Faculty Bibliography

A methionine substitution at lysine-27 on histone H3 variants (H3K27M) characterizes ~80% of diffuse intrinsic pontine gliomas (DIPG) and inhibits polycomb repressive complex 2 (PRC2) in a dominant-negative fashion. Yet, the mechanisms for this inhibition and abnormal epigenomic landscape have not been resolved. Using quantitative proteomics, we discovered that robust PRC2 inhibition requires levels of H3K27M greatly exceeding those of PRC2, seen in DIPG. While PRC2 inhibition requires interaction with H3K27M, we found that this interaction on chromatin is transient, with PRC2 largely being released from H3K27M. Unexpectedly, inhibition persisted even after PRC2 dissociated from H3K27M-containing chromatin, suggesting a lasting impact on PRC2. Furthermore, allosterically activated PRC2 is particularly sensitive to H3K27M, leading to the failure to spread H3K27me from PRC2 recruitment sites and consequently abrogating PRC2's ability to establish H3K27me2-3 repressive chromatin domains. In turn, levels of polycomb antagonists such as H3K36me2 are elevated, suggesting a more global, downstream effect on the epigenome. Together, these findings reveal the conditions required for H3K27M-mediated PRC2 inhibition and reconcile seemingly paradoxical effects of H3K27M on PRC2 recruitment and activity.

Visual inspection of histopathology slides is one of the main methods used by pathologists to assess the stage, type and subtype of lung tumors. Adenocarcinoma (LUAD) and squamous cell carcinoma (LUSC) are the most prevalent subtypes of lung cancer, and their distinction requires visual inspection by an experienced pathologist. In this study, we trained a deep convolutional neural network (inception v3) on whole-slide images obtained from The Cancer Genome Atlas to accurately and automatically classify them into LUAD, LUSC or normal lung tissue. The performance of our method is comparable to that of pathologists, with an average area under the curve (AUC) of 0.97. Our model was validated on independent datasets of frozen tissues, formalin-fixed paraffin-embedded tissues and biopsies. Furthermore, we trained the network to predict the ten most commonly mutated genes in LUAD. We found that six of them-STK11, EGFR, FAT1, SETBP1, KRAS and TP53-can be predicted from pathology images, with AUCs from 0.733 to 0.856 as measured on a held-out population. These findings suggest that deep-learning models can assist pathologists in the detection of cancer subtype or gene mutations. Our approach can be applied to any cancer type, and the code is available at https://github.com/ncoudray/DeepPATH .

BACKGROUND AND PURPOSE/OBJECTIVE:Few studies have shown MR imaging features and ADC correlating with molecular markers and survival in patients with glioma. Our purpose was to correlate MR imaging features and ADC with molecular subtyping and survival in adult diffuse gliomas. MATERIALS AND METHODS/METHODS:promoter methylation, and overall survival. RESULTS:wild-type glioma. Other MR imaging features were not statistically significant predictors of survival. CONCLUSIONS:wild-type gliomas.

In the version of this Letter originally published, the competing interests statement was missing. The authors declare no competing interests; this statement has now been added in all online versions of the Letter.

CNS-PNET is no longer regarded a single disease but encompassed many distinct molecular entities. After removal of the term from the 2016 WHO classification of CNS tumours, diagnostic and therapeutic uncertainty remains. Through a world-wide collaboration, tumour samples from patients with the "historic" diagnosis of CNS-PNET were re-evaluated by DNA methylation profiling (n=405) and blinded neuropathological panel review (n=256). Clinical data on treatment and outcome were pooled with data on previously published patients. The given numbers represent preliminary data of the ongoing project. The re-evaluation by DNA methylation identified many distinct entities as expected, including high grade glioma (HGG, n=70), embryonal tumors with multilayered rosettes (ETMR, n=57) and CNS-neuroblastoma with FOXR2 alteration (CNS-NB-FOXR2, n=42) as the most frequent molecular diagnostic categories. Poor clinical outcome was confirmed for patients with HGG (5y-PFS 12%/5y-OS 12%, n=24), and ETMR (5y-PFS 12%/5y-OS 18%, n=62), while most patients with CNS-NBFOXR2 survived (5y-PFS 52%/5y-OS 96%, n=31). Seven of 12 relapses/progressions of CNS-NB-FOXR2 occurred in radiotherapy-naive patients. Classification into other newly described and less common entities included HGNET-MN1 (n=19), HGNET-BCOR (n=11), and EFT-CIC (n=13). Independent neuropathological review demonstrated that samples of these entities presented as non-embryonal tumours. Furthermore, marked clinical differences exist (HGNET-MN1: 5y-PFS 25%/5y-OS 95%, n=22; HGNETBCOR: 5y-PFS 0%/5y-OS 44%, n=16; CNS-EFT-CIC: 5y-PFS 40%/5y-OS 60%, n=10). Our results show that implementation of DNA methylation profiling together with histopathological analysis will improve prospective diagnostic accuracy and facilitates the retrospective outcome analysis of treatment protocols used for this variety of biologically distinct entities

Several different entities with distinct clinical and histopathological features are described within in the group of pineal tumors. Whereas pineocytoma (PC), pineal parenchymal tumors of intermediate differentiation (PPTID) and papillary tumors of the pineal region (PTPR) predominantly affect adults and are associated with a relatively favorable prognosis, pineoblastoma (PB) occurs at young age and constitutes a highly aggressive tumor. Despite multimodal treatment, prognosis is poor (especially for infants and patients with metastatic disease), and therapeutically actionable molecular targets are lacking to date. Most PB arise sporadically, however, DICER1 or RB1 germline mutations are known to predispose to the development of PB, the latter in the context of trilateral retinoblastoma. Genome-wide DNA methylation profiling and copy-number analysis were used to investigate the biological features of 230 pineal tumors of different histologies. Known histopathological entities (PC, PTPR, PPTID) could be clearly separated by unsupervised clustering, as well as further distinct subgroups, including the previously hypothesized PTPR-A and PTPR-B groups. Notably, several biologically discrete subgroups were formed by the tumors initially diagnosed as PB or pineal primitive embryonal tumors/PNETs, which showed distinct clinical associations (e.g. age distribution). Somatic deletions of DROSHA, an endoribonuclease involved in miRNA processing upstream of DICER1, seem to be a recurrent feature of the largest subgroup. Rarer DICER1 and DGCR8 alterations in this group confirm a central role of altered miRNA biogenesis in the development of this subset of PB. Further molecular and functional characterization of novel sub-clusters, including ongoing miRNAseq, will provide insights into the oncogenesis of PB

BACKGROUND: Pineoblastoma is a rare and highly aggressive brain cancer of childhood, histologically belonging to the spectrum of primitive neuroectodermal tumors. Patients with germline mutations in DICER1, a ribonuclease involved in microRNA processing, have increased risk of pineoblastoma, but genetic drivers of sporadic pineoblastoma remain unknown. METHODS: We analyzed pediatric and adult pineoblastoma samples (n=23) using integrated genomic studies, including genome-wide DNA methylation profiling, whole-exome or whole-genome sequencing, and whole-transcriptome analysis. RESULTS: Pediatric and adult pineoblastomas showed distinct methylation profiles, the latter clustering with lower grade pineal tumors and normal pineal gland. Recurrent somatic mutations were found in genes involved in PKA-and NF-kappaB signaling, as well as in chromatin remodeling genes. We identified recurrent homozygous deletions of DROSHA, acting upstream of DICER1 in microRNA processing, and a novel microduplication involving chromosomal region 1q21 containing PDE4DIP (myomegalin), comprising the ancient DUF1220 protein domain. Expression of PDE4DIP and DUF1220 proteins was present exclusively in pineoblastoma with PDE4DIP gain. Whole-transcriptome analysis showed that homozygous loss of DROSHA led to distinct changes in RNA expression profile. Disruption of the DROSHA locus in human neural stem cells using the CRISPR/Cas9 system, led to decrease of the DROSHA protein, and massive loss of miRNAs. CONCLUSION: We identified recurrent homozygous deletions of DROSHA in pineoblastoma, suggesting that different mechanisms disrupting miRNA processing are involved in the pathogenesis of familial versus sporadic pineoblastoma. Furthermore, a novel microduplication of PDE4DIP leading to upregulation of DUF1220 protein suggests DUF1220 as a novel oncogenic driver in pineoblastoma

Program cell death ligand-1 (PD-L1) membranous expression on >5% tumor cells (PD-L1 positive tumors) is an unfavorable prognostic marker in many solid tumors. We previously showed that approximately 40% of neurofibromatosis type 2 (NF2) meningiomas are PD-L1 positive tumors. However, due to the invasive nature of biopsies, collection of tumor tissue is not always feasible. Thus, a non-invasive alternative is needed to evaluate the status of tumor growth and confirm PD-L1 positive tumors before the consideration of immunotherapy. It has recently been revealed that expression of PD-L1 on tumor associated macrophages is also a strong prognostic indicator. We retrieved formalin-fixed paraffin-embedded (FFPE) tissue from 10 NF2 meningioma cases to identify PD-L1 expression on macrophages and/or monocytes. We found that 3 out of 4 PD-L1 positive tumors were associated with expression of PDL-1 on CD68 (-) monocyte-like cells located in the peri-and intravascular lumens. These cells were only observed in 1 out of 6 PD-L1 negative tumors. Compared to others, tumors with PD-L1 expression on monocyte-like cells presented a higher Ki-67 proliferative index that was above 10%. Our results suggest that PD-L1 positive circulating CD68 (-) monocyte-like cells are correlated with tumor cell PD-L1 expression and progression in NF2 meningiomas

We previously reported an increased prevalence of pre-treatment lymphopenia in pediatric patients with medulloblastoma, the most common childhood malignant brain tumor, suggesting tumor-induced systemic immune suppression present at the time of diagnosis. Tumor-induced systemic immune suppression, including lymphopenia, has been recognized in adult high-grade gliomas for several decades. Pediatric midline gliomas express distinct epigenetic and genetic alterations, such as the histone H3 K27M mutation. Patients with this mutation often experience a more aggressive clinical course and poor outcome. We confirmed the K27M mutation from tumor biopsies in 10 pediatric midline glioma patients by whole genome DNA methylation array analysis, and analyzed complete blood counts (CBC) at the time of diagnosis prior to surgery and chemotherapy. Compared to a control group of pilocytic astrocytoma, patients with K27M-mutated gliomas have higher monocyte-to-lymphocyte ratios (MLR) and absolute monocyte counts (AMC), reaching statistical significance (p<0.05). Furthermore, the prevalence of monocytosis, characterized by AMC above normal age-adjusted means, was statistically significantly higher (p<0.05) in the study group (80%) than the control group (33%). High pre-operative MLR has previously been reported in various solid tumors in association with worse prognosis and poor overall survival. This is the first study to our knowledge that identifies MLR as a potential valuable biomarker in pediatric gliomas. Our findings suggest monocytosis and overall tumor-induced systemic immune suppression present at the time of diagnosis in K27M mutant tumors. Aberrant circulating immune cell ratios may affect the development of immune therapies for malignant pediatric midline gliomas with the histone H3 K27M mutation

In contrast to adulthood, meningiomas are rare among children and adolescents. Although recent papers have characterized the genomics of adult meningiomas, the molecular profiles of childhood meningiomas have not been elucidated in detail. We analyzed 41 tumor samples from 37 pediatric meningioma patients (female: 17, male: 20; age range: 1-21 years). Atypical meningioma WHO grade II was the most frequent histological subtype (N=14, 38%). Most tumors were located at the convexity (N=18) or the skull base (N=15). Lack of SMO, AKT, KLF4/TRAF7 mutations by Sanger sequencing (n=22) prompted whole genome sequencing of a subset (n=7). All seven cases exhibited bi-allelic inactivation of NF2 (combined large deletion and germline (5/7) or somatic (2/7) base exchanges/frameshifts). Subsequently, tumor samples from all 37 patients were subjected to 450K DNA methylation profiling and targeted DNA sequencing using brain tumor specific gene panel. Loss of chromosome 22 was frequent (N=28, 76%), followed by loss of chromosome 1 (N=12, 32%) and chromosome 18 (N=7, 19%). Moreover, separation into three groups was evident: One encompassing all clear-cell meningiomas with enrichment for SMARCE1 mutations, a second dominated by atypical meningiomas, and a third group composed of benign meningiomas, as well as rare subtypes such as rhabdoid meningiomas. When analyzed with 105 adult tumors, most of pediatric meningiomas (28/37) clustered into a separate methylation group both by unsupervised hierarchical clustering and t-stochastic nearest neighbor embedding (t-SNE). Four recurrences were similar to the primary tumor. These data suggest that pediatric meningiomas are genetically distinct from adult counterparts