Faculty Bibliography

Molecular classification has improved diagnosis and treatment for patients with malignant gliomas. However, classification has relied on individual assays that are both costly and slow, leading to frequent delays in treatment. Here, we propose the use of DNA methylation, as an emerging clinical diagnostic platform, to classify gliomas based on major genomic alterations and provide insight into subtype characteristics. We show that using machine learning models, DNA methylation signatures can accurately predict somatic alterations and show improvement over existing classifiers. The established Unified Diagnostic Pipeline (UniD) we develop is rapid and cost-effective for genomic alterations and gene expression subtypes diagnostic at early clinical phase and improves over individual assays currently in clinical use. The significant relationship between genetic alteration and epigenetic signature indicates broad applicability of our approach to other malignancies.

Image-based analysis as a method for mutation detection can be advantageous in settings when tumor tissue is limited or unavailable for direct testing. Here, we utilize two distinct and complementary machine learning methods of analyzing whole slide images (WSI) for predicting mutated BRAF. In the first method, WSI of melanomas from 256 patients were used to train a deep convolutional neural network (CNN) in order to develop a fully automated model that first selects for tumor-rich areas (Area Under the Curve AUC=0.96) then predicts for mutated BRAF (AUC=0.71). Saliency mapping was performed and revealed that pixels corresponding to nuclei were the most relevant to network learning. In the second method, WSI were analyzed using a pathomics pipeline that first annotates nuclei and then quantifies nuclear features, demonstrating that mutated BRAF nuclei were significantly larger and rounder nuclei compared to BRAF WT nuclei. Lastly, we developed a model that combines clinical information, deep learning, and pathomics that improves the predictive performance for mutated BRAF to AUC=0.89. Not only does this provide additional insights on how BRAF mutations affect tumor structural characteristics, machine learning-based analysis of WSI has the potential to be integrated into higher order models for understanding tumor biology.

Liquid biopsy offers a noninvasive approach to monitor cancer burden during therapy and surveillance. However, in pediatric brain cancers, liquid biopsy methods from the blood have been unsuccessful due to a low tumor burden and low number of mutations in coding regions. In contrast with targeted panels, whole genome sequencing (WGS)-derived patient specific mutational signature from a matched tumor-normal WGS can provide a personalized, highly specific approach to detect mutations in circulating cell free tumor DNA (ctDNA) and provide blood-based monitoring in pediatric patients with high sensitivity. Furthermore, it can be performed on lower amount of peripheral blood since WGS requires less depth compared to targeted ctDNA panels. We have profiled a diverse cohort of brain tumors including medulloblastomas, ependymomas, low- and high-grade gliomas. Using WGS of matched tumor-normal and plasma samples, we could derive a personalized mutational pattern and used an AI-based error suppression model for quantification and ultra-sensitive detection of ctDNA in plasma samples. A patient-specific personalized genome-wide compendium of somatic mutations could be established across all tumor types and ctDNA tested at the time of diagnosis, during the therapy or surveillance period. An AI-based error suppression model is implemented to filter out the noise in the cell free DNA (cfDNA) while the personalized mutational signature was used to detect the ctDNA in the cfDNA and to amplify the somatic signal contained in it. The ctDNA Tumor Fraction (TF) is compared to the clinical status and MR-based imaging. All subtypes of pediatric brain tumors contain sufficient number of mutations to derive personalized signatures and corelate with the clinical status. Patient-specific WGS tumor signature in ctDNA from blood can be used for sensitive monitoring of children with brain tumors. However, correlation between ctDNA levels and therapeutic response need to be established for various subtypes of brain tumors

In the recent 5th edition of the WHO classification of CNS tumors, 'Astroblastoma, MN1 altered' is recognized a distinct brain tumor type, occurring in children and young adults. Due to its rarity and novelty, little is known about clinical and molecular traits. Therefore, we initiated an international effort and collected tissue samples, clinical and molecular data from 176 patients with Astroblastoma, MN1 altered, identified by their distinct DNA methylation profiles. DNA methylation-based t-SNE clustering analyses revealed that Astroblastoma, MN1 altered tumors form one distinct main cluster (n=158) showing MN1:BEND2 and single cases with EWSR1:BEND2 fusions and a further adjacent, but distinct smaller cluster (n=18) mostly defined by MN1:CXXC5 fusions. Both fusion partner-defined groups show a median age of 12 years but distinct copy-number aberrations, characteristically a gain of chromosome 5 in one third of the CXXC5-fused group and a loss of chromosome 16q in one third of BEND2-fused cases. As previously reported, a vast majority of Astroblastoma, MN1 altered patients are female, which we confirm for the BEND2-fused group (85%). The CXXC5-fused group, however, shows 75% male patients. Interestingly, 9/10 tumors of the few male patients observed in the BEND2-fused group were all located infratentorially or in the spinal cord, whereas almost all female cases show a supratentorial location (85/87). Histologically, the BEND2-fused group was primarily reported as Astroblastoma (39%), whereas in the CXXC5-fused cases, 31% CNS-PNET and only 8% Astroblastoma histologies were originally assigned. Preliminary clinical analyses showed that the BEND2-fused group has a relatively good 5/10-year OS of 97%/89%, but a less favorable 5/10-year PFS of 48%/35%, in line with previous studies. Patients showing CXXC5-fused tumors (n=8) indicated 5/10-year OS and PFS rates of 83%/83% and 60%/60%, respectively. Additional survival and molecular analyses are being conducted to further characterize Astroblastoma, MN1 altered tumors and its molecular subgroups

The histone-lysine N-methyltransferase EZH2 is the catalytic component of the PRC2 complex and is overexpressed in several medulloblastoma subtypes. However, its role in medulloblastoma tumorigenesis has been shown to be context-dependent using genetic approaches. Furthermore, pharmacological approaches have been limited by the very poor blood-brain barrier (BBB) penetration of current EZH2 inhibitors in use. Using laser capture microdissection and RNA-Seq analysis of human nodular/desmoplastic SHH medulloblastoma FFPE tissue, we provide data for the spatial epigenetic heterogeneity of primitive/proliferative regions compared to nodular/mature regions. Bioinformatic analysis identifies ~120 differentially expressed genes between primitive and mature regions with enrichment for genes regulated by H3K4me3 and H3K27me3 or SUZ12. ChIP-Seq analysis shows striking differences in H3K27me3 enrichment between primitive and mature medulloblastoma cells including at the EZH2 locus. Utilizing a genetically-engineered mouse model of SHH medulloblastoma, we show that conditional EZH2 genetic ablation within medulloblastoma cells results in wide-spread tumor cell differentiation (n=31 mice;*p=2e-07). Conversely, conditional EZH2 (Y641F) activation in this GEM model prevents tumor cell differentiation. Notably, we have found that the CDNK2A (p16) locus is an important EZH2 target that regulates tumor cell differentiation. qRT-PCR analysis of SHH medulloblastoma in wild-type and Ezh2 knockout settings show significant reduction in Gli1 and CCND1 and increase p15 and p16 expression in Ezh2 knockout mice compared to Ezh2 wildtype mice (*p<0.05). Importantly, genetic ablation of p16 conditionally in SHH MB EZH2 double knockout mice rescues the widespread tumor cell differentiation (n=9 mice;*p=3e-06) seen in Ezh2 single knockout SHH medulloblastoma mice. Finally, we developed a novel fucoidan-based nanoparticle strategy to deliver the EZH2 inhibitor (EPZ-6438) across the intact BBB of this GEM model to achieve significant extension of mouse survival (median 70 days compared to 19 days in control mice;*p=0.01, Mantel-Cox) with potential utility for other pediatric brain tumors

Myxopapillary ependymoma (MPE) is a heterogeneous disease regarding histopathology and outcome. The underlying molecular biology is poorly understood, and markers that reliably predict the patients' clinical course are unknown. We assembled a cohort of 185 tumors classified as MPE based on DNA methylation from pediatric, adolescent, and adult patients. Methylation patterns, copy number profiles, and MGMT promoter methylation were analyzed for all tumors, 106 tumors were evaluated histomorphologically, and RNA sequencing was performed for 37 cases. Based on methylation profiling, we defined two subtypes MPE-A and MPEB, and explored associations with epidemiological, clinical, pathological, and molecular characteristics of these tumors. Tumors in the methylation class MPE were histologically diagnosed as WHO grade I (59%), WHO grade II (37%), or WHO grade III tumors (4%). 75/77 analyzed tumors expressed HOXB13, which is a diagnostic feature not detected in other spinal ependymal tumors. Based on DNA methylation, our series split into two subtypes. MPE-A occurred in younger patients (median age 27 vs. 45 years, p=7.3e-05). They were enriched with WHO grade I tumors and associated with papillary morphology and MGMT promoter hypermethylation (all p<0.001). MPE-B included most tumors initially diagnosed as WHO grade II and cases with tanycytic morphology. Copy number alterations were more common in MPE-A. RNA sequencing revealed an enrichment for extracellular matrix and immune system-related signatures in MPE-A. 15/30 MPE-A could not be totally resected compared to 1/58 MPE-B (p=6.3e-08), and progressionfree survival was significantly better for MPE-B (p=3.4e-06, 10-year relapse rate 33% vs. 85%). We unraveled the morphological and clinical heterogeneity of MPE by identifying two molecularly distinct subtypes. These subtypes significantly differed in progression-free survival and will likely need different protocols for surveillance and treatment

Ependymoma (EPN) is an aggressive pediatric tumor that occurs throughout the central nervous system. The two most aggressive molecular subgroups of EPN are the supratentorial ZFTA-fusion associated group (ST-EPN-ZFTA) and the posterior fossa group A (PF-EPN-A). Although the molecular characteristics underlying the tumorigenesis of these subgroups have been extensively studied, these tumors remain difficult to treat. Hence, innovative therapeutic approaches are urgently needed. Here, we used genome-wide chromosome conformation capture (Hi-C), complemented with CTCF (insulators) and H3K27ac (active enhancers) ChIP-seq, as well as gene expression and whole-genome DNA methylation profiling in primary and relapsed EPN tumors and cell lines, to identify chromosomal rearrangements and regulatory mechanisms underlying aberrant expression of genes that are essential for EPN tumorigenesis. By integrating these heterogenous data types, we have observed the formation of new topologically associated domains ('neo-TADs') caused by intra-and inter-chromosomal structural variants in both tumors. In addition, we observed 3D chromatin complexes of regulatory elements, and the replacement of CTCF insulators by DNA hyper-methylation in PF-EPN-A tumors. These tumor-specific 3D genome conformations can be associated with the transcriptional upregulation of nearby genes. Through inhibition experiments we validated that these newly identified genes, including RCOR2, ITGA6, LAMC1, and ARL4C, are highly essential for the survival of patient-derived EPN cell lines in a disease subgroup-specific manner. Thus, our study identifies novel potential therapeutic vulnerabilities in EPN and extends our ability to reveal tumor-dependency genes and pathways by oncogenic 3D genome conformations even in tumors that lack known genetic alterations

DNA methylation array profiling is an emerging modality in molecular diagnostics, which is particularly useful for cases with unclear or conflicting histology. Lesions with mixed subependymoma and ependymoma components have been well documented in the literature but grading and management of these lesions is unclear as they may represent unique entities with an intermediate prognosis to pure subependymoma and ependymomas or a histologic variant within one of these established diagnostic entities. We identified 10 cases between the years 2015-2021 that were classified as subependymoma by DNA methylation array profiling despite the given histologic diagnoses of low-grade ependymal neoplasm (1), ependymoma (5), mixed ependymomasubependymoma (2), or pilocytic astrocytoma (2). Eight of the surgical specimens were from men while 2 were from women, with the range of age at time of surgery being 30 - 75 years old (median = 57.5 years old). The majority of these tumors arose in the posterior fossa (8), including all of those with an original histologic diagnosis of ependymoma and mixed ependymoma-subependymoma, while the remaining 2 cases (1 pilocytic astrocytoma and 1 low grade ependymal neoplasm) were supratentorial. Laterality of the lesions was distributed between left (2), right (3), and midline (4). All but 2 cases were from initial resections, with one of the 2 recurrent cases having a PFS of 6 years before re-resection (unknown for the second recurrent case). Short-term follow-up data was available for 3 patients ranging from 13 - 21 months during which time none had any evidence of progression. These cases illustrate the value of methylation array profiling in characterizing ependymomas and related tumors, particularly in the posterior fossa

Ependymomas encompass multiple, clinically relevant tumor types based on localization, genetic alterations, and epigenetic and transcriptomic profiles. Tumors belonging to the methylation class of spinal ependymoma (SP-EPN) represent the most common intramedullary neoplasms in children and adults. However, molecular data of SP-EPN are scarce, and clear treatment recommendations are lacking. The only known recurrent genetic events in SP-EPN are loss of chromosome 22q and NF2 mutations. Yet, it remains unclear whether SP-EPN with germline or sporadic NF2 mutations or with NF2 wild type status differ clinically or molecularly. To provide a comprehensive molecular profile of SP-EPN, we integrated epigenetic, genomic, transcriptomic, and histological analyses of up to 237 cases. Clustering of methylation data revealed two distinct molecular SP-EPN subtypes. The distribution of NF2 mutated cases differed significantly across these subtypes (p <0.0001): The vast majority of tumors harboring either a previously known NF2 germline mutation or a sporadic mutation were assigned to subtypes A, whereas subtype B tumors mainly contained NF2 wild type sequences. In addition, subtype A tumors showed a lower frequency of MGMT promoter methylation (p= 0.018) and contained almost all pediatric patients of the cohort. Whole-exome sequencing (30 cases) identified numerous mutations in NF2 wild type and mutated tumors. Mutated genes in NF2 wild type tumors were enriched for genes associated with cell cycle and cytoskeleton. RNA sequencing revealed two distinct transcriptional groups with upregulation of proliferative genes in one group and upregulation of cilial genes in the other group. The molecular subtypes displayed subtle, but significant differences in the appearance of histopathological characteristics, such as surfaces, inflammation, and hyalinized vessels. Investigation of clinical parameters is ongoing and will complete the picture of SP-EPN heterogeneity as an important basis for future clinical decision-making