Grading of gliomas is a piece of critical information related to prognosis and survival. Classifying glioma grade by semantic radiological features is subjective, requires multiple MRI sequences, is quite complex and clinically demanding, and can very often result in erroneous radiological diagnosis. We used a radiomics approach with machine learning classifiers to determine the grade of gliomas. Eighty-three patients with histopathologically proven gliomas underwent MRI of the brain. Whenever available, immunohistochemistry was additionally used to augment the histopathological diagnosis. Segmentation was performed manually on the T2W MR sequence using the TexRad texture analysis softwareTM, Version 3.10. Forty-two radiomics features, which included first-order features and shape features, were derived and compared between high-grade and low-grade gliomas. Features were selected by recursive feature elimination using a random forest algorithm method. The classification performance of the models was measured using accuracy, precision, recall, f1 score, and area under the curve (AUC) of the receiver operating characteristic curve. A 10-fold cross-validation was adopted to separate the training and the test data. The selected features were used to build five classifier models: support vector machine, random forest, gradient boost, naive Bayes, and AdaBoost classifiers. The random forest model performed the best, achieving an AUC of 0.81, an accuracy of 0.83, f1 score of 0.88, a recall of 0.93, and a precision of 0.85 for the test cohort. The results suggest that machine-learning-based radiomics features extracted from multiparametric MRI images can provide a non-invasive method for predicting glioma grades preoperatively. In the present study, we extracted the radiomics features from a single cross-sectional image of the T2W MRI sequence and utilized these features to build a fairly robust model to classify low-grade gliomas from high-grade gliomas (grade 4 gliomas).
Purpose: Medulloblastoma is a heterogenous disease comprising four molecular subgroups: wingless (WNT), sonic hedgehog (SHH), group 3, and group 4, respectively. Excellent long-term outcomes have prompted deintensification of therapy in WNT-pathway medulloblastoma. We assessed the safety of avoiding upfront craniospinal irradiation (CSI) in children with low-risk WNT-pathway medulloblastoma. Patients and Methods: Children with low-risk WNT-pathway medulloblastoma were treated with postoperative focal conformal radiotherapy, avoiding upfront CSI, followed by six cycles of adjuvant systemic chemotherapy. A group-sequential design (triangular test) with predefined stopping rules if the rate of relapse exceeded 15% at 2 years was incorporated to ensure the safety of study participants. Results: 7 children with low-risk WNT-pathway medulloblastoma were accrued after written informed consent/assent and treated as per protocol. One child died of neutropenic sepsis and multiorgan dysfunction during chemotherapy. Three children were detected with neuraxial failure (supratentorial brain and/or spine) on surveillance neuro-imaging within 2 years from index diagnosis, leading to premature study termination. At relapse, children were treated with salvage CSI plus boost irradiation of metastatic deposits followed by second-line chemotherapy. Two of them continue to be in remission (32 and 26 months after first relapse), while one child developed a second relapse, necessitating further systemic chemotherapy and craniospinal reirradiation, resulting in excellent clinico-radiologic response. At a median follow-up of 42 months, the 2-year Kaplan–Meier estimates of event-free survival, recurrence-free survival, and overall survival were 42.9%, 50%, and 85.7% respectively. Conclusions: Omission of upfront CSI in low-risk WNT-pathway medulloblastoma is associated with an unacceptably high risk of neuraxial failure. See related commentary by Remke and Ramaswamy, p. 4161
Papillary tumors of the pineal region (PTPRs) are rare and are recognized as a distinct entity in the recent 2007 World Health Organization (WHO) classification of tumors of nervous system. Till date, only 55 cases have been reported. We report another three patients of PTPRs with characteristic radiological features. Histologically, all the three patients exhibited remarkable uniform histology of epithelioid morphology with variable presence of single to multilayered papillary architecture. Consistent absence of fibrillary matrix was noted in all the three cases. None of the cases showed aggressive histology. A large multicenter study is essential for upfront characterizing the biological behavior, as frequency of these tumors is very low.
Background: Intracerebral schwannomas are rare tumors resembling their peripheral nerve sheath counterparts but localized in the central nervous system (CNS). They are not classified as a separate tumor type in the 2021 World Health Organization classification. This study aimed to compile and characterize these rare neoplasms morphologically and molecularly. Methods: We analyzed 20 tumor samples by histology, RNA next-generation sequencing, DNA-methylation profiling, copy number analyses, and single-nucleus RNA sequencing (snRNA-seq). Clinical data, including age, sex, and disease progression, were collected. Magnetic resonance imaging (MRI) series were included when available. Results: All cases with tissue available for histology review (n = 13) were morphologically consistent with intracerebral schwannoma, but differed in their extent of glial fibrillary acidic protein staining. All (n = 20) shared DNA-methylation profiles distinct from other CNS tumors, as well as from Vestigial-like family (VGLL)-altered peripheral nerve sheath tumors. Most cases (n = 14/17) harbored fusions of either Vestigial-like family member 3 (VGLL3) or Vestigial-like Family member 1 (VGLL1) (CHD7::VGLL3 [n = 9/17] and EWSR1::VGLL1 [n = 5/17]). In 2 cases, the presence of a VGLL3 fusion was also confirmed by copy number analyses (n = 2/17). MRI (n = 4) showed well-defined, nodular tumors with strong, homogeneous enhancement and no diffusion restriction. Tumors were located throughout the neuroaxis (supratentorial [n = 15], infratentorial [n = 4], and spinal [n = 1]). snRNA-seq of a VGLL1-fused tumor indicated VGLL1 upregulation in 28.6% of tumor cells (n = 1). During a median follow-up of 1.8 years (range 3 months-9 years), none of the tumors recurred (n = 10). Conclusions: We identify and define a new benign tumor class, designated VGLL-altered intraparenchymal CNS schwannomas. These tumors feature VGLL alterations and a specific DNA-methylation profile, with schwannoma-like histopathology and CNS localization, akin to previously classified intracerebral schwannomas.