Functional magnetic resonance imaging has become a critical research tool for evaluating brain function during active tasks and resting states. This has improved our understanding of developmental ...trajectories in children as well as the plasticity of neural networks in disease states. In the clinical setting, functional maps of eloquent cortex in patients with brain lesions and/or epilepsy provides crucial information for presurgical planning. Although children are inherently challenging to scan in this setting, preparing them appropriately and providing adequate resources can help achieve useful clinical data. This article will review the basic underlying physiologic aspects of functional magnetic resonance imaging, review clinically relevant research applications, describe known validation data compared to gold standard techniques and detail future directions of this technology.
•Three SISCOM patterns were seen in pediatric temporal lobe epilepsy and unilateral anterior hyperperfusion is most identified.•Type of focal cortical dysplasia and presence of hippocampal sclerosis ...were significant determinants of SISCOM patterns.•Lobar concordance of SISCOM and ictal and interictal scalp EEG significantly correlated with postoperative outcomes.•There was no significant correlation between SISCOM patterns and one-year postoperative outcomes.
We evaluated SISCOM patterns and their relationship with surgical outcome in children with temporal lobe epilepsy (TLE) who had undergone a temporal lobe surgery.
This was an observational study evaluating SISCOM patterns in 40 children with TLE. We classified SISCOM patterns into 4 categories; (i) unilateral anteromesial and/or anterolateral temporal pattern; (ii) unilateral anteromesial and/or anterolateral temporal plus posterior extension pattern; (iii) bilateral anteromesial and/or anterolateral temporal pattern; and (iv) atypical pattern. Determinants of SISCOM pattern and correlation between postoperative outcomes and SISCOM patterns were evaluated.
Pattern (i), (ii), (iii), and (iv) were identified in 10 (25%), 14 (35%), 0 (0%), and 16 (40%) patients, respectively. There was no significant correlation between patterns and postoperative outcomes. SISCOM patterns significantly associated with the presence of hippocampal sclerosis and type of focal cortical dysplasia (p-value = 0.048 and 0.036, respectively). Patients with HS had 5 times the odds of having unilateral temporal pattern, compared to patients with other neuropathology (OR = 5, 95% CI 0.92 to 27.08). Patients with FCD type 2 had 9.71 times the odds of having atypical pattern, compared to patients with other types of FCD (OR = 9.71, 95% CI 0.92 to 103.04). Lobar concordance of SISCOM and ictal and interictal scalp EEG significantly correlated with postoperative outcomes (p-value = 0.018 and 0.013, respectively).
Three SISCOM patterns were seen. Patients with HS had increased odds of having unilateral temporal pattern while patients with FCD type 2 had increased odds of having atypical pattern. However, there was no significant correlation between SISCOM patterns and postoperative outcomes. Lobar concordance of SISCOM and ictal and interictal scalp EEG significantly correlated with postoperative outcome.
This study shows that the distribution of SISCOM patterns and their relationship with postoperative outcomes in children with TLE are different from adult population. Besides, SISCOM may add only limited diagnostic and prognostic information in children with drug-resistant TLE undergoing epilepsy surgery. Further evaluation to identify patient populations that may benefit from SISCOM is desirable.
•Subtraction ictal SPECT coregistered to MRI (SISCOM) sensitivity and specificity in identifying the epileptogenic zone (EZ) was 64.8% and 40.7% in children with drug-resistant epilepsy, ...respectively.•Injection latency and seizure duration did not significantly impact SISCOM performance.•Further study of factors contributing to SISCOM performance is needed.
We evaluated the impact of radiotracer injection latency and post-injection seizure duration on subtraction ictal SPECT co-registered to MRI (SISCOM) test performance in identifying the epileptogenic zone (EZ) in children with drug-resistant epilepsy who had undergone a resective epilepsy surgery.
SISCOM concordance with the EZ was retrospectively reviewed to evaluate its performance in 113 children. The impact of radiotracer injection latency and post-injection seizure duration was evaluated for their predictive value of SISCOM localization accuracy.
The overall sensitivity and specificity of SISCOM in identifying an EZ was 64.8% (95%CI = 50.6–77.3) and 40.7% (95%CI = 28.1–54.3). The positive likelihood ratio and diagnostic odd ratio was 1.09 (95%CI = 0.80–1.48) and 1.26 (95%CI = 0.59–2.71), respectively. Logistic regression showed that injection latency and post-injection seizure duration did not significantly predict the probability of true positive SISCOM (p-value = 0.45 and 0.29, respectively).
Radiotracer injection latency and post-injection seizure duration were not shown to have a statistical significant impact on SISCOM performance in identifying the EZ.
This study demonstrates that further study of factors contributing to the performance of SISCOM in EZ identification in children is needed.
•Resection of HGM language sites is associated with worse neuropsychological outcomes.•Working memory was significantly worse in patients with resection of HGM naming sites.•Neuropsychological ...evaluation of children with epilepsy should include multiple cognitive and language domains.
Language mapping with high-gamma modulation (HGM) has compared well with electrical cortical stimulation mapping (ESM). However, there is limited prospective data about its functional validity. We compared changes in neuropsychological evaluation (NPE) performed before and 1-year after epilepsy surgery, between patients with/without resection of cortical sites showing HGM during a visual naming task.
Pediatric drug-resistant epilepsy (DRE) patients underwent pre-surgical language localization with ESM and HGM using a visual naming task. Surgical decisions were based solely on ESM results. NPE difference scores were compared between patients with/without resection of HGM naming sites using principal component (PC) analysis. Follow-up NPE scores were modeled with resection group as main effect and respective pre-surgical score as a covariate, using analysis of covariance.
Seventeen native English speakers (12 females), aged 6.5–20.2 years, were included. One year after epilepsy surgery, first PC score increased by (mean ± standard deviation) 14.4 ± 16.5 points in patients without resection, whereas it decreased by 7.6 ± 24.6 points in those with resection of HGM naming sites (p = 0.040). This PC score represented verbal comprehension, working memory, perceptual reasoning (Wechsler subscales); Woodcock-Johnson Tests of Achievement; and Peabody Picture Vocabulary Test. Subsequent analysis showed significant difference in working memory score between patients with/without resection of HGM naming sites (−15.2 points, 95% confidence limits −29.7 to −0.7, p = 0.041).
We highlight the functional consequences of resecting HGM language sites, and suggest that NPE of DRE patients should include comprehensive assessment of multiple linguistic and cognitive domains besides naming ability.
Abstract Purpose A proportion of patients with childhood and juvenile absence epilepsies (CAE, JAE) are likely to be classified as medically refractory. In view of evidence gap for the treatment of ...such patients, this series is reported to generate estimate for efficacy of vagus nerve stimulation (VNS) in this patient population. Methods Patients were identified by a chart review of all VNS recipients between January 1, 2006 and December 31, 2011. The diagnosis of CAE and JAE was based on conventional criteria. Details of demography, epilepsy phenomenology, management and outcomes were extracted. The outcome measures included reduction in daily seizure frequency measured as a percentage of pre-VNS seizure frequency and classified on International League Against Epilepsy (ILAE) outcome scale. Results Nine patients (7 CAE, 2 JAE) with a mean age of seizure onset of 5.4 years (±3.9) were identified. Mean duration of epilepsy prior to VNS implant was found to be 3.9 years (±1.4). These patients had failed a median of 5 anti-epileptic drugs before being referred for consideration of surgical treatment. After a mean follow-up of 33.9 months (±25.5, minimum 4 months), 1 patient attained complete seizure freedom (ILAE class 1), 6 had ILAE class 4 and 2 had ILAE class 5 outcomes, respectively. Mean reduction in daily seizure frequency was found to be 53.5 ± 60.3% (1-sided p -value for paired t -test = 0.04), with a 50% responder rate of 55.6%. Conclusion VNS may be considered as a therapeutic option in patients with medically refractory absence epilepsy.
This study explored predictors of response to vagus nerve stimulation in childhood-onset epilepsy. This retrospective chart review included all patients with new vagus nerve stimulator insertion ...between January 1, 2006, and December 31, 2011. Primary outcome was change in seizure frequency classified on the International League Against Epilepsy scale. Overall, 67.4% (95% confidence limits 53.3%-81.6%) of the patients had outcome of class 4 or better, and 4 patients (9.3%, 95% confidence interval 0.5%-18.1%) achieved complete seizure freedom (mean follow-up 3.5 y). Absence of magnetic resonance imaging (MRI) lesion (odds ratio 6.068, 95% confidence interval 1.214-30.329, P = .028) and duration of epilepsy before implantation (odds ratio 1.291, 95% confidence interval 1.015-1.642, P = .038) were found to be statistically significant predictors of good outcome and provided a sufficient fit to the data (area under the receiver operating characteristic curve .80, Hosmer-Lemeshow goodness of fit P = .92). This study provides preliminary evidence that nonlesional patients are significantly more likely to have better outcome with vagus nerve stimulation.
Objective: We describe the development and evaluation of a system that uses machine learning and natural language processing techniques to identify potential candidates for surgical intervention for ...drug-resistant pediatric epilepsy. The data are comprised of free-text clinical notes extracted from the electronic health record (EHR). Both known clinical outcomes from the EHR and manual chart annotations provide gold standards for the patient's status. The following hypotheses are then tested: 1) machine learning methods can identify epilepsy surgery candidates as well as physicians do and 2) machine learning methods can identify candidates earlier than physicians do. These hypotheses are tested by systematically evaluating the effects of the data source, amount of training data, class balance, classification algorithm, and feature set on classifier performance. The results support both hypotheses, with F-measures ranging from 0.71 to 0.82. The feature set, classification algorithm, amount of training data, class balance, and gold standard all significantly affected classification performance. It was further observed that classification performance was better than the highest agreement between two annotators, even at one year before documented surgery referral. The results demonstrate that such machine learning methods can contribute to predicting pediatric epilepsy surgery candidates and reducing lag time to surgery referral.
Summary
Objective
The predictive value of intraoperative electrocorticography (ECoG) in pediatric epilepsy surgery is unknown. In a population of children undergoing ECoG followed typically by ...invasive extraoperative monitoring (IEM) and resection, we aimed to determine the relationship between frequent ECoG abnormalities and the seizure onset zone and outcome after resection.
Methods
We retrospectively identified 103 children with preresection ECoG of sufficient technical quality. ECoG records were scored based on electrode location and frequency, blinded to the seizure‐onset zone and outcome. Electrographic seizure and spike locations were identified. Locations of seizures and spike populations were then compared to the location of seizure‐onset zone defined by IEM using subdural electrodes and resection margin.
Results
Electrographic seizures were identified in 11 (11%) of 103 patients. A spike population of one or more was noted in 79 (77%) of 103 patients. In 50 (63%) of 79 patients, spike populations correlated with seizure‐onset zone location. The overall surgical outcome was good (ILAE 1 to 3) in 53 (52%) of 101 patients. Outcome was good in seven (78%) of nine patients when electrographic seizure location was resected. The best outcomes were obtained with resection of both the seizure‐onset zone and ECoG abnormalities to include seizures and spike locations (22/33 good outcome, 67%, p = 0.008). There was a significantly better outcome in children with complete resection of ECoG‐identified spike populations (14/26, 62% good outcome) compared to when none were resected (4/14, 29%, p = 0.043).
Significance
Electrographic seizures and frequent spikes are frequently seen on pre‐resection ECoG in children. The brain locations corresponding to these discharges are highly concordant with the seizure‐onset zone; resection of these regions is correlated with good seizure outcome. Further research is needed to design interventions that increase the reliability of ECoG prediction of the epileptogenic zone and obviate the need for IEM.
Highlights • 29% of children with an MRI-identified FCD did not have epilepsy • The prevalence of epilepsy and drug-resistant epilepsy was 71% and 33%, respectively • Lobar location predicted the ...likelihood of developing epilepsy • Family history of seizures predicted the likelihood of developing epilepsy • Age of seizure onset predicted the chance of developing drug-resistance
•Fosphenytoin (FOS) pre-medication decreased electrical stimulation induced seizures.•FOS pre-medication increased electrical stimulation temporal language threshold.•FOS (10–12 mg/kg) prevented ...seizures but did not affect temporal language threshold.•FOS pre-medication may be considered before electrical stimulation mapping.
We studied the effect of fosphenytoin (FOS) pre-medication on the incidence and thresholds of after-discharges (ADs), seizures, and functional responses during electrical stimulation mapping (ESM).
As individualized by the attending epileptologist, FOS was given intravenously at 2 mg-phenytoin-equivalents (PE)/kg/min or 150 mg-PE/min (whichever slower). Patients who received and did not receive FOS were compared for the incidence and thresholds of ADs, seizures, and functional responses.
Before ESM, 40 and 82 patients respectively were pre-medicated/not pre-medicated with FOS. The incidence of ESM-induced seizures was significantly lower in FOS pre-medicated patients (22.5% vs. 42.7%, p = 0.044), whereas temporal language threshold was higher (9.2 vs. 6.5 mA, p = 0.032). FOS was more efficacious in preventing ESM-induced seizures in patients with symptomatogenic zone ipsilateral to the side of ESM.
Although FOS dose had no significant effect on minimum language, minimum motor, or AD thresholds; seizure and temporal language thresholds showed trends approaching significance, intersecting at 12.2 mg-PE/kg. The incidence of ESM-induced seizures was significantly lower in those who received FOS at a dose of ≤12 mg/kg (9.1%) compared to those who did not receive any FOS (42.7%, p = 0.046), while the temporal language thresholds were not significantly different (6.3 vs. 6.5 mA, p = 0.897).
This study provides class III evidence that FOS pre-medication before ESM decreases the incidence of ESM-induced seizures, but increases temporal language threshold. FOS pre-medication may thus be considered before ESM. Future studies should prospectively verify these observations and characterize dose-response relationships.
PDF
