy et al. (2008) [16] reviewed Long-term outcome of extratemporal epilepsy surgery among 154 adult patients. In reviewing seizure outcome in relation to pathological findings, 93.8% of their patients with neoplastic lesions had favorable outcomes after surgery, which dropped to 56% after 14 years of follow up (Engel’s class 1).
The results of these studies are closer to what we found in our series than other studies that included all tumoral and non tumoral lesions and justify the good results of our lesionectomies being related mostly to the tumoral nature and total resection.
Limitations
The paucity of resources limited our perioperative neurophysiological work up. However, we could rely on concordant evidence of MRI and seizure semiology. Moreover, the lack of a database documenting all the cases presenting with seizures prevented us from including cases on a retrospective manner, and limited our overall number of cases, with which we couldn’t withdraw data of statistical significance. We could, however, compare our results with those of literature to validate our current practice.