Abstract
Leon-Rojas JE, Iqbal S, Vos SB, Rodionov R, Miserocchi A, McEvoy AW, Vakharia VN, Mancini L, Galovic M, Sparks RE, Ourselin S, Cardoso JM, Koepp MJ, Duncan JS. Br J Neurosurg. 2021;1-6. doi:10.1080/02688697.2021.1966385
Background: The piriform cortex (PC) occupies both banks of the entorhinal sulcus and has an important role in the pathophysiology of temporal lobe epilepsy (TLE). A recent study showed that resection of more than 50% of PC increased the odds of becoming seizure free by a factor of 16. Objective: We report the feasibility of manual segmentation of PC and application of the Geodesic Information Flows (GIF) algorithm to automated segmentation, to guide resection. Methods: Manual segmentation of PC was performed by two blinded independent examiners in 60 patients with TLE (55% Left TLE, 52% female) with a median age of 35 years (IQR, 29–47 years) and 20 controls (60% Women) with a median age of 39.5 years (IQR, 31–49). The GIF algorithm was used to create an automated pipeline for parcellating PC which was used to guide excision as part of temporal lobe resection for TLE. Results: Right PC was larger in patients and controls. Parcellation of PC was used to guide anterior temporal lobe resection, with subsequent seizure freedom and no visual field or language deficit. Conclusion: Reliable segmentation of PC is feasible and can be applied prospectively to guide neurosurgical resection that increases the chances of a good outcome from temporal lobe resection for TLE.
Commentary
Epilepsy affects 65 million people worldwide. Continued seizures despite optimal medical management occurs in 30% to 40% of people with epilepsy. 1 For those with temporal lobe epilepsy (TLE), randomized controlled trials have demonstrated temporal lobectomy is superior to continued medical management alone. 2 But even with surgical therapy, many patients have recurrent postoperative seizures. 3 The quest for lifelong seizure freedom without adverse effects remains out of reach for many and electrical storms of the brain continue. Identifying predictors of outcome after temporal lobectomy is critical. 4 A new predictor of outcome appears to be the extent of resection of the piriform cortex (PC). 5 -7
The PC includes the primary olfactory cortex. It is likely the site of olfactory auras which are not uncommon in TLE. The anatomical and network relationships of the PC to the hippocampus is responsible for the common phenomenon where smelling particular odors triggers vivid memories and emotions. Piriform cortex spans the entorhinal sulcus with an anterior part in the base of the frontal lobe and posterior part in the temporal lobe superior to the amygdala. The PC is the most epileptogenic cortex and a segment of the PC is referred to as the “Area Tempesta” due to its dramatic epileptogenicity. Phylogenetically, the PC and hippocampus both have a 3-layered cortex that is vulnerable to excitotoxic injury and responsible for kindling in animal models of epilepsy. 8,9 The PC in humans has been shown to be involved in focal epilepsy where it is also epileptogenic and involved in the propagation of epileptic discharges. 10
Resection of the PC is challenging as it is indistinct in humans and surrounds the M1 portion of the middle cerebral artery and its lenticulostriate branches. Despite the surgical challenges, several studies have demonstrated increased resection of the PC is strongly associated with improved seizure-free outcomes without an increase in complications. Galovic et al, performed analysis of preoperative and postoperative magnetic resonance imaging scans (MRIs) following standard anterior temporal lobectomies by a single surgeon in 107 patients and showed that removal of at least 50% of the PC was required to achieve seizure freedom after temporal lobectomy. Only 9% of patients become seizure free if less than half of the PC was resected, whereas 60% of patients became seizure free if at least 50% of the PC had been resected. These findings were validated in 2 independent external cohorts. The authors concluded that if the epileptogenic network involving the PC is sufficiently disrupted by removing at least half of this area, the patient has 16 times higher odds of postoperative seizure freedom. Reassuringly, the patients with a larger percent of PC removal did not have adverse cognitive or psychiatric outcomes. 5
In a slightly different patient population and with different imaging and volumetric analysis, studies by Borger et al showed that transsylvian selective amygdalohippocampectomy including resection of at least 27% of the temporal part of the PC (tPC) resulted more patients with seizure freedom compared to less than 27% resection of the tPC (83% vs 39%; P = .0007). Among those patients with at least 27% resected proportion of PC, there were significantly more patients with seizure freedom compared to those with less of PC resected (83% vs 39%, P = .0007). In those patients who were seizure free, 96% had removal of at least 27% of the tPC. 6 This strong correlation continued with longer term follow-up with a mean follow-up duration of 3.75 ± 1.61 years. The resected proportions of hippocampus and amygdala did not significantly affect outcome. In addition, the volume of resected tPC was the only independent significant predictor for successful reduction or withdrawal of anti-seizure medications. 7
Technological advances in epilepsy surgery are improving outcomes and minimizing complications. Computer assisted navigation including visualization of the location of anatomically relevant structures with the operating microscope are improving therapeutic options. With respect to the PC, Leon-Rojas et al developed a reliable manual PC segmentation and application of the Geodesic Information Flows (GIF) algorithm for automated segmentation to be used in guiding resection. 11 Using ITK-SNAP, two independent blinded examiners performed 160 manual segmentations (2 segmentations in 60 patients and 20 controls) that can be used surgically for guiding anterior temporal lobe resection. A total of 320 PC volumes were obtained and a surgical boundary mask of the temporal portion of the PC to be removed during surgery while preserving the frontal component. The segmentation process and its accuracy are facilitated by the GIF algorithm which propagates and extrapolates manual labels into a new dataset of patients. The authors created a template for automatic PC segmentation that can be applied to individual patients scheduled for TLE surgery. The authors developed the data for automatic parcellation of the PC which resulted in segmentation of the relevant models for intraoperative use through the operative microscope. This preoperative 3D model was employed in the operation of a 24-year-old woman with DRE since age 6 due to encephalitis and resultant right mesial temporal sclerosis. Following the anterior temporal lobectomy and macroscopic resection of the PC, neuronavigation and 3D representation of the PC was used in the operative microscope. An intraoperative MRI scan was performed prior to closure to confirm removal of the temporal portion of the PC without compromise of the frontal portion. The patient had no complications or deficits as a result of the surgery. Twelve months following surgery she remained seizure free. Neuropsychological testing showed improved working memory with slight decreases in verbal recall and learning. In this patient, the GIF algorithm resulted in an accurate segmentation of the PC that successfully guided temporal lobe resection. The authors achieved proof of principle and are planning a prospective series to verify the benefits and identify any shortcomings of this approach.
More studies are needed to address whether the potential increased risk to lateral lenticulostriate arteries and risk of damage to neighboring structures outweigh the potential benefits. This approach should be compared to outcomes seen with laser interstitial thermal therapy (LITT) including whether the extent of resection of the tPC is more important than the extent of ablation of other temporal lobe regions (eg, anterior, medial, inferior, and parahippocampal regions). Techniques in LITT are evolving to allow for ablation of the PC. Liu et al showed how a two-trajectory laser amygdalohippocampotomy including hippocampus, amygdala, and piriform/entorhinal/perirhinal cortex was superior in full ablation and seizure outcomes. 12 This more extensive ablation including the PC was correlated with the outcome of seizure freedom. Hwang et al performed preablation and postablation volumetric analyses of hippocampus, amygdala, PC, and ablation volumes in 39 patients with mTLE who underwent LITT and found that with multivariable logistic regression only percent PC ablation was a significant predictor of seizure freedom at 6 months (Odds Ratio (OR) 1.085, 95% Confidence Interval (CI) [1.012-1.193], P = .019) and at 1 year (OR 1.074, 95% CI [1.003-1.178], P = .041). 13 In the future, electrical stimulation of the PC could potentially be offered for treatment of epilepsy. 14
The recently renewed attention on the role of the PC in epilepsy and the technological advances allowing for successful removal, ablation, and potential neuromodulation may be important in our goal of quelling the area tempesta and creating “la quiete dopo la tempesta” (quiet after the storm, ie, seizure freedom).
