Interpretable surface-based detection of focal cortical dysplasias: a Multi-centre Epilepsy Lesion Detection study

One outstanding challenge for machine learning in diagnostic biomedical imaging is algorithm interpretability. A key application is the identification of subtle epileptogenic focal cortical dysplasias (FCDs) from structural MRI. FCDs are difficult to visualize on structural MRI but are often amenable to surgical resection. We aimed to develop an open-source, interpretable, surface-based machine-learning algorithm to automatically identify FCDs on heterogeneous structural MRI data from epilepsy surgery centres worldwide. The Multi-centre Epilepsy Lesion Detection (MELD) Project collated and harmonized a retrospective MRI cohort of 1015 participants, 618 patients with focal FCD-related epilepsy and 397 controls, from 22 epilepsy centres worldwide. We created a neural network for FCD detection based on 33 surface-based features. The network was trained and cross-validated on 50% of the total cohort and tested on the remaining 50% as well as on 2 independent test sites. Multidimensional feature analysis and integrated gradient saliencies were used to interrogate network performance. Our pipeline outputs individual patient reports, which identify the location of predicted lesions, alongside their imaging features and relative saliency to the classifier. On a restricted 'gold-standard' subcohort of seizure-free patients with FCD type IIB who had T1 and fluid-attenuated inversion recovery MRI data, the MELD FCD surface-based algorithm had a sensitivity of 85%. Across the entire withheld test cohort the sensitivity was 59% and specificity was 54%. After including a border zone around lesions, to account for uncertainty around the borders of manually delineated lesion masks, the sensitivity was 67%. This multicentre, multinational study with open access protocols and code has developed a robust and interpretable machine-learning algorithm for automated detection of focal cortical dysplasias, giving physicians greater confidence in the identification of subtle MRI lesions in individuals with epilepsy.

LGG-09. A Nationwide Service Evaluation of Safety, Radiologic and Visual Outcome Refining Bevacizumab-based Treatments in Children with Progressive Low-Grade Glioma

INTRODUCTION: Bevacizumab/Irinotecan is currently 3rd-line treatment in the UK for progressive Paediatric Low-Grade Glioma (PLGG) based on limited evidence. A nationwide service evaluation was conducted to review its safety and efficacy amongst a larger cohort.

METHODS: Data from children with PLGG receiving Bevacizumab-based Treatments (BBT) from 11 UK Centres (2009-2020) were reviewed. Radiological and visual outcomes were based on standardized measurements. Clinical-radiological correlation was investigated. Time to progression from BBT stop, progression free-survival (PFS) curves and multivariate analysis of prognostic factors (p 0.05) were performed.

RESULTS: 88 children with PLGG (88% OPG, 24% NF1) had BBT for radiological (43%), visual (20%) or combined (27%) progression, after 40 months (median) from diagnosis. Amongst OPG cases, visual acuity (VA) per eye (better/worse) before BBT was logMAR 0.0-0.3 (23/7) 0.3 - 1.0 (27/20), > 1.0 (14/18) and LP/NLP (8/27), with 19/8 children respectively blind (LP/NLP) in one or both eyes. Bevacizumab 10 mg/kg every 14 days (median 24 doses) was given as 3rd line+ with Irinotecan (85%) or alongside 1st/2nd line chemo (15%) leading to remarkable radiological (88%) and visual (74%) responses (stable or improved) within 3-6 months, with limited toxicity. 12% progressed on treatment, and 8% died unrelated to BBT. After initial response 65% progressed at a median of 8 months (4 - 23) after BBT, resulting in 3-year-all-causes-PFS of 16% and 3-yr-visual-PFS of 45% from start of BBT. Visual concordance with MRI was poor (36%) but increases (47%) when better-eye determines visual outcome. Lack of NF1 and diencephalic syndrome (DS) at presentation were independent negative prognostic factors for PFS.

CONCLUSIONS: A remarkable but transient effect of BBT has been confirmed. Visual > radiological responses can be sustained after BBT. Variations in current BBT strategies justifies further research, including the potential upfront use alongside conventional first-line chemotherapy as sight-saving strategy.

Re-Examining the Cochlea in Branchio-Oto-Renal Syndrome: Genotype-Phenotype Correlation

BACKGROUND AND PURPOSE

Temporal bone imaging plays an important role in the work-up of branchio-oto-renal syndrome. Previous reports have suggested that the unwound or offset cochlea is a highly characteristic marker for branchio-oto-renal syndrome. Our goals were to examine the prevalence of this finding in a branchio-oto-renal syndrome cohort and analyze genetic-phenotypic associations not previously established.

MATERIALS AND METHODS

This multicenter retrospective study included 38 ears in 19 unrelated individuals with clinically diagnosed branchio-oto-renal syndrome and confirmed mutations in the EYA1 or SIX1 genes. Two blinded neuroradiologists independently reviewed and documented temporal bone imaging findings in 13 categories for each ear. Imaging phenotypes were correlated with genotypes.

RESULTS

There was excellent interrater agreement for all 13 phenotypic categories (κ ≥ 0.80). Of these, 9 categories showed statistically significant differences between patients with EYA1-branchio-oto-renal syndrome and SIX1-branchio-oto-renal syndrome. Cochlear offset was present in 100% of patients with EYA1-branchio-oto-renal syndrome, but in only 1 ear (12.5%) among patients with SIX1-branchio-oto-renal syndrome. A short thorny appearance of the cochlear apical turn was observed in most patients with SIX1-branchio-oto-renal syndrome.

CONCLUSIONS

An offset cochlea is associated with the EYA1-branchio-oto-renal syndrome genotype. The SIX1-branchio-oto-renal syndrome genotype is associated with a different cochlear phenotype that almost always is without offset and has a short thorny tip as the apical turn. Therefore, cochlear offset is not a characteristic marker for all patients with branchio-oto-renal syndrome. The lack of a cochlear offset in a patient with clinically suspected branchio-oto-renal syndrome does not exclude the diagnosis and, in fact, may be predictive of the SIX1 genotype.

A 40-Year Cohort Study of Evolving Hypothalamic Dysfunction in Infants and Young Children (<3 years) with Optic Pathway Gliomas

Simple Summary

Serious, poorly understood health issues affect young children with optic pathway tumours. We studied the risk of developing life-limiting hormonal, metabolic, and neurobehavioural disorders by tumour position, recurrence, and treatment, in those diagnosed under 3 years. We found the highest risk for future complex health issues in those presenting with failure to thrive, under one year of age, and/or a tumour involving a deep midbrain area called the hypothalamus. Time, repeated tumour growth, and salvage therapies (radiotherapy, surgery) contributed. We highlight the need for a better understanding of tumour-induced midbrain injury and for neurobehavioural and hormonal support to inform rehabilitation in the growing years, during and beyond cure, to optimise quality of life for these survivors and their families. This might inform oncology treatment strategies and determine new areas for support and collaborative neuroscience research in this high survival group.

Abstract

Despite high survival, paediatric optic pathway hypothalamic gliomas are associated with significant morbidity and late mortality. Those youngest at presentation have the worst outcomes. We aimed to assess presenting disease, tumour location, and treatment factors implicated in the evolution of neuroendocrine, metabolic, and neurobehavioural morbidity in 90 infants/children diagnosed before their third birthday and followed-up for 9.5 years (range 0.5–25.0). A total of 52 (57.8%) patients experienced endo-metabolic dysfunction (EMD), the large majority (46) of whom had hypothalamic involvement (H+) and lower endocrine event-free survival (EEFS) rates. EMD was greatly increased by a diencephalic syndrome presentation (85.2% vs. 46%, p = 0.001)), H+ (OR 6.1 95% CI 1.7–21.7, p 0.005), radiotherapy (OR 16.2, 95% CI 1.7–158.6, p = 0.017) and surgery (OR 4.8 95% CI 1.3–17.2, p = 0.015), all associated with anterior pituitary disorders. Obesity occurred in 25% of cases and was clustered with the endocrinopathies. Neurobehavioural deficits occurred in over half (52) of the cohort and were associated with H+ (OR 2.5 95% C.I. 1.1–5.9, p = 0.043) and radiotherapy (OR 23.1 C.I. 2.9–182, p = 0.003). Very young children with OPHG carry a high risk of endo-metabolic and neurobehavioural comorbidities which deserve better understanding and timely/parallel support from diagnosis to improve outcomes. These evolve in complex, hierarchical patterns over time whose aetiology appears predominantly determined by injury from the hypothalamic tumour location alongside adjuvant treatment strategies.

Spectrum of Neuroradiologic Findings Associated with Monogenic Interferonopathies

The genetic interferonopathies are a heterogeneous group of disorders thought to be caused by the dysregulated expression of interferons and are now commonly considered in the differential diagnosis of children presenting with recurrent or persistent inflammatory phenotypes. With emerging therapeutic options, recognition of these disorders is increasingly important, and neuroimaging plays a vital role. In this article, we discuss the wide spectrum of neuroradiologic features associated with monogenic interferonopathies by reviewing the literature and illustrate these with cases from our institutions. These cases include intracerebral calcifications, white matter T2 hyperintensities, deep WM cysts, cerebral atrophy, large cerebral artery disease, bilateral striatal necrosis, and masslike lesions. A better understanding of the breadth of the neuroimaging phenotypes in conjunction with clinical and laboratory findings will enable earlier diagnosis and direct therapeutic strategies.

The role of preoperative diffusion tensor imaging in predicting and improving functional outcome in pediatric patients undergoing epilepsy surgery: a systematic review

OBJECTIVE

Diffusion tensor imaging (DTI) is a useful neuroimaging technique for surgical planning in adult patients. However, no systematic review has been conducted to determine its utility for pre-operative analysis and planning of Pediatric Epilepsy surgery. We sought to determine the benefit of pre-operative DTI in predicting and improving neurological functional outcome after epilepsy surgery in children with intractable epilepsy.

METHODS

A systematic review of articles in English using PubMed, EMBASE and Scopus databases, from inception to January 10, 2020 was conducted. All studies that used DTI as either predictor or direct influencer of functional neurological outcome (motor, sensory, language and/or visual) in pediatric epilepsy surgical candidates were included. Data extraction was performed by two blinded reviewers. Risk of bias of each study was determined using the QUADAS 2 Scoring System.

RESULTS

13 studies were included (6 case reports/series, 5 retrospective cohorts, and 2 prospective cohorts) with a total of 229 patients. Seven studies reported motor outcome; three reported motor outcome prediction with a sensitivity and specificity ranging from 80 to 85.7 and 69.6 to 100%, respectively; four studies reported visual outcome. In general, the use of DTI was associated with a high degree of favorable neurological outcomes after epilepsy surgery.

CONCLUSION

Multiple studies show that DTI helps to create a tailored plan that results in improved functional outcome. However, more studies are required in order to fully assess its utility in pediatric patients. This is a desirable field of study because DTI offers a non-invasive technique more suitable for children.

ADVANCES IN KNOWLEDGE

This systematic review analyses, exclusively, studies of pediatric patients with drug-resistant epilepsy and provides an update of the evidence regarding the role of DTI, as part of the pre-operative armamentarium, in improving post-surgical neurological sequels and its potential for outcome prediction.

Narrative review of epilepsy: getting the most out of your neuroimaging

Neuroimaging represents an important step in the evaluation of pediatric epilepsy. The crucial role of brain imaging in the diagnosis, follow-up and presurgical assessment of patients with epilepsy is noted and has to be familiar to all neuroradiologists and trainees approaching pediatric brain imaging. Morphological qualitative imaging shows the majority of cerebral lesions/alterations underlying focal epilepsy and can highlight some features which are useful in the differential diagnosis of the different types of epilepsy. Recent advances in MRI acquisitions including diffusion-weighted imaging (DWI), post-acquisition image processing techniques, and quantification of imaging data are increasing the accuracy of lesion detection during the last decades. Functional MRI (fMRI) can be really useful and helps to identify cortical eloquent areas that are essential for language, motor function, and memory, and diffusion tensor imaging (DTI) can reveal white matter tracts that are vital for these functions, thus reducing the risk of epilepsy surgery causing new morbidities. Also positron emission tomography (PET), single photon emission computed tomography (SPECT), simultaneous electroencephalogram (EEG) and fMRI, and electrical and magnetic source imaging can be used to assess the exact localization of epileptic foci and help in the design of intracranial EEG recording strategies. The main role of these "hybrid" techniques is to obtain quantitative and qualitative informations, a necessary step to evaluate and demonstrate the complex relationship between abnormal structural and functional data and to manage a "patient-tailored" surgical approach in epileptic patients.

FP2-7 Is whole spine imaging necessary in the evaluation of children with lumbosacral lipoma?

Objectives

Paediatric lumbosacral lipomas are commonly assessed by whole spine MRI, which is costly, time consuming and reduces imaging detail at the region of interest. This study evaluates the utility of whole spine vs local spine MRI in paediatric lumbosacral lipoma (LSL).

Design

Single centre, retrospective, radiological review.

Subjects

119 patients (77F:46M) aged 0.8 years (0.2–3.9) (median ±IQR) with complex LSL who had whole spine MRI between 2001–2017.

Methods

Lumbosacral lipomas were identified from a prospectively collected database. Cases where whole spine MRI was available were included. Type of lipoma and any associated spinal anomalies were recorded.

Results

347 patients had LSL, of which 119 (34%) patients had whole spine MRI. 3% were excluded due diagnosis of fatty filum terminale. 13% scans were unavailable. 119 patients were studied: 77F:46M aged 0.8 years at scan. Type of LML: transitional 36%; caudal 18%; dorsal 19%; unclassified 26%. Additional imaging abnormalities included: syrinx 18%; Chiari 1%; dermal sinus tract 13%; vertebral segmentation 13%; other 2%. None were associated with clinical symptoms nor required surgical treatment. Investigating the cost-benefit analysis of modality, lumbar spine required less MR time and cost less (40 vs 20 mins; £228.69 vs £282.39).

Conclusions

The low incidence of clinically relevant secondary lesions suggests that lumbosacral MRI only is necessary in the assessment of LSL. Imaging quality of the ROI can be optimised and MRI time and costs reduced.

MACF1 Mutations Encoding Highly Conserved Zinc-Binding Residues of the GAR Domain Cause Defects in Neuronal Migration and Axon Guidance

To date, mutations in 15 actin- or microtubule-associated genes have been associated with the cortical malformation lissencephaly and variable brainstem hypoplasia. During a multicenter review, we recognized a rare lissencephaly variant with a complex brainstem malformation in three unrelated children. We searched our large brain-malformation databases and found another five children with this malformation (as well as one with a less severe variant), analyzed available whole-exome or -genome sequencing data, and tested ciliogenesis in two affected individuals. The brain malformation comprised posterior predominant lissencephaly and midline crossing defects consisting of absent anterior commissure and a striking W-shaped brainstem malformation caused by small or absent pontine crossing fibers. We discovered heterozygous de novo missense variants or an in-frame deletion involving highly conserved zinc-binding residues within the GAR domain of MACF1 in the first eight subjects. We studied cilium formation and found a higher proportion of mutant cells with short cilia than of control cells with short cilia. A ninth child had similar lissencephaly but only subtle brainstem dysplasia associated with a heterozygous de novo missense variant in the spectrin repeat domain of MACF1. Thus, we report variants of the microtubule-binding GAR domain of MACF1 as the cause of a distinctive and most likely pathognomonic brain malformation. A gain-of-function or dominant-negative mechanism appears likely given that many heterozygous mutations leading to protein truncation are included in the ExAC Browser. However, three de novo variants in MACF1 have been observed in large schizophrenia cohorts.