No abstract

Diffuse intrinsic pontine glioma (DIPG) remains a fatal brainstem tumor demanding innovative therapies. As B7-H3 (CD276) is expressed on central nervous system (CNS) tumors, we designed B7-H3-specific chimeric antigen receptor (CAR) T cells, confirmed their preclinical efficacy, and opened BrainChild-03 (NCT04185038), a first-in-human phase 1 trial employing repeated locoregional B7-H3CARs to children with recurrent/refractory CNS tumors and DIPG. Here, we report results of the first 3 evaluable patients with DIPG (including two who enrolled after progression), who received 40 infusions with no dose-limiting toxicities. One patient had sustained clinical and radiographic improvement through 12 months on study. Patients exhibited correlative evidence of local immune activation and persistent cerebrospinal fluid (CSF) B7-H3CARs. Targeted mass spectrometry of CSF biospecimens revealed modulation in B7-H3 and critical immune analytes (CD14, CD163, CSF-1, CXCL13, and VCAM-1). Our data suggest the feasibility of repeated intracranial B7-H3CAR dosing, and that intracranial delivery may induce local immune activation.

BACKGROUND AND PURPOSE: Posterior fossa tumors are the most common pediatric brain tumors. MR imaging is key to tumor detection, diagnosis, and therapy guidance. We sought to develop an MR imaging-based deep learning model for posterior fossa tumor detection and tumor pathology classification. MATERIALS AND METHODS:The study cohort comprised 617 children (median age, 92 months; 56% males) from 5 pediatric institutions with posterior fossa tumors: diffuse midline glioma of the pons (n ¼ 122), medulloblastoma (n ¼ 272), pilocytic astrocytoma (n ¼ 135), and ependymoma (n ¼ 88). There were 199 controls. Tumor histology served as ground truth except for diffuse midline glioma of the pons, which was primarily diagnosed by MR imaging. A modified ResNeXt-50-32x4d architecture served as the backbone for a multitask classifier model, using T2-weighted MRIs as input to detect the presence of tumor and predict tumor class. Deep learning model performance was compared against that of 4 radiologists.

Focal malformations of cortical development including focal cortical dysplasia, hemimegalencephaly and megalencephaly, are a spectrum of neurodevelopmental disorders associated with brain overgrowth, cellular and architectural dysplasia, intractable epilepsy, autism and intellectual disability. Importantly, focal cortical dysplasia is the most common cause of focal intractable paediatric epilepsy. Gain and loss of function variants in the PI3K-AKT-MTOR pathway have been identified in this spectrum, with variable levels of mosaicism and tissue distribution. In this study, we performed deep molecular profiling of common PI3K-AKT-MTOR pathway variants in surgically resected tissues using droplet digital polymerase chain reaction (ddPCR), combined with analysis of key phenotype data. A total of 159 samples, including 124 brain tissue samples, were collected from 58 children with focal malformations of cortical development. We designed an ultra-sensitive and highly targeted molecular diagnostic panel using ddPCR for six mutational hotspots in three PI3K-AKT-MTOR pathway genes, namely PIK3CA (p.E542K, p.E545K, p.H1047R), AKT3 (p.E17K) and MTOR (p.S2215F, p.S2215Y). We quantified the level of mosaicism across all samples and correlated genotypes with key clinical, neuroimaging and histopathological data. Pathogenic variants were identified in 17 individuals, with an overall molecular solve rate of 29.31%. Variant allele fractions ranged from 0.14 to 22.67% across all mutation-positive samples. Our data show that pathogenic MTOR variants are mostly associated with focal cortical dysplasia, whereas pathogenic PIK3CA variants are more frequent in hemimegalencephaly. Further, the presence of one of these hotspot mutations correlated with earlier onset of epilepsy. However, levels of mosaicism did not correlate with the severity of the cortical malformation by neuroimaging or histopathology. Importantly, we could not identify these mutational hotspots in other types of surgically resected epileptic lesions (e.g. polymicrogyria or mesial temporal sclerosis) suggesting that PI3K-AKT-MTOR mutations are specifically causal in the focal cortical dysplasia-hemimegalencephaly spectrum. Finally, our data suggest that ultra-sensitive molecular profiling of the most common PI3K-AKT-MTOR mutations by targeted sequencing droplet digital polymerase chain reaction is an effective molecular approach for these disorders with a good diagnostic yield when paired with neuroimaging and histopathology.

AHT is a frequent cause of neurologic injury in children, particularly in infants in the first year of life. Imaging evaluation plays a vital role in determining the diagnosis and prognosis. A review of the intracranial injuries that are common in AHT cases has been provided. Understanding the common patterns of abusive head injury can help increase diagnostic accuracy both by increasing recognition of injuries with a high specificity for AHT and by avoiding unwarranted concern in patients with concordant injury patterns and clinical history.

Background Diffuse Intrinsic pontine gliomas (DIPGs) are lethal pediatric brain tumors. Presently, MRI is the mainstay of disease diagnosis and surveillance. We identify clinically significant computational features from MRI and create a prognostic machine learning model. Methods We isolated tumor volumes of T1-post contrast (T1) and T2-weighted (T2) MRIs from 177 treatment-naïve DIPG patients from an international cohort for model training and testing. The Quantitative Image Feature Pipeline and PyRadiomics was used for feature extraction. Ten-fold cross-validation of LASSO Cox regression selected optimal features to predict overall survival (OS) in the training dataset and tested in the independent testing dataset. We analyzed model performance using clinical variables (age at diagnosis and sex) only, radiomics only, and radiomics plus clinical variables. Results All selected features were intensity and texture-based on the wavelet filtered images (three T1 grey-level co-occurrence matrix (GLCM) texture features, T2 GLCM texture feature, and T2 first order-mean). This multivariable Cox model demonstrated a concordance of 0.68 [95% CI: 0.61-0.74] in the training dataset, significantly outperforming the clinical-only model (C=0.57 [95% CI: 0.49-0.64]). Adding clinical features to radiomics slightly improved performance (C=0.70 [95% CI: 0.64-0.77]). The combined radiomics and clinical model was validated in the independent testing dataset (C=0.59 [95% CI: 0.51-0.67], Noether’s test p=0.02). Conclusion In this international study, we demonstrate the use of radiomic signatures to create a machine learning model for DIPG prognostication. Standardized, quantitative approaches that objectively measure DIPG changes, including computational MRI evaluation, could offer new approaches to assessing tumor phenotype and serve a future role for optimizing clinical trial eligibility and tumor surveillance.

No abstract

BACKGROUND Clinicians and machine classifiers reliably diagnose pilocytic astrocytoma (PA) on magnetic resonance imaging (MRI) but less accurately distinguish medulloblastoma (MB) from ependymoma (EP). One strategy is to first rule out the most identifiable diagnosis. OBJECTIVE To hypothesize a sequential machine-learning classifier could improve diagnostic performance by mimicking a clinician's strategy of excluding PA before distinguishing MB from EP. METHODS We extracted 1800 total Image Biomarker Standardization Initiative (IBSI)-based features from T2- and gadolinium-enhanced T1-weighted images in a multinational cohort of 274 MB, 156 PA, and 97 EP. We designed a 2-step sequential classifier – first ruling out PA, and next distinguishing MB from EP. For each step, we selected the best performing model from 6-candidate classifier using a reduced feature set, and measured performance on a holdout test set with the microaveraged F1 score. RESULTS Optimal diagnostic performance was achieved using 2 decision steps, each with its own distinct imaging features and classifier method. A 3-way logistic regression classifier first distinguished PA from non-PA, with T2 uniformity and T1 contrast as the most relevant IBSI features (F1 score 0.8809). A 2-way neural net classifier next distinguished MB from EP, with T2 sphericity and T1 flatness as most relevant (F1 score 0.9189). The combined, sequential classifier was with F1 score 0.9179. CONCLUSION An MRI-based sequential machine-learning classifiers offer high-performance prediction of pediatric posterior fossa tumors across a large, multinational cohort. Optimization of this model with demographic, clinical, imaging, and molecular predictors could provide significant advantages for family counseling and surgical planning.