Same same but different: A Web-based deep learning application revealed classifying features for the histopathologic distinction of cortical malformations.

OBJECTIVE: The microscopic review of hematoxylin-eosin-stained images of focal cortical dysplasia type IIb and cortical tuber of tuberous sclerosis complex remains challenging. Both entities are distinct subtypes of human malformations of cortical development that share histopathological features consisting of neuronal dyslamination with dysmorphic neurons and balloon cells. We trained a convolutional neural network (CNN) to classify both entities and visualize the results. Additionally, we propose a new Web-based deep learning application as proof of concept of how deep learning could enter the pathologic routine. METHODS: A digital processing pipeline was developed for a series of 56 cases of focal cortical dysplasia type IIb and cortical tuber of tuberous sclerosis complex to obtain 4000 regions of interest and 200 000 subsamples with different zoom and rotation angles to train a neural network. Guided gradient-weighted class activation maps (Guided Grad-CAMs) were generated to visualize morphological features used by the CNN to distinguish both entities. RESULTS: Our best-performing network achieved 91% accuracy and 0.88 area under the receiver operating characteristic curve at the tile level for an unseen test set. Novel histopathologic patterns were found through the visualized Guided Grad-CAMs. These patterns were assembled into a classification score to augment decision-making in routine histopathology workup. This score was successfully validated by 11 expert neuropathologists and 12 nonexperts, boosting nonexperts to expert level performance. SIGNIFICANCE: Our newly developed Web application combines the visualization of whole slide images with the possibility of deep learning-aided classification between focal cortical dysplasia IIb and tuberous sclerosis complex. This approach will help to introduce deep learning applications and visualization for the histopathologic diagnosis of rare and difficult-to-classify brain lesions.

Efficacy and safety of Everolimus in children with TSC - associated epilepsy - Pilot data from an open single-center prospective study.

BACKGROUND: Epilepsy occurs in up to 90 % of all individuals with tuberous sclerosis complex (TSC). In 67 % disease onset is during childhood. In ≥ 50 % seizures are refractory to currently available treatment options. The mTOR-Inhibitor Everolimus (Votubia®) was approved for the treatment of subependymal giant cell astrocytoma (SEGA) and renal angiomyolipoma (AML) in Europe in 2011. It's anticonvulsive/antiepileptic properties are promising, but evidence is still limited. Study aim was to evaluate the efficacy and safety of Everolimus in children and adolescents with TSC-associated epilepsies. METHODS: Inclusion-criteria of this investigator-initiated, single-center, open, prospective study were: 1) the ascertained diagnosis of TSC; 2) age ≤ 18 years; 3) treatment indication for Votubia® according to the European Commission guidelines; 4) drug-resistant TSC-associated epilepsy, 5) prospective continuous follow-up for at least 6 months after treatment initiation and 6) informed consent to participate. Votubia® was orally administered once/day, starting with 4.5 mg/m2 and titrated to achieve blood trough concentrations between 5 and 15 ng/ml. Primary endpoint was the reduction in seizure frequency of ≥ 50 % compared to baseline. RESULTS: Fifteen patients (nine male) with a median age of six (range; 1-18) years fulfilled the inclusion criteria. 26 % (4/15) had TSC1, 66 % (10/15) had TSC2 mutations. In one patient no mutation was found. Time of observation after treatment initiation was median 22 (range; 6-50) months. At last observation, 80 % (12/15) of the patients were responders, 58 % of them (7/12) were seizure free. The overall reduction in seizure frequency was 60 % in focal seizures, 80 % in generalized tonic clonic seizures and 87 % in drop attacks. The effect of Everolimus was seen already at low doses, early after treatment initiation. Loss of efficacy over time was not observed. Transient side effects were seen in 93 % (14/15) of the patients. In no case the drug had to be withdrawn. CONCLUSION: Everolimus seems to be an effective treatment option not only for SEGA and AML, but also for TSC-related epilepsies. Although there are potential serious side effects, treatment was tolerated well by the majority of patients, provided that patients are under close surveillance of epileptologists who are familiar with immunosuppressive agents.

Improvement of language development after successful hemispherotomy.

PURPOSE: To investigate language development after functional hemispherotomy and to evaluate prognostic factors for (un-)favourable outcomes. METHODS: Children and adolescents who had vertical perithalamic hemispherotomy at the Medical University Wien (MUW) paediatric epilepsy centre were identified from a prospectively maintained database. Inclusion criteria were: complete clinical, neurophysiological and neuropsychological data, seizure freedom and a minimum follow-up of 12 months after surgery. The language quotients (LQ) prior to surgery and at last follow-up were calculated for each child. In addition, associations between pre- to post-surgical changes in LQ and the following variables were examined: age at epilepsy-onset, age at surgery and duration of epilepsy prior to surgery, aetiology, side of surgery, interictal EEG including sleep organization before and 12 months after surgery and antiepileptic-drug (AED) withdrawal state at last follow-up. Analyses were carried out in SPSS version 20.0 (SPSS Inc., Chicago, IL, USA). Nonparametric Wilcoxon and chi-square tests were applied, as required. RESULTS: Data from 28 children (14 female) were analyzed. The median age at epilepsy surgery was 64.5 months. The median follow-up after surgery was 3.0 years (±2.6 years, range 12 months to 12 years). Significant gains in LQs at last follow-up were found in 31% of the children (p=0.008). Short disease duration prior to surgery, acquired pathology, lack of epileptiform EEG discharges in the contralateral hemisphere and/or normalization of EEG sleep patterns after surgery, and successful AED withdrawal were linked to favourable language outcomes. CONCLUSION: Successful and early hemispherotomy results in improvement of language function in the intact hemisphere.

Vertical perithalamic hemispherotomy: a single-center experience in 40 pediatric patients with epilepsy.

PURPOSE: The current concept for hemispherotomy includes various lateral techniques and the vertical perithalamic hemispherotomy introduced by Delalande in 1992. We have chosen the vertical approach because of advantages that possibly influence outcome: the possibility to completely disconnect the hemisphere at the level of the thalamus obviating both the need to resect the insula and the need to open and dissect the subarachnoid space of the Sylvian fissure. METHODS: We retrospectively analyzed prospectively collected data of all patients who underwent vertical hemispherotomy at the Vienna pediatric epilepsy center. Seizure outcome was classified according to the International League Against Epilepsy (ILAE) proposal 2001. KEY FINDINGS: Follow-up data of 40 patients (22 male/18 female; median age 5.5 years; range 4.4 months to 20.1 years) were analyzed. Hemispherotomy was left in 26 and right in 14 patients. The underlying pathology was ischemic vascular in 19, malformation of cortical development (MCD) in 11, and other pathology in 10. No serious intraoperative complications were encountered. Only two infants (5.0%) needed blood replacement. There was one death on the fourth day after surgery caused by intractable hyponatremic brain edema. Three patients developed cerebrospinal fluid (CSF) disturbances, but only one needed a permanent ventriculoperitoneal (VP) shunt (2.5%). For outcome analysis we included 37 of 40 children with at least 12 months of follow-up. Thirty-four (91.9%) of 37 children were seizure-free (class 1a) after a median follow-up time of 3.7 years (range 12 month to 14.8 years). SIGNIFICANCE: We confirm the efficacy and safety of vertical parasagittal hemispherotomy as described by Delalande in a consecutive series of patients treated at our center since 1998. In addition, complete disconnection of the hemisphere in patients with MCD and/or patients with significant involvement of the insula was possible without the complications usually reported with other techniques.

Disconnective surgery in posterior quadrantic epilepsy: experience in a consecutive series of 10 patients.

OBJECT: Outcomes following functional hemispherotomy in patients with drug-resistant epilepsy have been well described. However, studies reporting long-term longitudinal outcomes after subhemispheric disconnective epilepsy surgery are still limited. METHODS: The authors conducted a retrospective review of prospectively collected data of 10 children who underwent temporoparietooccipital (TPO) disconnective surgery at the Vienna Pediatric Epilepsy Center. RESULTS: There were 3 males and 7 females (median age 8.7 years; range 4.2-22.1 years). The affected hemisphere was the left in 3 patients and the right in 7. The patients' median age at seizure onset was 3.0 years (range 0.2-8.3 years). The median duration of epilepsy before surgery was 5.2 years (range 1.3-17.2 years). The underlying pathology was TPO malformation of cortical development in 5 patients, and venous infarction, posterior hemispheric quadrant atrophy, Sturge-Weber syndrome, cortical involvement of a systemic lupus erythematosus, and gliosis after cerebral tumor treatment in 1 each. In 6 children, a pure TPO disconnection was performed; in 2 patients, the temporal lobe was resected and parietooccipital disconnection was performed. The 2 remaining patients had had previous epilepsy surgery that was extended to a TPO disconnection: disconnection of the occipital lobe (n = 1) and resection of the temporal lobe (n = 1). The authors encountered no complications while performing surgery. No patient needed blood replacement therapy. No patient developed CSF disturbances that warranted treatment. Nine of 10 patients are currently seizure free since surgery (Wieser Class 1a) at a median follow-up time of 2.1 years (range 4 months to 8.1 years). CONCLUSIONS: Temporoparietooccipital disconnection is a safe and effective motor-sparing epilepsy surgery in selected cases. Technical adjuncts facilitate a better intraoperative visualization and orientation, thereby enabling a less invasive approach than previously suggested.