The Molecular Landscape of Primary CNS Lymphomas (PCNSLs) in Children and Young Adults.

Pediatric brain tumors are often noted to be different from their adult counterparts in terms of molecular features. Primary CNS lymphomas (PCNSLs) are mostly found in elderly adults and are uncommon in children and teenagers. There has only been scanty information about the molecular features of PCNSLs at a young age. We examined PCNSLs in 34 young patients aged between 7 and 39 years for gene rearrangements of BCl2, BCL6, CCND1, IRF4, IGH, IGL, IGK, and MYC, homozygous deletions (HD) of CDKN2A, and HLA by FISH. Sequencing was performed using WES, panel target sequencing, or Sanger sequencing due to the small amount of available tissues. The median OS was 97.5 months and longer than that for older patients with PCNSLs. Overall, only 14 instances of gene rearrangement were found (5%), and patients with any gene rearrangement were significantly older (p = 0.029). CDKN2A HD was associated with a shorter OS (p < 0.001). Only 10/31 (32%) showed MYD88 mutations, which were not prognostically significant, and only three of them were L265P mutations. CARD11 mutations were found in 8/24 (33%) cases only. Immunophenotypically, the cases were predominantly GCB, in contrast to older adults (61%). In summary, we showed that molecular findings identified in the PCNSLs of the older patients were only sparingly present in pediatric and young adult patients.

Rare Pediatric Cerebellar High-Grade Gliomas Mimic Medulloblastomas Histologically and Transcriptomically and Show p53 Mutations.

Pediatric high-grade gliomas (HGG) of the cerebellum are rare, and only a few cases have been documented in detail in the literature. A major differential diagnosis for poorly differentiated tumors in the cerebellum in children is medulloblastoma. In this study, we described the histological and molecular features of a series of five pediatric high-grade gliomas of the cerebellum. They actually showed histological and immunohistochemical features that overlapped with those of medulloblastomas and achieved high scores in NanoString-based medulloblastoma diagnostic assay. Methylation profiling demonstrated these tumors were heterogeneous epigenetically, clustering to GBM_MID, DMG_K27, and GBM_RTKIII methylation classes. MYCN amplification was present in one case, and PDGFRA amplification in another two cases. Interestingly, target sequencing showed that all tumors carried TP53 mutations. Our results highlight that pediatric high-grade gliomas of the cerebellum can mimic medulloblastomas at histological and transcriptomic levels. Our report adds to the rare number of cases in the literature of cerebellar HGGs in children. We recommend the use of both methylation array and TP53 screening in the differential diagnoses of poorly differentiated embryonal-like tumors of the cerebellum.

Molecular landscapes of longitudinal NF2/22q and non-NF2/22q meningiomas show different life histories.

Recurrence is a major complication of some meningiomas. Although there were many studies on biomarkers associated with higher grades or increased aggressiveness, few studies specifically examined longitudinal samples of primary meningiomas and recurrences from the same patients for molecular life history. We studied 99 primary and recurrent meningiomas from 42 patients by FISH for 22q, 1q, 1p, 3p, 5q, 6q, 10p, 10q, 14q, 18q, CDKN2A/B homozygous deletion, ALT (Alternative Lengthening of Telomere), TERT re-arrangement, targeted sequencing and TERTp sequencing. Although NF2 mutation and 22q were well known to be aetiological events in meningiomas, we found that in these paired meningiomas, combining the two events resulted in an NF2/22q group (57 tumors from 25 patients) which were almost mutually exclusive with those cases without these two changes (42 tumors from 17 patients) for NF2/22q. No other molecular changes were totally unique to NF2/22q or non-NF2/22q tumors. For molecular evolution, NF2/22q meningiomas had higher cytogenetic abnormalities than non-NF2/22q meningiomas (p = 0.003). Most of the cytogenetic changes in NF2/22q meningiomas were present from the outset whereas for non-NF2/22q meningiomas, cytogenetic events were uncommon in the primary tumors and most were acquired in recurrences. For non-NF2/22q tumors, CDKN2A/B homozygous deletion, 1q gain, 18p loss, 3p loss, and ALT were preferentially found in recurrences. Mutations were largely conserved between primary and recurrent tumors. Phylogenetic trees showed 11/11 patients with multiple recurrent tumors had a conserved evolutionary pattern. We conclude that for molecular life history, NF2 and 22q should be regarded as a group. NF2/22q recurring meningiomas showed more cytogenetic abnormalities in the primary tumors, whereas non-NF2/22q meningiomas showed CDKN2A/B deletion and other cytogenetic abnormalities and ALT at recurrences. Although chromosome 1p loss is a known poor prognostic marker in meningiomas, it was also associated with a shorter TBR (time between resection) in this cohort (p = 0.002).

Molecular landscape of IDH-wild type, pTERT-wild type adult glioblastomas.

Telomerase reverse transcriptase (TERT) promoter (pTERT) mutation has often been described as a late event in gliomagenesis and it has been suggested as a prognostic biomarker in gliomas other than 1p19q codeleted tumors. However, the characteristics of isocitrate dehydrogenase (IDH) wild type (wt) (IDHwt), pTERTwt glioblastomas are not well known. We recruited 72 adult IDHwt, pTERTwt glioblastomas and performed methylation profiling, targeted sequencing, and fluorescence in situ hybridization (FISH) for TERT structural rearrangement and ALT (alternative lengthening of telomeres). There was no significant difference in overall survival (OS) between our cohort and a the Cancer Genome Atlas (TCGA) cohort of IDHwt, pTERT mutant (mut) glioblastomas, suggesting that pTERT mutation on its own is not a prognostic factor among IDHwt glioblastomas. Epigenetically, the tumors clustered into classic-like (11%), mesenchymal-like (32%), and LGm6-glioblastoma (GBM) (57%), the latter far exceeding the corresponding proportion seen in the TCGA cohort of IDHwt, pTERTmut glioblastomas. LGm6-GBM-clustered tumors were enriched for platelet derived growth factor receptor alpha (PDGFRA) amplification or mutation (p = 0.008), and contained far fewer epidermal growth factor receptor (EGFR) amplification (p < 0.01), 10p loss (p = 0.001) and 10q loss (p < 0.001) compared with cases not clustered to this group. LGm6-GBM cases predominantly showed ALT (p = 0.038). In the whole cohort, only 35% cases showed EGFR amplification and no case showed combined chromosome +7/-10. Since the cases were already pTERTwt, so the three molecular properties of EGFR amplification, +7/-10, and pTERT mutation may not cover all IDHwt glioblastomas. Instead, EGFR and PDGFRA amplifications covered 67% and together with their mutations covered 71% of cases of this cohort. Homozygous deletion of cyclin dependent kinase inhibitor 2A (CDKN2A)/B was associated with a worse OS (p = 0.031) and was an independent prognosticator in multivariate analysis (p = 0.032). In conclusion, adult IDHwt, pTERTwt glioblastomas show epigenetic clustering different from IDHwt, pTERTmut glioblastomas, and IDHwt glioblastomas which are pTERTwt may however not show EGFR amplification or +7/-10 in a significant proportion of cases. CDKN2A/B deletion is a poor prognostic biomarker in this group.

Combinations of Single-Gene Biomarkers Can Precisely Stratify 1,028 Adult Gliomas for Prognostication.

Advanced genomic techniques have now been incorporated into diagnostic practice in neuro-oncology in the literature. However, these assays are expensive and time-consuming and demand bioinformatics expertise for data interpretation. In contrast, single-gene tests can be run much more cheaply, with a short turnaround time, and are available in general pathology laboratories. The objective of this study was to establish a molecular grading scheme for adult gliomas using combinations of commonly available single-gene tests. We retrospectively evaluated molecular diagnostic data of 1,275 cases of adult diffuse gliomas from three institutions where we were testing for IDH1/2 mutation, TERTp mutation, 1p19q codeletion, EGFR amplification, 10q deletion, BRAF V600E, and H3 mutations liberally in our regular diagnostic workup. We found that a molecular grading scheme of Group 1 (1p19q codeleted, IDH mutant), Group 2 (IDH mutant, 1p19q non-deleted, TERT mutant), Group 3 (IDH mutant, 1p19q non-deleted, TERT wild type), Group 4 (IDH wild type, BRAF mutant), Group 5 (IDH wild type, BRAF wild type and not possessing the criteria of Group 6), and Group 6 (IDH wild type, and any one of TERT mutant, EGFR amplification, 10q deletion, or H3 mutant) could significantly stratify this large cohort of gliomas for risk. A total of 1,028 (80.6%) cases were thus classifiable with sufficient molecular data. There were 270 cases of molecular Group 1, 59 cases of molecular Group 2, 248 cases of molecular Group 3, 27 cases of molecular Group 4, 117 cases of molecular Group 5, and 307 cases of molecular Group 6. The molecular groups were independent prognosticators by multivariate analyses and in specific instances, superseded conventional histological grades. We were also able to validate the usefulness of the Groups with a cohort retrieved from The Cancer Genome Atlas (TCGA) where similar molecular tests were liberally available. We conclude that a single-gene molecular stratification system, useful for fine prognostication, is feasible and can be adopted by a general pathology laboratory.

Molecular landscape of pediatric type IDH wildtype, H3 wildtype hemispheric glioblastomas.

The WHO (2021) Classification classified a group of pediatric-type high-grade gliomas as IDH wildtype, H3 wildtype but as of currently, they are characterized only by negative molecular features of IDH and H3. We recruited 35 cases of pediatric IDH wildtype and H3 wildtype hemispheric glioblastomas. We evaluated them with genome-wide methylation profiling, targeted sequencing, RNAseq, TERT promoter sequencing, and FISH. The median survival of the cohort was 27.6 months. With Capper et al.'s36 methylation groups as a map, the cases were found to be epigenetically heterogeneous and were clustered in proximity or overlay of methylation groups PXA-like (n = 8), LGG-like (n = 10), GBM_MYCN (n = 9), GBM_midline (n = 5), and GBM_RTKIII (n = 3). Histology of the tumors in these groups was not different from regular glioblastomas. Methylation groups were not associated with OS. We were unable to identify groups specifically characterized by EGFR or PDGFRA amplification as proposed by other authors. EGFR, PDGFRA, and MYCN amplifications were not correlated with OS. 4/9 cases of the GBM_MYCN cluster did not show MYCN amplification; the group was also enriched for EGFR amplification (4/9 cases) and the two biomarkers overlapped in two cases. Overall, PDGFRA amplification was found in only four cases and they were not restricted to any groups. Cases in proximity to GBM_midline were all hemispheric and showed loss of H3K27me3 staining. Fusion genes ALK/NTRK/ROS1/MET characteristic of infantile glioblastomas were not identified in 17 cases successfully sequenced. BRAF V600E was only found in the PXA group but CDKN2A deletion could be found in other methylation groups. PXA-like cases did not show PXA histological features similar to findings by other authors. No case showed TERT promoter mutation. Mutations of mismatch repair (MMR) genes were poor prognosticators in single (p ≤ 0.001) but not in multivariate analyses (p = 0.229). MGMT had no survival significance in this cohort. Of the other common biomarkers, only TP53 and ATRX mutations were significant poor prognosticators and only TP53 mutation was significant after multivariate analyses (p = 0.024). We conclude that IDH wildtype, H3 wildtype pediatric hemispheric glioblastomas are molecularly heterogeneous and in routine practice, TP53, ATRX, and MMR status could profitably be screened for risk stratification in laboratories without ready access to methylation profiling.

Molecular landscape of IDH-wild-type, H3-wild-type glioblastomas of adolescents and young adults.

OBJECTIVE: We aimed to characterise glioblastomas of adolescents and young adults (AYAs) that were isocitrate dehydrogenase (IDH) wild type (wt) and H3wt. MATERIALS AND METHODS: Fifty such patients (aged 16-32) were studied by methylation profiling, targeted sequencing and targeted RNA-seq. RESULTS: Tumours predominantly clustered into three methylation classes according to the terminology of Capper et al. (2018): (anaplastic) pleomorphic xanthoastrocytoma (PXA) (21 cases), GBM_midline (15 cases) and glioblastoma RTK/mesenchymal (seven cases). Two cases clustered with ANA_PA, four cases with LGG classes and one with GBM_MYCN. Only fifteen cases reached a calibrated score >0.84 when the cases were uploaded to DKFZ Classifier. GBM_midline-clustered tumours had a poorer overall survival (OS) compared with the PXA-clustered tumours (p = 0.030). LGG-clustered cases had a significantly better survival than GBM_midline-clustered tumours and glioblastoma RTK/mesenchymal-clustered tumours. Only 13/21 (62%) of PXA-clustered cases were BRAF V600E mutated. Most GBM_midline-clustered cases were not located in the midline. GBM_midline-clustered cases were characterised by PDGFRA amplification/mutation (73.3%), mutations of mismatch repair genes (40.0%), and all showed H3K27me3 and EZH1P loss, and an unmethylated MGMT promoter. Across the whole cohort, MGMT promoter methylation and wt TERT promoter were favourable prognosticators. Mismatch repair gene mutations were poor prognosticators and together with methylation class and MGMT methylation, maintained their significance in multivariate analyses. BRAF mutation was a good prognosticator in the PXA-clustered tumours. CONCLUSION: Methylation profiling is a useful tool in the diagnosis and prognostication of AYA glioblastomas, and the methylation classes have distinct molecular characteristics. The usual molecular diagnostic criteria for adult IDHwt glioblastoma should be applied with caution within the AYA age group.

Mismatch repair proteins PMS2 and MLH1 can further refine molecular stratification of IDH-mutant lower grade astrocytomas.

The diagnostic role of Isocitrate Dehydrogenase (IDH) mutation status in adult lower grade astrocytomas was first formally presented within the WHO Classification of Tumours of the Central Nervous System (2016). IDH-mutant astrocytomas are not as common as IDH-wildtype astrocytomas but are of better prognosis. Our previous study provided an evident that IDH-mutant lower grade astrocytomas is not a homogeneous group and could be further stratified by PDGFRA amplification, CDK4 amplification and CDKN2A deletion. In this study, we detected the expressions of DNA mismatch repair (MMR) proteins (PMS2, MLH1, MSH2, MSH6) and PD-L1 by immunohistochemistry in 147 IDH-mutant lower grade astrocytomas and explored their clinical relevance. The loss of was identified in 28.6%, 1.4%, 8.8% and 13.6%, respectively. PD-L1 expression was detected in 1.4% of this cohort. Survival analysis revealed that loss of PMS2 was correlated with shorter OS (p < 0.001) and PFS (p = 0.005). Loss of PMS2 or MLH1 was associated with shorter OS (p < 0.001) and PFS (p = 0.008). In IDH-mutant lower grade astrocytomas without CDKN2A deletion, loss of PMS2 was associated with poorer OS (p < 0.001) and PFS (p = 0.001). Furthermore, among IDH-mutant lower grade astrocytomas lacking the three biomarkers (PDGFRA, CDK4 and CDKN2A), loss of PMS2 was also associated with a poorer OS (p < 0.001) and PFS (p = 0.003). Our data illustrated the potential application of MMR genes in stratification of IDH-mutant lower grade astrocytomas without PDGFRA, CDK4 and CDKN2A copy number alterations.

Molecular landscape of IDH-mutant primary astrocytoma Grade IV/glioblastomas.

WHO 2016 classified glioblastomas into IDH-mutant and IDH-wildtype with the former having a better prognosis but there was no study on IDH-mutant primary glioblastomas only, as previous series included secondary glioblastomas. We recruited a series of 67 IDH-mutant primary glioblastomas/astrocytoma IV without a prior low-grade astrocytoma and examined them using DNA-methylation profiling, targeted sequencing, RNA sequencing and TERT promoter sequencing, and correlated the molecular findings with clinical parameters. The median OS of 39.4 months of 64 cases and PFS of 25.9 months of 57 cases were better than the survival data of IDH-wildtype glioblastomas and IDH-mutant secondary glioblastomas retrieved from datasets. The molecular features often seen in glioblastomas, such as EGFR amplification, combined +7/-10, and TERT promoter mutations were only observed in 6/53 (11.3%), 4/53 (7.5%), and 2/67 (3.0%) cases, respectively, and gene fusions were found only in two cases. The main mechanism for telomere maintenance appeared to be alternative lengthening of telomeres as ATRX mutation was found in 34/53 (64.2%) cases. In t-SNE analyses of DNA-methylation profiles, with an exceptional of one case, a majority of our cases clustered to IDH-mutant high-grade astrocytoma subclass (40/53; 75.5%) and the rest to IDH-mutant astrocytoma subclass (12/53; 22.6%). The latter was also enriched with G-CIMP high cases (12/12; 100%). G-CIMP-high status and MGMT promoter methylation were independent good prognosticators for OS (p = 0.022 and p = 0.002, respectively) and TP53 mutation was an independent poor prognosticator (p = 0.013) when correlated with other clinical parameters. Homozygous deletion of CDKN2A/B was not correlated with OS (p = 0.197) and PFS (p = 0.278). PDGFRA amplification or mutation was found in 16/59 (27.1%) of cases and was correlated with G-CIMP-low status (p = 0.010). Aside from the three well-known pathways of pathogenesis in glioblastomas, chromatin modifying and mismatch repair pathways were common aberrations (88.7% and 20.8%, respectively), the former due to high frequency of ATRX involvement. We conclude that IDH-mutant primary glioblastomas have better prognosis than secondary glioblastomas and have major molecular differences from other commoner glioblastomas. G-CIMP subgroups, MGMT promoter methylation, and TP53 mutation are useful prognostic adjuncts.

Clinical and mutational profiles of adult medulloblastoma groups.

Adult medulloblastomas are clinically and molecularly understudied due to their rarity. We performed molecular grouping, targeted sequencing, and TERT promoter Sanger sequencing on a cohort of 99 adult medulloblastomas. SHH made up 50% of the cohort, whereas Group 3 (13%) was present in comparable proportion to WNT (19%) and Group 4 (18%). In contrast to paediatric medulloblastomas, molecular groups had no prognostic impact in our adult cohort (p = 0.877). Most frequently mutated genes were TERT (including promoter mutations, mutated in 36% cases), chromatin modifiers KMT2D (31%) and KMT2C (30%), TCF4 (31%), PTCH1 (27%) and DDX3X (24%). Adult WNT patients showed enrichment of TP53 mutations (6/15 WNT cases), and 3/6 TP53-mutant WNT tumours were of large cell/anaplastic histology. Adult SHH medulloblastomas had frequent upstream pathway alterations (PTCH1 and SMO mutations) and few downstream alterations (SUFU mutations, MYCN amplifications). TERT promoter mutations were found in 72% of adult SHH patients, and were restricted to this group. Adult Group 3 tumours lacked hallmark MYC amplifications, but had recurrent mutations in KBTBD4 and NOTCH1. Adult Group 4 tumours harboured recurrent mutations in TCF4 and chromatin modifier genes. Overall, amplifications of MYC and MYCN were rare (3%). Since molecular groups were not prognostic, alternative prognostic markers are needed for adult medulloblastoma. KMT2C mutations were frequently found across molecular groups and were associated with poor survival (p = 0.002). Multivariate analysis identified histological type (p = 0.026), metastasis (p = 0.031) and KMT2C mutational status (p = 0.046) as independent prognosticators in our cohort. In summary, we identified distinct clinical and mutational characteristics of adult medulloblastomas that will inform their risk stratification and treatment.

Radiomic Features From Multi-Parameter MRI Combined With Clinical Parameters Predict Molecular Subgroups in Patients With Medulloblastoma.

The 2016 WHO classification of central nervous system tumors has included four molecular subgroups under medulloblastoma (MB) as sonic hedgehog (SHH), wingless (WNT), Grade 3, and Group 4. We aimed to develop machine learning models for predicting MB molecular subgroups based on multi-parameter magnetic resonance imaging (MRI) radiomics, tumor locations, and clinical factors. A total of 122 MB patients were enrolled retrospectively. After selecting robust, non-redundant, and relevant features from 5,529 extracted radiomics features, a random forest model was constructed based on a training cohort (n = 92) and evaluated on a testing cohort (n = 30). By combining radiographic features and clinical parameters, two combined prediction models were also built. The subgroup can be classified using an 11-feature radiomics model with a high area under the curve (AUC) of 0.8264 for WNT and modest AUCs of 0.6683, 0.6004, and 0.6979 for SHH, Group 3, and Group 4 in the testing cohort, respectively. Incorporating location and hydrocephalus into the radiomics model resulted in improved AUCs of 0.8403 and 0.8317 for WNT and SHH, respectively. After adding gender and age, the AUCs for WNT and SHH were further improved to 0.9097 and 0.8654, while the accuracies were 70 and 86.67% for Group 3 and Group 4, respectively. Prediction performance was excellent for WNT and SHH, while that for Group 3 and Group 4 needs further improvements. Machine learning algorithms offer potentials to non-invasively predict the molecular subgroups of MB.

Incremental prognostic value and underlying biological pathways of radiomics patterns in medulloblastoma.

BACKGROUND: To develop a radiomics signature for predicting overall survival (OS)/progression-free survival (PFS) in patients with medulloblastoma (MB), and to investigate the incremental prognostic value and biological pathways of the radiomics patterns. METHODS: A radiomics signature was constructed based on magnetic resonance imaging (MRI) from a training cohort (n = 83), and evaluated on a testing cohort (n = 83). Key pathways associated with the signature were identified by RNA-seq (GSE151519). Prognostic value of pathway genes was assessed in a public GSE85218 cohort. FINDINGS: The radiomics-clinicomolecular signature predicted OS (C-index 0.762) and PFS (C-index 0.697) better than either the radiomics signature (C-index: OS: 0.649; PFS: 0.593) or the clinicomolecular signature (C-index: OS: 0.725; PFS: 0.691) alone, with a better calibration and classification accuracy (net reclassification improvement: OS: 0.298, P = 0.022; PFS: 0.252, P = 0.026). Nine pathways were significantly correlated with the radiomics signature. Average expression value of pathway genes achieved significant risk stratification in GSE85218 cohort (log-rank P = 0.016). INTERPRETATION: This study demonstrated radiomics signature, which associated with dysregulated pathways, was an independent parameter conferring incremental value over clinicomolecular factors in survival predictions for MB patients. FUNDING: A full list of funding bodies that contributed to this study can be found in the Acknowledgements section.

Pattern of Relapse and Treatment Response in WNT-Activated Medulloblastoma.

Over the past decade, wingless-activated (WNT) medulloblastoma has been identified as a candidate for therapy de-escalation based on excellent survival; however, a paucity of relapses has precluded additional analyses of markers of relapse. To address this gap in knowledge, an international cohort of 93 molecularly confirmed WNT MB was assembled, where 5-year progression-free survival is 0.84 (95%, 0.763-0.925) with 15 relapsed individuals identified. Maintenance chemotherapy is identified as a strong predictor of relapse, with individuals receiving high doses of cyclophosphamide or ifosphamide having only one very late molecularly confirmed relapse (p = 0.032). The anatomical location of recurrence is metastatic in 12 of 15 relapses, with 8 of 12 metastatic relapses in the lateral ventricles. Maintenance chemotherapy, specifically cumulative cyclophosphamide doses, is a significant predictor of relapse across WNT MB. Future efforts to de-escalate therapy need to carefully consider not only the radiation dose but also the chemotherapy regimen and the propensity for metastatic relapses.

Molecular subgrouping of medulloblastoma based on few-shot learning of multitasking using conventional MR images: a retrospective multicenter study.

BACKGROUND: The determination of molecular subgroups-wingless (WNT), sonic hedgehog (SHH), Group 3, and Group 4-of medulloblastomas is very important for prognostication and risk-adaptive treatment strategies. Due to the rare disease characteristics of medulloblastoma, we designed a unique multitask framework for the few-shot scenario to achieve noninvasive molecular subgrouping with high accuracy. METHODS: We introduced a multitask technique based on mask regional convolutional neural network (Mask-RCNN). By effectively utilizing the comprehensive information including genotyping, tumor mask, and prognosis, multitask technique, on the one hand, realized multi-purpose modeling and simultaneously, on the other hand, promoted the accuracy of the molecular subgrouping. One hundred and thirteen medulloblastoma cases were collected from 4 hospitals during the 8-year period in the retrospective study, which were divided into 3-fold cross-validation cohorts (N = 74) from 2 hospitals and independent testing cohort (N = 39) from the other 2 hospitals. Comparative experiments of different auxiliary tasks were designed to illustrate the effect of multitasking in molecular subgrouping. RESULTS: Compared to the single-task framework, the multitask framework that combined 3 tasks increased the average accuracy of molecular subgrouping from 0.84 to 0.93 in cross-validation and from 0.79 to 0.85 in independent testing. The average area under the receiver operating characteristic curves (AUCs) of molecular subgrouping were 0.97 in cross-validation and 0.92 in independent testing. The average AUCs of prognostication also reached to 0.88 in cross-validation and 0.79 in independent testing. The tumor segmentation results achieved the Dice coefficient of 0.90 in both cohorts. CONCLUSIONS: The multitask Mask-RCNN is an effective method for the molecular subgrouping and prognostication of medulloblastomas with high accuracy in few-shot learning.

IDH mutant lower grade (WHO Grades II/III) astrocytomas can be stratified for risk by CDKN2A, CDK4 and PDGFRA copy number alterations.

In the 2016, WHO classification of tumors of the central nervous system, isocitrate dehydrogenase (IDH) mutation is a main classifier for lower grade astrocytomas and IDH-mutated astrocytomas is now regarded as a single group with longer survival. However, the molecular and clinical heterogeneity among IDH mutant lower grade (WHO Grades II/III) astrocytomas have only rarely been investigated. In this study, we recruited 160 IDH mutant lower grade (WHO Grades II/III) astrocytomas, and examined PDGFRA amplification, CDKN2A deletion and CDK4 amplification by FISH analysis, TERT promoter mutation by Sanger sequencing and ATRX loss and p53 expression by immunohistochemistry. We identified PDGFRA amplification, CDKN2A homozygous deletion and CDK4 amplification in 18.8%, 15.0% and 18.1% of our cohort respectively, and these alterations occurred in a mutually exclusive fashion. PDGFRA amplification was associated with shorter PFS (P = 0.0003) and OS (P < 0.0001). In tumors without PDGFRA amplification, CDKN2A homozygous deletion or CDK4 amplification was associated with a shorter OS (P = 0.035). Tumors were divided into three risk groups based on the presence of molecular alterations: high risk (PDGFRA amplification), intermediate risk (CDKN2A deletion or CDK4 amplification) and low risk (neither CDKN2A deletion and CDK4 amplification nor PDGFRA amplification). These three risk groups were significantly different in overall survival with mean survivals of 40.5, 62.9 and 71.5 months. The high-risk group also demonstrated a shorter PFS compared to intermediate- (P = 0.036) and low-risk (P < 0.0001) groups. One limitation of this study is the relatively short follow-up period, a common confounding factor for studies on low-grade tumors. Our data illustrate that IDH mutant lower grade astrocytomas is not a homogeneous group and should be molecularly stratified for risk.

Identification of subsets of IDH-mutant glioblastomas with distinct epigenetic and copy number alterations and stratified clinical risks.

BACKGROUND: IDH-mutant glioblastoma is classified by the 2016 CNS WHO as a group with good prognosis. However, the actual number of cases examined in the literature is relatively small. We hypothesize that IDH-mutant glioblastoma is not a uniform group and should be further stratified. METHODS: We conducted methylation profiles and estimated copy number variations of 57 IDH-mutant glioblastomas. RESULTS: Our results showed that 59.6% and 40.4% of tumors belonged to glioma-CpG island methylator phenotype (G-CIMP)-high and G-CIMP-low methylation subgroups, respectively. G-CIMP-low subgroup was associated with significantly worse overall survival (OS) as compared to G-CIMP-high (P = .005). CDKN2A deletion (42.1%) was the most common gene copy number variation, and was significantly associated with G-CIMP-low subgroup (P = .004). Other frequent copy number changes included mesenchymal-epithelial transition (MET) (5.3%), CCND2 (19.3%), PDGFRA (14.0%), CDK4 (12.3%), and EGFR (12.3%) amplification. Both CDKN2A deletion (P = .036) and MET amplification (P < .001) were associated with poor OS in IDH-mutant glioblastomas. Combined epigenetic signature and gene copy number variations separated IDH-mutant glioblastomas into Group 1 (G-CIMP-high), Group 2 (G-CIMP-low without CDKN2A nor MET alteration), and Group 3 (G-CIMP-low with CDKN2A and/or MET alteration). Survival analysis revealed Groups 1 and 2 exhibited a favorable OS (median survival: 619 d [20.6 mo] and 655 d [21.8 mo], respectively). Group 3 exhibited a significant shorter OS (median survival: 252 d [8.4 mo]). Multivariable analysis confirmed the independent prognostic significance of our Groups. CONCLUSIONS: IDH-mutant glioblastomas should be stratified for risk with combined epigenetic signature and CDKN2A/MET status and some cases have poor outcome.