Characterisation of paediatric brain tumours by their MRS metabolite profiles

1H-magnetic resonance spectroscopy (MRS) has the potential to improve the noninvasive diagnostic accuracy for paediatric brain tumours. However, studies analysing large, comprehensive, multicentre datasets are lacking, hindering translation to widespread clinical practice. Single-voxel MRS (point-resolved single-voxel spectroscopy sequence, 1.5 T: echo time [TE] 23-37 ms/135-144 ms, repetition time [TR] 1500 ms; 3 T: TE 37-41 ms/135-144 ms, TR 2000 ms) was performed from 2003 to 2012 during routine magnetic resonance imaging for a suspected brain tumour on 340 children from five hospitals with 464 spectra being available for analysis and 281 meeting quality control. Mean spectra were generated for 13 tumour types. Mann-Whitney U-tests and Kruskal-Wallis tests were used to compare mean metabolite concentrations. Receiver operator characteristic curves were used to determine the potential for individual metabolites to discriminate between specific tumour types. Principal component analysis followed by linear discriminant analysis was used to construct a classifier to discriminate the three main central nervous system tumour types in paediatrics. Mean concentrations of metabolites were shown to differ significantly between tumour types. Large variability existed across each tumour type, but individual metabolites were able to aid discrimination between some tumour types of importance. Complete metabolite profiles were found to be strongly characteristic of tumour type and, when combined with the machine learning methods, demonstrated a diagnostic accuracy of 93% for distinguishing between the three main tumour groups (medulloblastoma, pilocytic astrocytoma and ependymoma). The accuracy of this approach was similar even when data of marginal quality were included, greatly reducing the proportion of MRS excluded for poor quality. Children's brain tumours are strongly characterised by MRS metabolite profiles readily acquired during routine clinical practice, and this information can be used to support noninvasive diagnosis. This study provides both key evidence and an important resource for the future use of MRS in the diagnosis of children's brain tumours.

Applying machine learning classifiers to automate quality assessment of paediatric dynamic susceptibility contrast (DSC-) MRI data

OBJECTIVE

Investigate the performance of qualitative review (QR) for assessing dynamic susceptibility contrast (DSC-) MRI data quality in paediatric normal brain and develop an automated alternative to QR.

METHODS

1027 signal-time courses were assessed by Reviewer 1 using QR. 243 were additionally assessed by Reviewer 2 and % disagreements and Cohen's κ (κ) were calculated. The signal drop-to-noise ratio (SDNR), root mean square error (RMSE), full width half maximum (FWHM) and percentage signal recovery (PSR) were calculated for the 1027 signal-time courses. Data quality thresholds for each measure were determined using QR results. The measures and QR results trained machine learning classifiers. Sensitivity, specificity, precision, classification error and area under the curve from a receiver operating characteristic curve were calculated for each threshold and classifier.

RESULTS

Comparing reviewers gave 7% disagreements and κ = 0.83. Data quality thresholds of: 7.6 for SDNR; 0.019 for RMSE; 3 s and 19 s for FWHM; and 42.9 and 130.4% for PSR were produced. SDNR gave the best sensitivity, specificity, precision, classification error and area under the curve values of 0.86, 0.86, 0.93, 14.2% and 0.83. Random forest was the best machine learning classifier, giving sensitivity, specificity, precision, classification error and area under the curve of 0.94, 0.83, 0.93, 9.3% and 0.89.

CONCLUSION

The reviewers showed good agreement. Machine learning classifiers trained on signal-time course measures and QR can assess quality. Combining multiple measures reduces misclassification.

ADVANCES IN KNOWLEDGE

A new automated quality control method was developed, which trained machine learning classifiers using QR results.

Metabolite selection for machine learning in childhood brain tumour classification

MRS can provide high accuracy in the diagnosis of childhood brain tumours when combined with machine learning. A feature selection method such as principal component analysis is commonly used to reduce the dimensionality of metabolite profiles prior to classification. However, an alternative approach of identifying the optimal set of metabolites has not been fully evaluated, possibly due to the challenges of defining this for a multi-class problem. This study aims to investigate metabolite selection from in vivo MRS for childhood brain tumour classification. Multi-site 1.5 T and 3 T cohorts of patients with a brain tumour and histological diagnosis of ependymoma, medulloblastoma and pilocytic astrocytoma were retrospectively evaluated. Dimensionality reduction was undertaken by selecting metabolite concentrations through multi-class receiver operating characteristics and compared with principal component analysis. Classification accuracy was determined through leave-one-out and k-fold cross-validation. Metabolites identified as crucial in tumour classification include myo-inositol (P < 0.05, AUC = 0 . 81 ± 0 . 01 ), total lipids and macromolecules at 0.9 ppm (P < 0.05, AUC = 0 . 78 ± 0 . 01 ) and total creatine (P < 0.05, AUC = 0 . 77 ± 0 . 01 ) for the 1.5 T cohort, and glycine (P < 0.05, AUC = 0 . 79 ± 0 . 01 ), total N-acetylaspartate (P < 0.05, AUC = 0 . 79 ± 0 . 01 ) and total choline (P < 0.05, AUC = 0 . 75 ± 0 . 01 ) for the 3 T cohort. Compared with the principal components, the selected metabolites were able to provide significantly improved discrimination between the tumours through most classifiers (P < 0.05). The highest balanced classification accuracy determined through leave-one-out cross-validation was 85% for 1.5 T 1 H-MRS through support vector machine and 75% for 3 T 1 H-MRS through linear discriminant analysis after oversampling the minority. The study suggests that a group of crucial metabolites helps to achieve better discrimination between childhood brain tumours.

European Society for Paediatric Oncology (SIOPE) MRI guidelines for imaging patients with central nervous system tumours

INTRODUCTION

Standardisation of imaging acquisition is essential in facilitating multicentre studies related to childhood CNS tumours. It is important to ensure that the imaging protocol can be adopted by centres with varying imaging capabilities without compromising image quality.

MATERIALS AND METHOD

An imaging protocol has been developed by the Brain Tumour Imaging Working Group of the European Society for Paediatric Oncology (SIOPE) based on consensus among its members, which consists of neuroradiologists, imaging scientists and paediatric neuro-oncologists. This protocol has been developed to facilitate SIOPE led studies and regularly reviewed by the imaging working group.

RESULTS

The protocol consists of essential MRI sequences with imaging parameters for 1.5 and 3 Tesla MRI scanners and a set of optional sequences that can be used in appropriate clinical settings. The protocol also provides guidelines for early post-operative imaging and surveillance imaging. The complementary use of multimodal advanced MRI including diffusion tensor imaging (DTI), MR spectroscopy and perfusion imaging is encouraged, and optional guidance is provided in this publication.

CONCLUSION

The SIOPE brain tumour imaging protocol will enable consistent imaging across multiple centres involved in paediatric CNS tumour studies.

Classification of paediatric brain tumours by diffusion weighted imaging and machine learning

To determine if apparent diffusion coefficients (ADC) can discriminate between posterior fossa brain tumours on a multicentre basis. A total of 124 paediatric patients with posterior fossa tumours (including 55 Medulloblastomas, 36 Pilocytic Astrocytomas and 26 Ependymomas) were scanned using diffusion weighted imaging across 12 different hospitals using a total of 18 different scanners. Apparent diffusion coefficient maps were produced and histogram data was extracted from tumour regions of interest. Total histograms and histogram metrics (mean, variance, skew, kurtosis and 10th, 20th and 50th quantiles) were used as data input for classifiers with accuracy determined by tenfold cross validation. Mean ADC values from the tumour regions of interest differed between tumour types, (ANOVA P < 0.001). A cut off value for mean ADC between Ependymomas and Medulloblastomas was found to be of 0.984 × 10⁻³ mm² s⁻¹ with sensitivity 80.8% and specificity 80.0%. Overall classification for the ADC histogram metrics were 85% using Naïve Bayes and 84% for Random Forest classifiers. The most commonly occurring posterior fossa paediatric brain tumours can be classified using Apparent Diffusion Coefficient histogram values to a high accuracy on a multicentre basis.

Droplet digital PCR-based detection of circulating tumor DNA from pediatric high grade and diffuse midline glioma patients

BACKGROUND

The use of liquid biopsy is of potential high importance for children with high grade (HGG) and diffuse midline gliomas (DMG), particularly where surgical procedures are limited, and invasive biopsy sampling not without risk. To date, however, the evidence that detection of cell-free DNA (cfDNA) or circulating tumor DNA (ctDNA) could provide useful information for these patients has been limited, or contradictory.

METHODS

We optimized droplet digital PCR (ddPCR) assays for the detection of common somatic mutations observed in pediatric HGG/DMG, and applied them to liquid biopsies from plasma, serum, cerebrospinal fluid (CSF), and cystic fluid collected from 32 patients.

RESULTS

Although detectable in all biomaterial types, ctDNA presented at significantly higher levels in CSF compared to plasma and/or serum. When applied to a cohort of 127 plasma specimens from 41 patients collected from 2011 to 2018 as part of a randomized clinical trial in pediatric non-brainstem HGG/DMG, ctDNA profiling by ddPCR was of limited use due to the small volumes (mean = 0.49 mL) available. In anecdotal cases where sufficient material was available, cfDNA concentration correlated with disease progression in two examples each of poor response in H3F3A_K27M-mutant DMG, and longer survival times in hemispheric BRAF_V600E-mutant cases.

CONCLUSION

Tumor-specific DNA alterations are more readily detected in CSF than plasma. Although we demonstrate the potential of the approach to assessing tumor burden, our results highlight the necessity for adequate sample collection and approach to improve detection if plasma samples are to be used.

EPEN-24. SIOP EPENDYMOMA II: CENTRAL EPENDYMOMA MANAGEMENT ADVISORY GROUP – THE UK EXPERIENCE

Paediatric Ependymoma is the second most common malignant brain tumour of childhood with approximately 50% of cases recurring. It has been described as a “surgical” disease since patients who have undergone a gross total surgical resection (GTR) have a better prognosis than those who have a subtotal resection (STR). Analysis of the UKCCSG/SIOP 1992 04 clinical trial has shown that only 49% of cases had a GTR, with 5-year survival rates for STR of 22–47% and GTR of 67–80%. As part of the SIOP II Ependymoma trial the UK established a panel of experts in the treatment of Ependymoma from Neuro-oncology, Neuro-radiology and Neuro-surgery. Meeting weekly, cases are discussed to provide a consensus on radiological review, ensuring central pathological review, trial stratification and whether further surgery should be advocated on any particular case. Evaluation of the first 68 UK patients has shown a GTR in 47/68 (69%) of patients and STR in 21/68 (31%) of patients. Following discussion at EMAG it was felt that 9/21 (43%) STR patients could be offered early second look surgery. Following this 2nd look surgery the number of cases with a GTR increased to 56/68 (82%). There has been a clear increase in the number of patients for whom a GTR has been achieved following discussion at EMAG and prior to them moving forwards with their oncological treatment. This can only have beneficial effects in decreasing their risk of tumour recurrence or CSF dissemination and also in reducing the target volume for radiotherapy.

IMG-04. RESPONSE ASSESSMENT IN PEDIATRIC HIGH-GRADE GLIOMA: RECOMMENDATIONS FROM THE RESPONSE ASSESSMENT IN PEDIATRIC NEURO-ONCOLOGY WORKING GROUP

INTRODUCTION

Response criteria for pediatric high-grade gliomas (pHGG) have varied historically and across clinical trials. Compared to adult HGG, pHGG response assessment has unique challenges. An international Response Assessment in Pediatric Neuro-Oncology (RAPNO) working group was established to develop pHGG response assessment criteria.

METHODS

Pediatric and adult neuro-oncologists, neuro-radiologists and experts in imaging informatics developed a consensus statement and established a unified response assessment for biopsy-proven pHGG, excluding DIPG. This was achieved by identifying major challenges, reviewing existing literature and current practices, and finally developing recommendations through an iterative process.

RESULTS

Categories for response assessment include complete response, partial response, minor response, stable disease and progressive disease. Refractory disease is excluded. Criteria used to determine response assessment include quantitative evaluation of measurable disease, qualitative assessment of diffusion imaging, presence or absence of new lesions, clinical status using performance score, and vascular endothelial growth factor inhibitor and/or corticosteroid use. Response is determined over 2-time points ≥ 8 weeks apart, and when progressive disease is unclear, guidance for repeat MRI imaging and/or utility of repeat biopsy is described. A number of recommendations are also given to standardize response assessment across clinical trials including MRI protocol sequence recommendations for brain and spine, definitions for measurable and non-measurable disease, and imaging time points with post-operative considerations. In addition, guidance is given for differentiating vasogenic edema versus tumor invasion in non-enhancing disease.

CONCLUSION

Consensus recommendations and response definitions have been established and, similar to other RAPNO recommendations, prospective validation in clinical trials is warranted.

Response assessment in paediatric high-grade glioma: recommendations from the Response Assessment in Pediatric Neuro-Oncology (RAPNO) working group

Response criteria for paediatric high-grade glioma vary historically and across different cooperative groups. The Response Assessment in Neuro-Oncology working group developed response criteria for adult high-grade glioma, but these were not created to meet the unique challenges in children with the disease. The Response Assessment in Pediatric Neuro-Oncology (RAPNO) working group, consisting of an international panel of paediatric and adult neuro-oncologists, clinicians, radiologists, radiation oncologists, and neurosurgeons, was established to address issues and unique challenges in assessing response in children with CNS tumours. We established a subcommittee to develop response assessment criteria for paediatric high-grade glioma. Current practice and literature were reviewed to identify major challenges in assessing the response of paediatric high-grade gliomas to various treatments. For areas in which scientific investigation was scarce, consensus was reached through an iterative process. RAPNO response assessment recommendations include the use of MRI of the brain and the spine, assessment of clinical status, and the use of corticosteroids or antiangiogenics. Imaging standards for brain and spine are defined. Compared with the recommendations for the management of adult high-grade glioma, for paediatrics there is inclusion of diffusion-weighted imaging and a higher reliance on T2-weighted fluid-attenuated inversion recovery. Consensus recommendations and response definitions have been established and, similar to other RAPNO recommendations, prospective validation in clinical trials is warranted.

Radiological Evaluation of Newly Diagnosed Non-Brainstem Pediatric High-Grade Glioma in the HERBY Phase II Trial

PURPOSE

The HERBY trial evaluated the benefit of the addition of the antiangiogenic agent Bevacizumab (BEV) to radiotherapy/temozolomide (RT/TMZ) in pediatric patients with newly diagnosed non-brainstem high-grade glioma (HGG). The work presented here aims to correlate imaging characteristics and outcome measures with pathologic and molecular data.

EXPERIMENTAL DESIGN

Radiological, pathologic, and molecular data were correlated with trial clinical information to retrospectively re-evaluate event-free survival (EFS) and overall survival (OS).

RESULTS

One-hundred thirteen patients were randomized to the RT/TMZ arm (n = 54) or the RT/TMZ+BEV (BEV arm; n = 59). The tumor arose in the cerebral hemispheres in 68 patients (Cerebral group) and a midline location in 45 cases (Midline group). Pathologic diagnosis was available in all cases and molecular data in 86 of 113. H3 K27M histone mutations were present in 23 of 32 Midline cases and H3 G34R/V mutations in 7 of 54 Cerebral cases. Total/near-total resection occurred in 44 of 68 (65%) Cerebral cases but in only 5 of 45 (11%) Midline cases (P < 0.05). Leptomeningeal metastases (27 cases, 13 with subependymal spread) at relapse were more frequent in Midline (17/45) than in Cerebral tumors (10/68, P < 0.05). Mean OS (14.1 months) and EFS (9.0 months) in Midline tumors were significantly lower than mean OS (20.7 months) and EFS (14.9 months) in Cerebral tumors (P < 0.05). Pseudoprogression occurred in 8 of 111 (6.2%) cases.

CONCLUSIONS

This study has shown that the poor outcome of midline tumors (compared with cerebral) may be related to (1) lesser surgical resection, (2) H3 K27M histone mutations, and (3) higher leptomeningeal dissemination.

WHO grade has no prognostic value in the pediatric high-grade glioma included in the HERBY trial

BACKGROUND

The World Health Organization (WHO) adult glioma grading system is questionable in pediatric high-grade gliomas (pHGGs), which are biologically distinct from adult HGGs. We took advantage of the neuropathological review data obtained during one of the largest prospective randomized pHGG trials, namely HERBY (NCT01390948), to address this issue in children with newly diagnosed non-brainstem HGG.

METHODS

HGG diagnosis was confirmed by pre-randomization, real-time central pathology review using WHO 2007 criteria, followed by a consensus review blinded to clinical factors and outcomes. We evaluated association between WHO 2007 grade and other clinical/radiological/biological characteristics and the prognostic value of WHO 2007 grade, midline location, and selected biomarkers (Ki-67 index/Olig2/CD34/EGFR/p53/H3F3A K27M mutation) on overall survival.

RESULTS

Real-time central neuropathological review was feasible in a multicenter study, with a mean time of 2.4 days, and led to the rejection of HGG diagnosis in 20 of 163 cases (12.3%). The different grading criteria and resulting WHO grade were not significantly associated with overall survival in the entire population (n = 118) or in midline and non-midline subgroups. H3F3A K27M mutation was significantly associated with poor outcome. No significant prognostic value was observed for grade, even after regrading H3F3A K27M-mutated midline glioma as grade IV (WHO 2016). Midline location and a high Ki-67 index (≥20%) were associated with poor outcome (P = 0.004 and P = 0.04, respectively). A 10% increase in Ki-67 index was associated with a hazard ratio of 1.53 (95% CI: 1.27-1.83; P < 0.0001).

CONCLUSION

Our findings suggest that WHO grade III versus IV has no prognostic value in pediatric HGG.

Evaluation of the Implementation of the Response Assessment in Neuro-Oncology Criteria in the HERBY Trial of Pediatric Patients with Newly Diagnosed High-Grade Gliomas

BACKGROUND AND PURPOSE

HERBY was a Phase II multicenter trial setup to establish the efficacy and safety of adding bevacizumab to radiation therapy and temozolomide in pediatric patients with newly diagnosed non-brain stem high-grade gliomas. This study evaluates the implementation of the radiologic aspects of HERBY.

MATERIALS AND METHODS

We analyzed multimodal imaging compliance rates and scan quality for participating sites, adjudication rates and reading times for the central review process, the influence of different Response Assessment in Neuro-Oncology criteria in the final response, the incidence of pseudoprogression, and the benefit of incorporating multimodal imaging into the decision process.

RESULTS

Multimodal imaging compliance rates were the following: diffusion, 82%; perfusion, 60%; and spectroscopy, 48%. Neuroradiologists' responses differed for 50% of scans, requiring adjudication, with a total average reading time per patient of approximately 3 hours. Pseudoprogression occurred in 10/116 (9%) cases, 8 in the radiation therapy/temozolomide arm and 2 in the bevacizumab arm (P < .01). Increased target enhancing lesion diameter was a reason for progression in 8/86 cases (9.3%) but never the only radiologic or clinical reason. Event-free survival was predicted earlier in 5/86 (5.8%) patients by multimodal imaging (diffusion, n = 4; perfusion, n = 1).

CONCLUSIONS

The addition of multimodal imaging to the response criteria modified the assessment in a small number of cases, determining progression earlier than structural imaging alone. Increased target lesion diameter, accounting for a large proportion of reading time, was never the only reason to designate disease progression.

Prospective multicentre evaluation and refinement of an analysis tool for magnetic resonance spectroscopy of childhood cerebellar tumours

BACKGROUND

A tool for diagnosing childhood cerebellar tumours using magnetic resonance (MR) spectroscopy peak height measurement has been developed based on retrospective analysis of single-centre data.

OBJECTIVE

To determine the diagnostic accuracy of the peak height measurement tool in a multicentre prospective study, and optimise it by adding new prospective data to the original dataset.

MATERIALS AND METHODS

Magnetic resonance imaging (MRI) and single-voxel MR spectroscopy were performed on children with cerebellar tumours at three centres. Spectra were processed using standard scanner software and peak heights for N-acetyl aspartate, creatine, total choline and myo-inositol were measured. The original diagnostic tool was used to classify 26 new tumours as pilocytic astrocytoma, medulloblastoma or ependymoma. These spectra were subsequently combined with the original dataset to develop an optimised scheme from 53 tumours in total.

RESULTS

Of the pilocytic astrocytomas, medulloblastomas and ependymomas, 65.4% were correctly assigned using the original tool. An optimized scheme was produced from the combined dataset correctly assigning 90.6%. Rare tumour types showed distinctive MR spectroscopy features.

CONCLUSION

The original diagnostic tool gave modest accuracy when tested prospectively on multicentre data. Increasing the dataset provided a diagnostic tool based on MR spectroscopy peak height measurement with high levels of accuracy for multicentre data.

Molecular, Pathological, Radiological, and Immune Profiling of Non-brainstem Pediatric High-Grade Glioma from the HERBY Phase II Randomized Trial

The HERBY trial was a phase II open-label, randomized, multicenter trial evaluating bevacizumab (BEV) in addition to temozolomide/radiotherapy in patients with newly diagnosed non-brainstem high-grade glioma (HGG) between the ages of 3 and 18 years. We carried out comprehensive molecular analysis integrated with pathology, radiology, and immune profiling. In post-hoc subgroup analysis, hypermutator tumors (mismatch repair deficiency and somatic POLE/POLD1 mutations) and those biologically resembling pleomorphic xanthoastrocytoma ([PXA]-like, driven by BRAF_V600E or NF1 mutation) had significantly more CD8+ tumor-infiltrating lymphocytes, and longer survival with the addition of BEV. Histone H3 subgroups (hemispheric G34R/V and midline K27M) had a worse outcome and were immune cold. Future clinical trials will need to take into account the diversity represented by the term "HGG" in the pediatric population.

Phase II, Open-Label, Randomized, Multicenter Trial (HERBY) of Bevacizumab in Pediatric Patients With Newly Diagnosed High-Grade Glioma

Purpose Bevacizumab (BEV) is approved in more than 60 countries for use in adults with recurrent glioblastoma. We evaluated the addition of BEV to radiotherapy plus temozolomide (RT+TMZ) in pediatric patients with newly diagnosed high-grade glioma (HGG). Methods The randomized, parallel group, multicenter, open-label HERBY trial ( ClinicalTrials.gov identifier: NCT01390948) enrolled patients age ≥ 3 years to ≤ 18 years with localized, centrally neuropathology-confirmed, nonbrainstem HGG. Eligible patients were randomly assigned to receive RT + TMZ (RT: 1.8 Gy, 5 days per week, and TMZ: 75 mg/m² per day for 6 weeks; 4-week treatment break; then up to 12 × 28-day cycles of TMZ [cycle 1: 150 mg/m² per day, days 1 to 5; cycles 2 to 12: 200 mg/m² per day, days 1 to 5]) with or without BEV (10 mg/kg every 2 weeks). The primary end point was event-free survival (EFS) as assessed by a central radiology review committee that was blinded to treatment. We report findings of EFS at 12 months after the enrollment of the last patient. Results One hundred twenty-one patients were enrolled (RT+TMZ [n = 59]; BEV plus RT+TMZ [n = 62]). Central radiology review committee-assessed median EFS did not differ significantly between treatment groups (RT+TMZ, 11.8 months; 95% CI, 7.9 to 16.4 months; BEV plus RT+TMZ, 8.2 months; 95% CI, 7.8 to 12.7 months; hazard ratio, 1.44; P = .13 [stratified log-rank test]). In the overall survival analysis, the addition of BEV did not reduce the risk of death (hazard ratio, 1.23; 95% CI, 0.72 to 2.09). More patients in the BEV plus RT+TMZ group versus the RT+TMZ group experienced one or more serious adverse events (n = 35 [58%] v n = 27 [48%]), and more patients who received BEV discontinued study treatment as a result of adverse events (n = 13 [22%] v n = 3 [5%]). Conclusion Adding BEV to RT+TMZ did not improve EFS in pediatric patients with newly diagnosed HGG. Our findings were not comparable to those of previous adult trials, which highlights the importance of performing pediatric-specific studies.

Application of pattern recognition techniques for classification of pediatric brain tumors by in vivo 3T 1 H-MR spectroscopy-A multi-center study

Purpose

3T magnetic resonance scanners have boosted clinical application of ¹H‐MR spectroscopy (MRS) by offering an improved signal‐to‐noise ratio and increased spectral resolution, thereby identifying more metabolites and extending the range of metabolic information. Spectroscopic data from clinical 1.5T MR scanners has been shown to discriminate between pediatric brain tumors by applying machine learning techniques to further aid diagnosis. The purpose of this multi‐center study was to investigate the discriminative potential of metabolite profiles obtained from 3T scanners in classifying pediatric brain tumors.

Methods

A total of 41 pediatric patients with brain tumors (17 medulloblastomas, 20 pilocytic astrocytomas, and 4 ependymomas) were scanned across four different hospitals. Raw spectroscopy data were processed using TARQUIN. Borderline synthetic minority oversampling technique was used to correct for the data skewness. Different classifiers were trained using linear discriminative analysis, support vector machine, and random forest techniques.

Results

Support vector machine had the highest balanced accuracy for discriminating the three tumor types. The balanced accuracy achieved was higher than the balanced accuracy previously reported for similar multi‐center dataset from 1.5T magnets with echo time 20 to 32 ms alone.

Conclusion

This study showed that 3T MRS can detect key differences in metabolite profiles for the main types of childhood tumors. Magn Reson Med 79:2359–2366, 2018. © 2017 The Authors Magnetic Resonance in Medicine published by Wiley Periodicals, Inc. on behalf of International Society for Magnetic Resonance in Medicine. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Imaging central veins in brain lesions with 3-T T2*-weighted magnetic resonance imaging differentiates multiple sclerosis from microangiopathic brain lesions

Background:

White matter lesions are frequently detected using brain magnetic resonance imaging (MRI) performed for various indications. Most are microangiopathic, but demyelination, including multiple sclerosis (MS), is an important cause; conventional MRI cannot always distinguish between these pathologies. The proportion of lesions with a central vein on 7-T T2*-weighted MRI prospectively distinguishes demyelination from microangiopathic lesions.

Objective:

To test whether 3-T T2*-weighted MRI can differentiate MS from microangiopathic brain lesions.

Methods:

A total of 40 patients were studied. Initially, a test cohort of 10 patients with MS and 10 patients with microangiopathic white matter lesions underwent 3-T T2*-weighted brain MRI. Anonymised scans were analysed blind to clinical data, and simple diagnostic rules were devised. These rules were applied to a validation cohort of 20 patients (13 with MS and 7 with microangiopathic lesions) by a blinded observer.

Results:

Within the test cohort, all patients with MS had central veins visible in >45% of brain lesions, while the rest had central veins visible in <45% of lesions. By applying diagnostic rules to the validation cohort, all remaining patients were correctly categorised.

Conclusion:

3-T T2*-weighted brain MRI distinguishes perivenous MS lesions from microangiopathic lesions. Clinical application of this technique could supplement existing diagnostic algorithms.

Response Assessment in Pediatric Neuro-Oncology: Implementation and Expansion of the RANO Criteria in a Randomized Phase II Trial of Pediatric Patients with Newly Diagnosed High-Grade Gliomas

Determination of tumor response to treatment in neuro-oncology is challenging, particularly when antiangiogenic agents are considered. Nontumoral factors (eg, blood-brain barrier disruption, edema, and necrosis) can alter contrast enhancement independent of true tumor response/progression. Furthermore, gliomas are often infiltrative, with nonenhancing components. In adults, the Response Assessment in Neuro-Oncology (RANO) criteria attempted to address these issues. No such guidelines exist yet for children. The ongoing randomized phase II trial, A Study of Avastin (bevacizumab) in Combination With Temolozomide (TMZ) and Radiotherapy in Paediatric and Adolescent Patients With High-Grade Glioma (HERBY), will establish the efficacy and safety of the antiangiogenic agent bevacizumab for the first-line treatment of newly diagnosed high-grade glioma in children (n = 121 patients, enrollment complete). The primary end point is event-free survival (tumor progression/recurrence by central review, second primary malignancy, or death). Determination of progression or response is based on predefined clinical and radiographic criteria, modeled on the RANO criteria and supported by expert pseudoprogression review and the use of standardized imaging protocols. The HERBY trial will also compare conventional MR imaging (T1-weighted and T2/fluid-attenuated inversion recovery sequences) with conventional MR imaging plus diffusion/perfusion imaging for response assessment. It is anticipated that HERBY will provide new insights into antiangiogenic-treated pediatric brain tumors. HERBY will also investigate the practicality of obtaining adequate quality diffusion/perfusion scans in a trial setting, and the feasibility of implementing standard imaging protocols across multiple sites. To date, 61/73 (83.6%) patients with available data have completed diffusion-weighted imaging (uptake of other nonconventional techniques has been limited). Harmonization of imaging protocols and techniques may improve the robustness of pediatric neuro-oncology studies and aid future trial comparability.

Fetal MRI demonstrating vein of Galen malformations in two successive pregnancies--a previously unreported occurrence

PURPOSE

Vein of Galen malformations are rare and are usually detected in utero using ultrasonography. No definite genetic predisposition has been described in the literature. We present a case with two successive pregnancies complicated by vein of Galen malformations, which were assessed using fetal MRI. The putative role of genetic mutations is also discussed.

METHODS

A 30-year-old primigravida presented in the third trimester with a fetus diagnosed with vein of Galen malformation on sonography. MRI and MR angiography were performed for further assessment. The subsequent pregnancy was again complicated by vein of Galen malformation. In addition to MRI, genetic analysis was carried out on both fetuses and on the parents.

RESULTS

MR angiography revealed that both fetuses suffered from the choroidal sub-type of vein of Galen malformation, with multiple arterial feeders fistulating onto a midline venous pouch. The visualised anatomy obtained was far superior than on sonography and allowed categorisation of vein of Galen malformation sub-type. Genetic analysis on the mother and both fetuses showed variant RASA1 gene mutation.

CONCLUSIONS

This case demonstrates that fetal MRI is a powerful tool in the investigation of in utero neurovascular malformations. A genetic mutation was identified, but this was of uncertain significance.

3D Texture Analysis of Heterogeneous MRI Data for Diagnostic Classification of Childhood Brain Tumours

Brain tumours are the most frequently occuring solid tumours affecting childhood, representing 27% of all cancers. The most common posterior fossa tumours are medulloblastoma, pilocytic astrocytoma and ependymoma. Texture Analysis (TA) of Magnetic Resonance Imaging (MRI) aims to represent pixel distributions, intensities and dependencies using mathematically defined features. Such features could potentially provide quantifiable information that is beyond the human vision capabilities, and hence be used to supplement qualitative assessments conducted by radiologists. The primary aim of this study was to carry out a multicentre investigation on the efficacy of 3D TA for diagnostic classification of childhood brain tumours, using conventional MRI images. The data used had been acquired at three different hospitals and consisted of pre-contrast T1 and T2-weighted MRI series, obtained from 121 children diagnosed with medulloblastoma, pilocytic astrocytoma and ependymoma. Using 3D textural features, based on first, second and higher order statistical methods, a support vector machine (SVM) classifier was trained and tested using the leave-one-out cross-validation (LOOCV) approach. An essential outcome of this study is that 3D TA demonstrated a good overall performance, when used on data acquired from a number of centres and using scanners made by different manufacturers and at different magnetic field strengths.

Subjective discomfort in children receiving 3 T MRI and experienced adults' perspective on children's tolerability of 7 T: a cross-sectional questionnaire survey

OBJECTIVES

To explore the possible discomfort perceived by children participating in 7 T MRI research, and the age range in which children are most likely to tolerate it well.

DESIGN

A cross-sectional survey using age-appropriate questionnaires containing six measures of subjective discomfort (general discomfort, dizziness, noisiness, claustrophobia and feeling of cold or warm).

SETTING

For children, 3 T clinical scanner in a tertiary referral teaching hospital; for adults, 3 and 7 T scanner in a university research building.

PARTICIPANTS

Non-sedated children and young people under 18 years of age who underwent 3 T clinical MRI for brain or musculoskeletal scans and adult volunteers attending 7 T with or without 3 T for brain scans.

RESULTS

83% (89/107) of involved individuals returned questionnaires. The most common discomfort among 31 children receiving 3 T MRI was noisiness (39%), followed by cold (19%), general discomfort (16%), dizziness (13%) and claustrophobia (10%). The noise was reported more frequently in children younger than 12 years than those older (p=0.021). The most common discomfort for 58 adults receiving 7 T MRI was noisiness (43%). In adults, there was a higher frequency of general discomfort during 7 than 3 T scans (p=0.031). More than 85% of adult respondents thought children aged 12-17 years would tolerate 7 T scans well, but only 35% and 15% thought children aged 10-11 and 8-9 years, respectively, would.

CONCLUSIONS

Noisiness was the most common discomfort across all ages in 3 and 7 T scanners. Although general discomfort was more common during 7 than 3 T scans in adults, most adults thought children aged 12 years or more would tolerate 7 T MRI well. Cautious enrolment of children in 7 T MRI study is warranted, but until there is more evidence of how well those aged 12 years or more tolerate 7 T MRI, we would caution against enrolling younger children.

Metrics and textural features of MRI diffusion to improve classification of pediatric posterior fossa tumors

BACKGROUND AND PURPOSE

Qualitative radiologic MR imaging review affords limited differentiation among types of pediatric posterior fossa brain tumors and cannot detect histologic or molecular subtypes, which could help to stratify treatment. This study aimed to improve current posterior fossa discrimination of histologic tumor type by using support vector machine classifiers on quantitative MR imaging features.

MATERIALS AND METHODS

This retrospective study included preoperative MRI in 40 children with posterior fossa tumors (17 medulloblastomas, 16 pilocytic astrocytomas, and 7 ependymomas). Shape, histogram, and textural features were computed from contrast-enhanced T2WI and T1WI and diffusivity (ADC) maps. Combinations of features were used to train tumor-type-specific classifiers for medulloblastoma, pilocytic astrocytoma, and ependymoma types in separation and as a joint posterior fossa classifier. A tumor-subtype classifier was also produced for classic medulloblastoma. The performance of different classifiers was assessed and compared by using randomly selected subsets of training and test data.

RESULTS

ADC histogram features (25th and 75th percentiles and skewness) yielded the best classification of tumor type (on average >95.8% of medulloblastomas, >96.9% of pilocytic astrocytomas, and >94.3% of ependymomas by using 8 training samples). The resulting joint posterior fossa classifier correctly assigned >91.4% of the posterior fossa tumors. For subtype classification, 89.4% of classic medulloblastomas were correctly classified on the basis of ADC texture features extracted from the Gray-Level Co-Occurence Matrix.

CONCLUSIONS

Support vector machine-based classifiers using ADC histogram features yielded very good discrimination among pediatric posterior fossa tumor types, and ADC textural features show promise for further subtype discrimination. These findings suggest an added diagnostic value of quantitative feature analysis of diffusion MR imaging in pediatric neuro-oncology.

Patterns of accidental craniocerebral injury occurring in early childhood

OBJECTIVE

To describe the range of intracranial injuries encountered in 0-2-year-olds in cases of accidental head injury where the mechanism of trauma was well characterised and to assess the clinical consequences.

DESIGN

A retrospective review of imaging and clinical data.

SETTING

Two tertiary paediatric referral centres.

PATIENTS

All children aged 0-2 undergoing cranial CT as indicated by National Institute for Health and Clinical Excellence guidance for head injury from 2006 to 2011. After exclusion criteria, 149 patients were included.

MAIN OUTCOME MEASURES

Rates of skull fracture, intracranial haemorrhage, parenchymal injuries and ischaemic change per type of mechanism of injury. Rates of neurological sequelae on follow-up.

RESULTS

Skull fractures were demonstrated in 54 (36%) patients of whom 17 (11%) had thin underlying subdural haemorrhage (SDH). Extradural haemorrhage complicated one fracture and two cases of isolated subdural haematomas were seen without skull fracture. Radiologically evident brain parenchymal injuries were present in three patients, all with mechanisms of injury involving high levels of force; severe neurological sequelae were only seen in one patient, who had diffuse hypoxic-ischaemic damage at presentation and whose (accidental) mechanism of injury involved extensive acceleration/deceleration/translational forces.

CONCLUSIONS

Skull fractures and focal SDH are relatively common following minor trauma in this age group but in the vast majority of cases there are no long-term neurological sequelae. Conversely, diffuse brain injury with severe subsequent neurological impairment was only seen in patients with a correspondingly severe mechanism of injury.