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

Human-Level Differentiation of Medulloblastoma from Pilocytic Astrocytoma: A Real-World Multicenter Pilot Study

Medulloblastoma and pilocytic astrocytoma are the two most common pediatric brain tumors with overlapping imaging features. In this proof-of-concept study, we investigated using a deep learning classifier trained on a multicenter data set to differentiate these tumor types. We developed a patch-based 3D-DenseNet classifier, utilizing automated tumor segmentation. Given the heterogeneity of imaging data (and available sequences), we used all individually available preoperative imaging sequences to make the model robust to varying input. We compared the classifier to diagnostic assessments by five readers with varying experience in pediatric brain tumors. Overall, we included 195 preoperative MRIs from children with medulloblastoma (n = 69) or pilocytic astrocytoma (n = 126) across six university hospitals. In the 64-patient test set, the DenseNet classifier achieved a high AUC of 0.986, correctly predicting 62/64 (97%) diagnoses. It misclassified one case of each tumor type. Human reader accuracy ranged from 100% (expert neuroradiologist) to 80% (resident). The classifier performed significantly better than relatively inexperienced readers (p < 0.05) and was on par with pediatric neuro-oncology experts. Our proof-of-concept study demonstrates a deep learning model based on automated tumor segmentation that can reliably preoperatively differentiate between medulloblastoma and pilocytic astrocytoma, even in heterogeneous data.

Brain stem tumors in children less than 3 months: Clinical and radiologic findings of a rare disease

PURPOSE

Brain stem tumors in children < 3 months at diagnosis are extremely rare. Our aim is to study a retrospective cohort to improve the understanding of the disease course and guide patient management.

METHODS

This is a multicenter retrospective analysis across the European Society for Pediatric Oncology SIOP-E HGG/DIPG Working Group linked centers, including patients with a brainstem tumor diagnosed between 2009 and 2020 and aged < 3 months at diagnosis. Clinical data were collected, and imaging characteristics were analyzed blindly and independently by two neuroradiologists.

RESULTS

Five cases were identified. No patient received any therapy. The epicenter of two tumors was in the medulla oblongata alone and in the medulla oblongata and the pons in three. For patients with tumor in equal parts in the medulla oblongata and the pons (n = 3), the extension at diagnosis involved the spinal cord; for the two patients with the tumor epicenter in the medulla oblongata alone (n = 2), the extension at diagnosis included the pons (n = 2) and the spinal cord (n = 1). Biopsy was performed in one patient identifying a pilocytic astrocytoma. Two patients died. In one patient, autopsy revealed a high-grade glioma (case 3). Three survivors showed either spontaneous tumor regression (n = 2) or stable disease (n = 1). Survivors were followed up for 10, 7, and 0.6 years, respectively. One case had the typical imaging characteristics of a dorsal exophytic low-grade glioma.

CONCLUSIONS

No patient fulfilled the radiologic criteria defining a high-grade glioma. Central neuroradiological review and biopsy may provide useful information regarding the patient management.

ESTRO-SIOPE guideline: Clinical management of radiotherapy in atypical teratoid/rhabdoid tumors (AT/RTs)

BACKGROUND AND PURPOSE

Treatment of patients with atypical teratoid/rhabdoid (AT/RT) is challenging, especially when very young (below the age of three years). Radiotherapy (RT) is part of a complex trimodality therapy. The purpose of this guideline is to provide appropriate recommendations for RT in the clinical management of patients not enrolled in clinical trials.

MATERIALS AND METHODS

Nine European experts were nominated to form a European Society for Radiotherapy and Oncology (ESTRO) guideline committee. A systematic literature search was conducted in PubMed/MEDLINE and Web of Science. They discussed and analyzed the evidence concerning the role of RT in the clinical management of AT/RT.

RESULTS

Recommendations on diagnostic imaging, therapeutic principles, RT considerations regarding timing, dose, techniques, target volume definitions, dose constraints of radiation-sensitive organs at risk, concomitant chemotherapy, and follow-up were considered. Treating children with AT/RT within the framework of prospective trials or prospective registries is of utmost importance.

CONCLUSION

The present guideline summarizes the evidence and clinical-based recommendations for RT in patients with AT/RT. Prospective clinical trials and international, large registries evaluating modern treatment approaches will contribute to a better understanding of the best treatment for these children in future.

Neurosurgical treatment of pediatric brain tumors - results from a single center multidisciplinary setup

OBJECTIVE

The challenge of pediatric brain tumor surgery is given due to a relative low prevalence but high heterogeneity in age, localization, and pathology. Improvements of long-term overall survival rates were achieved during the past decades stressing the importance of a multidisciplinary decision process guided by a national treatment protocol. We reviewed the entire spectrum of pediatric brain tumor surgeries from the perspective of an interdisciplinary pediatric neuro-oncology center in Germany.

METHODS

Every patient who underwent brain tumor surgery from January 2010 to June 2017 in our Pediatric Neurosurgery department was retrospectively included and evaluated regarding the course of treatment. Perioperative data such as tumor localization, timing of surgery, extent of resection, neuropathological diagnosis, transfusion rates, oncologic and radiation therapy, and neurological follow-up including morbidity and mortality were evaluated.

RESULTS

Two hundred ninety-three pediatric brain tumor patients were applicable (age: 8.28 ± 5.62 years, 1.22:1.0 m:f). A total of 531 tumor surgical interventions was performed within these patients (457 tumor resections, 74 tumor biopsies; mean interventions per patient 1.8 ± 1.2). Due to a critical neurologic status, 32 operations (6%) were performed on the day of admission. In 65.2% of all cases, tumor were approached supratentorially. Most frequent diagnoses of the cases were glial tumors (47.8%) and embryonal tumors (17.6%). Preoperative planned extent of resection was achieved in 92.7%. Pre- and postoperative neurologic deficits resolved completely in 30.7%, whereas symptom regressed in 28.6% of surgical interventions. New postoperative neurologic deficit was observed in 10.7%, which resolved or improved in 80% of these cases during 30 days. The mortality rate was 1%.

CONCLUSION

We outlined the center perspective of a specialized pediatric neuro-oncological center describing the heterogeneous distribution of cases regarding age-related prevalence, tumor localization, and biology, which requires a high multidisciplinary expertise. The study contributes to define challenges in treating pediatric brain tumors and to develop quality indicators for pediatric neuro-oncological surgery. We assume that an adequate volume load of patients within a interdisciplinary infrastructure is warranted to aim for effective treatment and decent quality of life for the majority of long-term surviving pediatric tumor patients.

Review of standard paediatric neuroradiology MRI protocols from 12 UK tertiary paediatric hospitals: is there much variation between centres?

AIM

To investigate how magnetic resonance imaging (MRI) examinations are protocolled in tertiary paediatric neuroradiology centres around the UK for some of the more common presentations encountered in paediatric neuroradiology, and to identify any variations of note.

MATERIALS AND METHODS

All 19 UK tertiary paediatric neuroradiology centres registered with the British Society of Neuroradiologists-Paediatric Group were contacted and asked if they could provide a copy of their standard MRI protocols. Twelve responded (63%) and 10 of the more common presentations were selected and the standard acquired sequences obtained at each participating centre were compared. Where available the collated protocols were also compared against current published guidance.

RESULTS

The basic sequences carried out by centres around the UK are similar; however, there are lots of variations overall. The only standardised protocol currently being implemented nationally in paediatric imaging is that for brain tumours. Otherwise, chosen protocols are generally dependent on the preferences and technical capabilities of individual centres. Suggested published protocols also exist for non-accidental injury (NAI), multiple sclerosis, epilepsy, and head and neck imaging.

CONCLUSIONS

The differences in MRI protocolling depend in part on technical capabilities and in part on the experience and preferences of the paediatric neuroradiologists at each centre. For most presentations, there is no consensus as to what constitutes the perfect protocol. The present results will be useful for specialist centres who may wish to review their current protocols, and for more generalist centres to use as a reference to guide their MRI protocolling.

Recommendations for neuroradiological examinations in children living with achondroplasia: a European Society of Pediatric Radiology and European Society of Neuroradiology opinion paper

Children living with achondroplasia are at an increased risk of developing neurological complications, which may be associated with acute and life-altering events. To remediate this risk, the timely acquisition of effective neuroimaging that can help to guide clinical management is essential. We propose imaging protocols and follow-up strategies for evaluating the neuroanatomy of these children and to effectively identify potential neurological complications, including compression at the cervicomedullary junction secondary to foramen magnum stenosis, spinal deformity and spinal canal stenosis. When compiling these recommendations, emphasis has been placed on reducing scan times and avoiding unnecessary radiation exposure. Standardized imaging protocols are important to ensure that clinically useful neuroimaging is performed in children living with achondroplasia and to ensure reproducibility in future clinical trials. The members of the European Society of Pediatric Radiology (ESPR) Neuroradiology Taskforce and European Society of Neuroradiology pediatric subcommittee, together with clinicians and surgeons with specific expertise in achondroplasia, wrote this opinion paper. The research committee of the ESPR also endorsed the final draft. The rationale for these recommendations is based on currently available literature, supplemented by best practice opinion from radiologists and clinicians with subject-specific expertise.

Optimizing reirradiation for relapsed medulloblastoma: identifying the ideal patient and tumor profiles

BACKGROUND

First-line therapies for medulloblastoma(MBL) are obtaining higher survival-rates while decreasing late-effects, but treatment at relapse is not standardized. We report here the experience with MBL re-irradiation(re-RT), its timing and outcome in different clinical settings and tumor groups.

METHODS

Patient's staging/treatment at diagnosis, histotypes/molecular subgroups, relapse site/s, re-treatments outcome are reported.

RESULTS

25 patients were included, with a median age of 11.4 years; 8 had metastases. According to 2016-2021 WHO-classification, 14 had SHH subgroup tumors(six TP53 mutated,one + MYC,one + NMYC amplification), 11 non-WNT/non-SHH (two with MYC/MYCN amplification).Thirteen had received HART-CSI, 11 standard-CSI, one HFRT; all post-radiation chemotherapy(CT), 16 also pre-RT. Median time to relapse (local-LR in nine, distant-DR in 14, LR + DR in two) was 26 months. Fourteen patients were re-operated, in five cases excising single DR-sites, thereafter three received CT, two after re-RT; out of 11 patients not re-operated, four had re-RT as first treatment and seven after CT. Re-RT was administered at median 32 months after first RT: focally in 20 cases, craniospinal-CSI in five. Median post-relapse-PFS/after re-RT was 16.7/8.2 months, while overall survival-OS was 35.1/23.9 months, respectively. Metastatic status both at diagnosis/relapse negatively affected outcome and re-surgery was prognostically favorable. PD after re-RT was however significantly more frequent in SHH (with a suggestive association with TP53 mutation, p = 0.050). We did not observe any influence of biological subgroups on PFS from recurrence while SHH showed apparently worse OS compared to non-WNT/non-SHH group.

CONCLUSIONS

Re-surgery + reRT can prolong survival; a substantial fraction of patients with worse outcome belongs to the SHH-subgroup.

[Pediatric supratentorial brain tumors]

BACKGROUND

Pediatric brain tumors differ regarding location and histopathological features compared to those in adults. In children, 30% of pediatric brain tumors are supratentorial lesions. Low-grade astrocytomas, e.g. pilocystic astrocytoma or craniopharyngioma, are the most common tumors.

IMAGING MODALITIES

Magnetic resonance imaging (MRI) is the default imaging technique that is used to evaluate the findings. Ultrasound and cranial computed tomography (CCT) accompany the imaging, although CCT is mainly used in emergency situations.

TOPICS COVERED

The following article describes the most common pediatric supratentorial brain tumors with reference to imaging criteria as well as changes in the World Health Organization (WHO) classification.

Comparison of the different imaging modalities used to image pediatric oncology patients: A COG diagnostic imaging committee/SPR oncology committee white paper

Diagnostic imaging is essential in the diagnosis and management, including surveillance, of known or suspected cancer in children. The independent and combined roles of the various modalities, consisting of radiography, fluoroscopy, ultrasonography (US), computed tomography (CT), magnetic resonance imaging (MRI), and nuclear medicine (NM), are both prescribed through protocols but also function in caring for complications that may occur during or subsequent to treatment such as infection, bleeding, or organ compromise. Use of a specific imaging modality may be based on situational circumstances such as a brain CT or MR for a new onset seizure, chest CT for respiratory signs or symptoms, or US for gross hematuria. However, in many situations, there are competing choices that do not easily lend themselves to a formulaic approach as options; these situations depend on the contributions of a variety of factors based on a combination of the clinical scenario and the strengths and limitations of the imaging modalities. Therefore, an improved understanding of the potential influence of the imaging decision pathways in pediatric cancer care can come from comparison among the individual diagnostic imaging modalities. The purpose of the following material to is to provide such a comparison. To do this, pediatric imaging content experts for the individual modalities of radiography and fluoroscopy, US, CT, MRI, and NM will discuss the individual modality strengths and limitations.

[Pediatric Intracranial Ependymoma - Recommendations for First-Line Treatment from the German HIT-MED study group]

Biological subtypes of ependymoma (EPN) have been introduced by the recent WHO classification and appear to have great impact on the clinical course, but have not yet found their way into clinical risk stratification. Further, the overall unfavorable prognosis underlines the fact that current therapeutic strategies need further evaluation for improvement. To date, there is no international consensus regarding first-line treatment for children with intracranial EPN. Extent of resection is known to be the most important clinical risk factor, leading to the consensus that consequent evaluation for re-surgery of postoperative residual tumor needs to have highest priority. Furthermore, efficacy of local irradiation is unquestioned and recommended for patients aged>1 year. In contrast, efficacy of chemotherapy is still under discussion. The European trial SIOP Ependymoma II aims at evaluating efficacy of different chemotherapy elements, leading to the recommendation to include German patients. The BIOMECA study, as biological accompanying study, aims at identifying new prognostic parameters. These results might help to develop targeted therapies for unfavorable biological subtypes. For patient who are not qualified for inclusion into the interventional strata, the HIT-MED Guidance 5.2 provides specific recommendations. This article is meant as an overview of national guidelines regarding diagnostics and treatment as well as of treatment according to the SIOP Ependymoma II trial protocol.

Towards an Automated Approach to the Semi-Quantification of [18F]F-DOPA PET in Pediatric-Type Diffuse Gliomas

BACKGROUND

This study aims to evaluate the use of a computer-aided, semi-quantification approach to [¹⁸F]F-DOPA positron emission tomography (PET) in pediatric-type diffuse gliomas (PDGs) to calculate the tumor-to-background ratio.

METHODS

A total of 18 pediatric patients with PDGs underwent magnetic resonance imaging and [¹⁸F]F-DOPA PET, which were analyzed using both manual and automated procedures. The former provided a tumor-to-normal-tissue ratio (T_N) and tumor-to-striatal-tissue ratio (T_S), while the latter provided analogous scores (t_n, t_s). We tested the correlation, consistency, and ability to stratify grading and survival between these methods.

RESULTS

High Pearson correlation coefficients resulted between the ratios calculated with the two approaches: ρ = 0.93 (p < 10^-4) and ρ = 0.814 (p < 10^-4). The analysis of the residuals suggested that t_n and t_s were more consistent than T_N and T_S. Similarly to T_N and T_S, the automatically computed scores showed significant differences between low- and high-grade gliomas (p ≤ 10^-4, t-test) and the overall survival was significantly shorter in patients with higher values when compared to those with lower ones (p < 10^-3, log-rank test).

CONCLUSIONS

This study suggested that the proposed computer-aided approach could yield similar results to the manual procedure in terms of diagnostic and prognostic information.

Acute care diagnostics in children for general radiologists - not alone in the hospital at night

BACKGROUND

Radiologic care for children and adolescents during night and weekend shifts is challenging. This is especially true when a dedicated pediatric radiology front or background service is not available.

METHODS

The purpose of this review is to present the approach, the most relevant diagnoses, and their differential diagnoses based on four common example cases - abdominal pain, respiratory/chest pain, headache, and refusal to walk. Essentials such as clinical classification (e. g., disease dynamics) and practical instructions (e. g., necessity of acute cross-sectional imaging) are presented.

RESULTS AND CONCLUSION

For the abdomen, appendicitis ranks first among acute diseases. Other important diseases are intussusception and volvulus. Far more frequently, however, gastroenteritis is the cause of abdominal pain. Usually no imaging is required in this case. In unclear clinical situations, ultrasound may be indicated. In suspected pulmonary infections, chest imaging is limited to inconclusive cases and suspicion of complications such as pleural empyema. Major emergencies include (spontaneous) pneumothorax and aspiration. Headache is a common symptom. Immediate imaging is only necessary in cases of suspected acute inflammatory (meningitis/encephalitis) or vascular disease (e. g., hemorrhage due to vascular malformations). MRI is the primary imaging modality in these cases. Restricted walking/refusal to walk is a classic nonspecific sign, particularly of acute musculoskeletal disease, especially in younger children. Clinical examination is essential to narrow down the field of investigation. Besides the frequent and symptomatic coxitis fugax, the rare but serious (septic) arthritis/osteomyelitis must not be overlooked.

KEY POINTS

· Radiological care of children and adolescents is challenging, especially during night and weekend shifts.. · However, in close cooperation with the referring colleagues/clinics, the appropriate approach can be effectively determined even if the symptoms are not clear.. · The selection of the optimal imaging method is based on guidance and guidelines, but also on the condition of the child/adolescent.. · A silent or whimpering child is cause for alarm..

CITATION FORMAT

· Beer M, Mentzel H, Steinborn M et al. Acute care diagnostics in children for general radiologists - not alone at night in the hospital. Fortschr Röntgenstr 2023; 195: 205 - 216.

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.

Imaging Guidelines and Recommendations for Diagnosis, Surveillance, and Management of Pediatric CNS and Spinal Tumors

Central nervous system (CNS) tumors are the second most common cause of cancer in children when incidence rates of cancer are estimated according to the Indian population dynamics based on 2011 consensus. As per the estimates, CNS tumors account for 20.1% of cancer burden in children aged between 0 and 14 years and 16.8% when 0 to 19 years age group is considered. The most common pediatric brain tumors are astrocytoma and medulloblastoma followed by other embryonal tumors, craniopharyngioma, and ependymal tumors. The incidence of CNS tumors in children from India is similar to the western high-income countries, other than slightly higher incidence of craniopharyngioma in Indian children.

ERN GENTURIS tumour surveillance guidelines for individuals with neurofibromatosis type 1

BACKGROUND

Neurofibromatosis type 1 (NF1) is a multisystem genetic disorder, predisposing development of benign and malignant tumours. Given the oncogenic potential, long-term surveillance is important in patients with NF1. Proposals for NF1 care and its specific manifestations have been developed, but lack integration within routine care. This guideline aims to assimilate available information on NF1 associated tumours (based on evidence and/or expert opinion) to assist healthcare professionals in undertaking tumour surveillance of NF1 individuals.

METHODS

By comprehensive literature review, performed March 18th 2020, guidelines were developed by a NF1 expert group and patient representatives, conversant with clinical care of the wide NF1 disease spectrum. We used a modified Delphi procedure to overcome issues of variability in recommendations for specific (national) health care settings, and to deal with recommendations based on indirect (scarce) evidence.

FINDINGS

We defined proposals for personalised and targeted tumour management in NF1, ensuring appropriate care for those in need, whilst reducing unnecessary intervention. We also incorporated the tumour-related psychosocial and quality of life impact of NF1.

INTERPRETATION

The guideline reflects the current care for NF1 in Europe. They are not meant to be prescriptive and may be adjusted to local available resources at the treating centre, both within and outside EU countries.

FUNDING

This guideline has been supported by the European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). ERN GENTURIS is funded by the European Union. DGE is supported by the Manchester NIHRBiomedical Research Centre (IS-BRC-1215-20007).

[18F]FET-PET in children and adolescents with central nervous system tumors: does it support difficult clinical decision-making?

PURPOSE

Positron emission tomography (PET) with O-(2-[¹⁸F]fluoroethyl)-L-tyrosine ([¹⁸F]FET) is a well-established tool for non-invasive assessment of adult central nervous system (CNS) tumors. However, data on its diagnostic utility and impact on clinical management in children and adolescents are limited.

METHODS

Twenty-one children and young adults (13 males; mean age, 8.6 ± 5.2 years; range, 1-19 at initial diagnosis) with either newly diagnosed (n = 5) or pretreated (n = 16) CNS tumors were retrospectively analyzed. All patients had previously undergone neuro-oncological work-up including cranial magnetic resonance imaging. In all cases, [¹⁸F]FET-PET was indicated in a multidisciplinary team conference. The impact of PET imaging on clinical decision-making was assessed. Histopathology (n = 12) and/or clinical and imaging follow-up (n = 9) served as the standard of reference.

RESULTS

The addition of [¹⁸F]FET-PET to the available information had an impact on further patient management in 14 out of 21 subjects, with avoidance of invasive surgery or biopsy in four patients, biopsy guidance in four patients, change of further treatment in another five patients, and confirmation of diagnosis in one patient.

CONCLUSION

[¹⁸F]FET-PET may provide important additional information for treatment guidance in pediatric and adolescent patients with CNS tumors.

The impact of 1.5-T intraoperative magnetic resonance imaging in pediatric tumor surgery: Safety, utility, and challenges

Objective

In this study, we present our experience with 1.5-T high-field intraoperative magnetic resonance imaging (ioMRI) for different neuro-oncological procedures in a pediatric population, and we discuss the safety, utility, and challenges of this intraoperative imaging technology.

Methods

A pediatric consecutive-case series of neuro-oncological surgeries performed between February 2020 and May 2022 was analyzed from a prospective ioMRI registry. Patients were divided into four groups according to the surgical procedure: intracranial tumors (group 1), intraspinal tumors (group 2), stereotactic biopsy for unresectable tumors (group 3), and catheter placement for cystic tumors (group 4). The goal of surgery, the volume of residual tumor, preoperative and discharge neurological status, and postoperative complications related to ioMRI were evaluated.

Results

A total of 146 procedures with ioMRI were performed during this period. Of these, 62 were oncology surgeries: 45 in group 1, two in group 2, 10 in group 3, and five in group 4. The mean age of our patients was 8.91 years, with the youngest being 12 months. ioMRI identified residual tumors and prompted further resection in 14% of the cases. The mean time for intraoperative image processing was 54 ± 6 min. There were no intra- or postoperative security incidents related to the use of ioMRI. The reoperation rate in the early postoperative period was 0%.

Conclusion

ioMRI in pediatric neuro-oncology surgery is a safe and reliable tool. Its routine use maximized the extent of tumor resection and did not result in increased neurological deficits or complications in our series. The main limitations included the need for strict safety protocols in a highly complex surgical environment as well as the inherent limitations on certain patient positions with available MR-compatible headrests.

Multimodal MRI radiomic models to predict genomic mutations in diffuse intrinsic pontine glioma with missing imaging modalities

Purpose

Predicting H3.1, TP53, and ACVR1 mutations in DIPG could aid in the selection of therapeutic options. The contribution of clinical data and multi-modal MRI were studied for these three predictive tasks. To keep the maximum number of subjects, which is essential for a rare disease, missing data were considered. A multi-modal model was proposed, collecting all available data for each patient, without performing any imputation.

Methods

A retrospective cohort of 80 patients with confirmed DIPG and at least one of the four MR modalities (T1w, T1c, T2w, and FLAIR), acquired with two different MR scanners was built. A pipeline including standardization of MR data and extraction of radiomic features within the tumor was applied. The values of radiomic features between the two MR scanners were realigned using the ComBat method. For each prediction task, the most robust features were selected based on a recursive feature elimination with cross-validation. Five different models, one based on clinical data and one per MR modality, were developed using logistic regression classifiers. The prediction of the multi-modal model was defined as the average of all possible prediction results among five for each patient. The performances of the models were compared using a leave-one-out approach.

Results

The percentage of missing modalities ranged from 6 to 11% across modalities and tasks. The performance of each individual model was dependent on each specific task, with an AUC of the ROC curve ranging from 0.63 to 0.80. The multi-modal model outperformed the clinical model for each prediction tasks, thus demonstrating the added value of MRI. Furthermore, regardless of performance criteria, the multi-modal model came in the first place or second place (very close to first). In the leave-one-out approach, the prediction of H3.1 (resp. ACVR1 and TP53) mutations achieved a balanced accuracy of 87.8% (resp. 82.1 and 78.3%).

Conclusion

Compared with a single modality approach, the multi-modal model combining multiple MRI modalities and clinical features was the most powerful to predict H3.1, ACVR1, and TP53 mutations and provided prediction, even in the case of missing modality. It could be proposed in the absence of a conclusive biopsy.

Guidelines for magnetic resonance imaging in pediatric head and neck pathologies: a multicentre international consensus paper

The use of standardized imaging protocols is paramount in order to facilitate comparable, reproducible images and, consequently, to optimize patient care. Standardized MR protocols are lacking when studying head and neck pathologies in the pediatric population. We propose an international, multicenter consensus paper focused on providing the best combination of acquisition time/technical requirements and image quality. Distinct protocols for different regions of the head and neck and, in some cases, for specific pathologies or clinical indications are recommended. This white paper is endorsed by several international scientific societies and it is the result of discussion, in consensus, among experts in pediatric head and neck imaging.

Imaging response assessment for CNS germ cell tumours: consensus recommendations from the European Society for Paediatric Oncology Brain Tumour Group and North American Children's Oncology Group

Homogeneous and common objective disease assessments and standardised response criteria are important for better international clinical trials for CNS germ cell tumours. Currently, European protocols differ from those of North America (the USA and Canada) in terms of criteria to assess radiological disease response. An international working group of the European Society for Paediatric Oncology Brain Tumour Group and North American Children's Oncology Group was therefore established to review existing literature and current practices, identify major challenges regarding imaging assessment, and develop consensus recommendations for imaging response assessment for patients with CNS germ cell tumours. New clinical imaging standards were defined for the most common sites of CNS germ cell tumour and for the definition of locoregional extension. These new standards will allow the evaluation of response to therapy in patients with CNS germ cell tumours to be more consistent, and facilitate direct comparison of treatment outcomes across international studies.

Relapsed Medulloblastoma in Pre-Irradiated Patients: Current Practice for Diagnostics and Treatment

Relapsed medulloblastoma (rMB) accounts for a considerable, and disproportionate amount of childhood cancer deaths. Recent advances have gone someway to characterising disease biology at relapse including second malignancies that often cannot be distinguished from relapse on imaging alone. Furthermore, there are now multiple international early-phase trials exploring drug-target matches across a range of high-risk/relapsed paediatric tumours. Despite these advances, treatment at relapse in pre-irradiated patients is typically non-curative and focuses on providing life-prolonging and symptom-modifying care that is tailored to the needs and wishes of the individual and their family. Here, we describe the current understanding of prognostic factors at disease relapse such as principal molecular group, adverse molecular biology, and timing of relapse. We provide an overview of the clinical diagnostic process including signs and symptoms, staging investigations, and molecular pathology, followed by a summary of treatment modalities and considerations. Finally, we summarise future directions to progress understanding of treatment resistance and the biological mechanisms underpinning early therapy-refractory and relapsed disease. These initiatives include development of comprehensive and collaborative molecular profiling approaches at relapse, liquid biopsies such as cerebrospinal fluid (CSF) as a biomarker of minimal residual disease (MRD), modelling strategies, and the use of primary tumour material for real-time drug screening approaches.

Incidence and survival for childhood central nervous system tumours in Australia, 1983-2016

PURPOSE

To investigate incidence and survival of childhood tumours of the central nervous system (CNS) by histological subtype, tumour behaviour and tumour grade.

METHODS

National, population-based data on all children under 15 years old diagnosed with a CNS tumour between 1983 and 2016 were sourced from the Australian Childhood Cancer Registry. Incidence rate trends were calculated using Joinpoint regression. Relative survival was calculated using the cohort method, with changes in survival over time by cancer type and tumour grade assessed by multivariable flexible parametric survival modelling.

RESULTS

The study cohort included 4914 patients, with astrocytoma (n = 2181, 44%) and embryonal tumours (n = 931, 19%) the most common diagnostic subgroups. Almost half (n = 2181, 44%) of all tumours were classified as high grade (III or IV). Incidence rates increased by 29% between 1983 and 2016, with high grade tumours rising by an annual average of + 1.1% (95% CI =  + 0.7%, + 1.5%, p < 0.001). 5-year survival for all patients combined was 72% (95% CI = 71-74%), ranging from 50% (46-54%) for those with other gliomas to 81% (79-83%) for astrocytoma (p < 0.001). Survival improved over time for grade II and III ependymomas but not for patients with astrocytoma irrespective of grade.

CONCLUSION

Improvements in diagnostic technology leading to more precise tumour classification are likely to explain some of the differences in incidence rate trends by histological type and grade. While improvements in survival over time were noted for some tumours, outcomes remained poor among patients with high-grade astrocytoma.

Combining multi-site magnetic resonance imaging with machine learning predicts survival in pediatric brain tumors

Brain tumors represent the highest cause of mortality in the pediatric oncological population. Diagnosis is commonly performed with magnetic resonance imaging. Survival biomarkers are challenging to identify due to the relatively low numbers of individual tumor types. 69 children with biopsy-confirmed brain tumors were recruited into this study. All participants had perfusion and diffusion weighted imaging performed at diagnosis. Imaging data were processed using conventional methods, and a Bayesian survival analysis performed. Unsupervised and supervised machine learning were performed with the survival features, to determine novel sub-groups related to survival. Sub-group analysis was undertaken to understand differences in imaging features. Survival analysis showed that a combination of diffusion and perfusion imaging were able to determine two novel sub-groups of brain tumors with different survival characteristics (p < 0.01), which were subsequently classified with high accuracy (98%) by a neural network. Analysis of high-grade tumors showed a marked difference in survival (p = 0.029) between the two clusters with high risk and low risk imaging features. This study has developed a novel model of survival for pediatric brain tumors. Tumor perfusion plays a key role in determining survival and should be considered as a high priority for future imaging protocols.