Molecular Subgroups Research Articles

MDB-99. SPATIAL AND TEMPORAL DEVELOPMENT OF GLI2-DRIVEN MEDULLOBLASTOMA

Abstract BACKGROUND Medulloblastoma (MB) is a highly malignant tumor of the cerebellum predominantly affecting children. It can be classified into four major molecular subgroups: WNT, SHH, Group 3, and Group 4. Among these subgroups, the SHH subgroup MB, characterized by GLI2 amplification, is associated with a poor prognosis. However, while GLI2 alterations have been linked to SHH-MB, evidence demonstrating its role as an oncogenic driver in an animal model is lacking. Understanding the mechanisms underlying GLI2-amplified MB initiation and progression is crucial for developing improved therapeutic strategies. METHODS By utilizing transgenic mouse models, our study aims to elucidate the role of GLI2 in SHH-MB tumorigenesis and uncover the underlying mechanisms driving tumor formation. This research effort seeks to identify novel therapeutic targets for the treatment of this disease. RESULTS We discovered that overexpression of GLI2 in embryonic Math1+ granule neuron precursors (GNPs) led to tumorigenesis within the cerebellum, resulting in 100% mortality in the mice. Importantly, the cellular and molecular characteristics of these tumors faithfully recapitulated those found in human SHH-MB. Further analysis revealed that only Math1+ GNPs specifically between E13.5 and E15.5 are susceptible to GLI2-induced tumorigenesis, while overexpression of GLI2 in embryonic GFAP+ stem cell/progenitors failed to induce tumorigenesis. Additionally, our single-cell RNA-seq analysis unveiled the heterogeneity of Math1+ cells during cerebellar development. Furthermore, we demonstrated that the MAPK signaling pathway was involved in the tumorigenesis of embryonic Math1+ GNPs by preventing cell apoptosis. CONCLUSIONS Our findings establish a pivotal oncogenic role of GLI2 in SHH-MB tumorigenesis, with GLI2-driven tumors developing within a specific spatial-temporal window from a distinct cell population during cerebellar development. Targeting the MAPK signaling pathway holds promise in preventing tumorigenesis and progression of GLI2-amplified MB.

ATRT-03. PRIMARY DIFFUSE LEPTOMENINGEAL ATYPICAL TERATOID/RHABDOID TUMOR (ATRT) OF CHILDHOOD DEMONSTRATING UNEXPECTED IMPROVEMENT UPON CESSATION OF CONVENTIONAL THERAPY FOR PRESUMED PROGRESSION; A MOLECULARLY CHARACTERIZED CASE REPORT

Abstract BACKGROUND Atypical teratoid/rhabdoid tumors (ATRTs) are malignant tumors of the central nervous system, typically presenting in the posterior fossa of very young children. Prognosis remains poor despite current therapy, while tumorigenesis implicates both genomic and epigenetic dysregulation. Primary diffuse leptomeningeal disease (PDL ATRT), without a defined intraparenchymal lesion, is an extremely rare ATRT variant with only seven cases reported. None of these preceding cases had undergone molecular characterization. We compared our case, who demonstrated an atypical clinical and radiological trajectory, against these seven cases. CASE PRESENTATION We describe a PDL ATRT, with hitherto novel molecular and imaging characteristics, in a pediatric patient who subsequently demonstrated radiological and clinical recovery from radiologically diagnosed metastatic relapse in the absence of conventional ATRT salvage treatment. Our case is the first to utilize Magnetic Resonance Imaging Arterial Spin Labelling perfusion imaging to suggest an underlying malignancy as the responsible etiology for this patient’s complex presentation, and the first to use methylation profiling to establish a tumor molecular subgroup for a PDL ATRT patient (identifying a MYC-subgroup in this case). CONCLUSIONS For this reported case of pediatric MYC-subgroup PDL ATRT, MR perfusional imaging aided in establishing the diagnosis and predicting potential relapse as part of a multi-parametric imaging assessment. We administered upfront intensive multimodal therapy, including high-dose CSRT to the patient, resulting in the longest overall survival for any patient reported with this extremely challenging ATRT variant. His unanticipated clinical and radiological improvement following a subsequent radiological diagnosis of tumor relapse highlights the importance of confirming recurrence by tissue extraction where feasible, but suggests that further evaluation of histone deacetylase inhibition as a potential future therapy in ATRT is warranted.

PATH-01. ACCURATE AND RAPID PAEDIATRIC BRAIN TUMOR CLASSIFICATION (PBT) BY METHYLATION PROFILING USING NOVEL NATIVE REAL TIME OXFORD NANOPORE TECHNOLOGY SEQUENCING TECHNOLOGY

Abstract BACKGROUND Pediatric brain tumors (PBTs) pose a significant challenge being the leading cause of cancer related mortality in children. The existing classification process is time-consuming and fails to capture the heterogeneity of PBTs. Urgent advancements in diagnostic and therapeutic approaches are needed. The WHO classification 5th edition for CNS tumors introduced DNA methylation analysis into the diagnostic process, especially for difficult-to-diagnose cases. However, the currently available DNA methylation array-based methods are laborious, costly, &require large amounts of DNA METHODS Nanopore methylation analysis is rapid, accurate, and inexpensive & requires minimal tissue. We applied this technology on 24 PBT samples that were classified into ten molecular subgroups. RESULTS Our results were compatible with those obtained through Heidelberg methylation analysis, standard histopathological examination, and molecular tests and were in concordance with the final neuro-pathological diagnosis. Moreover, for five prospective cases, that would have challenged standard methods, Nanopore yielded same-day results, guiding additional tests and providing a significant diagnostic advantage. In one case, a 2.2-year-old male, with a large left parieto-temporal tumor and a suprasellar lesioninitially treated elsewhere with upfront chemotherapy without biopsy, developed clinical deterioration. We resected the temporal lesion and frozen tissue diagnosis suggested low-grade glioma. However, same day Nanopore sequencing analysis indicated a high score for ‘Meningioma’. Final pathology examination 10 days later confirmed ‘‘Atypical meningioma, WHO grade 2” – an extremely rare diagnosis in this age group. Visual deterioration prompted immediate resection of the second lesion before final pathological confirmation. Thus, early Nanopore classification played a vital role in diagnosis and management. CONCLUSIONS Our preliminary results underscore the potential of Nanopore potential for accurate and rapid PBT classification from small brain biopsies, with DNA input as low as 100 nanograms. This cutting-edge technology promises to substantially reduce time to diagnosis, shaping future neurosurgical strategies for individualized PBT patient management.

MDB-49. OUTCOMES OF SONIC HEDGEHOG (SHH) MEDULLOBLASTOMA (MB) IN CHILDREN AND ADULTS TREATED WITH RISK-ADAPTED THERAPY ON A PROSPECTIVE INTERNATIONAL PROTOCOL (SJMB12)

Abstract BACKGROUND SJMB12 (NCT01878617) was among the first protocols to treat newly diagnosed MB incorporating molecular subgrouping. METHODS Tumors were subtyped and TP53 mutation assessed via immunohistochemistry. SHH patients were grouped into 2 strata: S1 (GTR/NTR; M0) and S2 (residual > 1.5 cm2; MYC/MYCN amplification; M+). Treatment consisted of maximal surgical resection followed by risk adapted craniospinal irradiation (CSI) (S1 = 23.4 Gy CSI; S2= 36); 54 Gy primary site. Patients then received 4 cycles of chemotherapy consisting of Vincristine 1 mg/m2 on Day 1 and Day 8; Cisplatin 75 mg/m2 on Day 1 and Cyclophosphamide 1.5 gm/m2 on Day 2 and 3 with Mesna and IV fluid support. PEG Filgrastim was administered on Day 4. Skeletally mature pts (males with bone age ≥ 17 yrs.; females with bone age ≥ 15 yrs.) were also treated with the smoothened inhibitor Vismodegib for 12 months. RESULTS Of the 660 patients accrued from 2013-2022, 110 were SHH (107 evaluable, 62 in S1 and 45 in S2). Median age was 12.2 yrs. (3.1-39.7) and 67 were males. 98 patients had GTR/NTR, 31 had M+ disease and 63 had nodular desmoplastic histology. TP53 mutation was detected in 7 (11.5%) S1 and 18 (40%) S2 pts. 5-year EFS for S1 and S2 were 84.5 ± 6.4% and 61 ± 11.5%, respectively. Five-year EFS of the S1 was 91.7 ± 5.4% and 28.6 ± 13.9% for TP53 wildtype and TP53 mutant cases, respectively, and 84.2 ± 10.1% and 25.0 ± 21.7% in the S2 cohort. CONCLUSION Patients with low-risk SHH MB lacking TP53 mutation have excellent survival and may be considered for therapy reduction. TP53 mutation, with or without M+ disease, leads to a dismal outcome that needs novel therapy for cure.

HGG-32. PNOC008: A PILOT TRIAL TESTING THE CLINICAL BENEFIT OF USING MOLECULAR PROFILING TO DETERMINE AN INDIVIDUALIZED TREATMENT PLAN IN CHILDREN AND YOUNG ADULTS WITH NEWLY DIAGNOSED HIGH-GRADE GLIOMA (EXCLUDING DIFFUSE INTRINSIC PONTINE GLIOMA)

Abstract BACKGROUND Children and young adults diagnosed with high-grade glioma (HGG) face extremely poor prognoses. Despite multiple clinical trials testing new treatments in this population, a survival advantage has yet to be achieved. Herein we assessed, in a single-arm, multi-center pilot trial, the feasibility of molecular profiling of primary HGG tumor tissue to create an individualized treatment plan with up to four FDA approved medications. METHODS Patients aged <21 years with newly diagnosed, localized, hemispheric HGG (Stratum A) or midline HGG (non-DIPG; Stratum B) were eligible. Tumor tissue underwent comprehensive molecular profiling (targeted gene panel, whole exome, and whole transcriptome sequencing). Based on detailed review of the molecular data by a dedicated tumor board, an individualized treatment plan that combined up to four FDA approved drugs was recommended. Circulating tumor DNA (ctDNA), imaging, and quality of life (QOL) measures were collected at multiple timepoints. RESULTS Fifty-five patients enrolled between 2018 and 2023 (median age 11 years [range 2-20], n=31 female, n=29 Stratum A). The most common integrated diagnoses included H3K27-altered diffuse midline glioma (n=17), H3/IDH-wildtype diffuse pediatric-type HGG (n=16), and H3G34-mutant diffuse hemispheric glioma (n=12). Median overall survival (OS) from the time of study enrollment was 26.5 months in Stratum A (lower 95% CI: 18.7) and 21.7 months in Stratum B (lower 95% CI: 16.8), with a median follow-up of 35.4 months for all patients (lower 95% CI: 32.5). The most common grade 3 or 4 treatment-related adverse events were decreased neutrophils (n=28), decreased platelets (n=22), and decreased white blood cells (n=16). As of December 2023, seven patients remain on therapy. Central imaging, ctDNA, and QOL analyses are underway. CONCLUSIONS A personalized treatment recommendation for children and young adults with HGG based on comprehensive transcriptomic and genomic analysis is feasible. Current survival data are encouraging, and molecular subgroup analyses are ongoing.

MDB-93. A TRANSCRIPTOME-BASED APPROACH TO DRUG REPURPOSING FOR AGGRESSIVE GROUP 3 MEDULLOBLASTOMAS

Abstract BACKGROUND Medulloblastoma (MB) is the most common malignant brain tumor in children, representing 20% of pediatric brain tumors. International consensus has identified molecular subgroups: WNT, SHH, group 3 (G3), and group 4 (G4). Current treatments, guided by histologic tumor grade and stage, do not fully harness these subgroups, resulting in varying prognoses. G3MB particularly faces challenges with rapid recurrence and aggressive disease, necessitating focused intervention. METHODS We conducted differential expression analysis on MB patient datasets (GSE148389, GSE164677) to generate a G3MB gene expression signature. This gene expression signature was analyzed against the LINCS database, yielding candidate antineoplastic compounds. Candidates were filtered for blood-brain barrier permeability and FDA approval. Functional assays evaluated drug effects on cytotoxicity, clonogenicity, wound healing, cell cycle, and apoptosis in G3MB cell lines (HDMB03 and D425). RNA sequencing of treated cells identified differentially regulated pathways to infer mechanism, followed by in vitro validation. Mitochondrial function and membrane potential were measured by Seahorse and TMRM assays, respectively. RESULTS Eighty-one candidates passed filtering. The leading drug class identified was antidepressants. Based on MTT assays, the top four candidates were nortriptyline, simvastatin, fluoxetine, and sertraline. Nortriptyline (NT) emerged as the lead compound, with IC50 values of ~7uM (HDMB03) and ~11uM (D425). Amongst the four drugs, NT induced consistent anti-neoplastic effects on wound healing, colony formation, medullosphere generation, and apoptosis. Mechanistically, NT induced mitochondrial stress, escalated mitochondrial superoxide, and decreased mitochondrial membrane potential. CHRM3, a receptor antagonized by NT and upregulated in G3MB samples was downregulated by NT treatment. Western blotting confirmed the inhibition of CHRM3 downstream effectors. CONCLUSIONS Our pipeline identified cytotoxic compounds against G3MB, including NT, which we hypothesize induces mitochondrial stress via CHRM3 modulation. This approach offers repurposable FDA-approved candidates for high-risk G3MB, potentially reducing current treatment toxicity and improving outcomes.

MDB-12. NANOPORE SEQUENCING AS A CUTTING-EDGE TECHNOLOGY FOR MEDULLOBLASTOMA CLASSIFICATION

Abstract INTRODUCTION Medulloblastoma (MB) is one of the most prevalent embryonal malignant brain tumors, divided into four distinct molecular subgroups (WNT, SHH, group 3, and group 4) with its individual prognosis. A more extensive classification of MB has recently been provided, identifying numerous subtypes, some with poor prognosis, leading to a significant number of patients succumbing to the disease while many survivors experience long-term side effects, highlighting the need for efficient subgrouping methods. Distinct genome-wide DNA methylation profiles characterize each MB subgroup, and the advent of Nanopore DNA sequencing has positioned itself as a reliable technique for comprehensive copy number and methylation profiling. MATERIAL AND METHODS We evaluated the ability of Nanopore sequencing to provide clinically relevant methylation and copy number profiles of MB, on a discovery cohort of 45 frozen MB samples, benchmarked against the gold standard EPIC array, and subsequently validated on a cohort of 116 MB samples. RESULTS The majority of MB in both the discovery cohort (43/45; 95.6%) and the validation cohort (110/116; 94.8%) were accurately subgrouped by Nanopore sequencing. Flongle flow cells for 18 MB allowed for a more rapid and cost-effective analysis, with 94.4% (17/18) being correctly classified. Additionally, Nanopore sequencing allowed us to correctly subtype 24/30 (80.0%) within MB groups. DISCUSSION We provided proof of concept that Nanopore sequencing can subgroup MB using DNA extracted from formalin-fixed paraffin-embedded (FFPE) samples and cell-free DNA obtained from cerebrospinal fluid. This first large-scale study establishes the proof of concept that this modern and innovative technology is well-suited for MB classification. We confirm the use of Nanopore sequencing on circulating DNA and present an initial exploration of its application on DNA extracted from FFPE samples. Its ability to deliver quick and cost-effective results firmly establishes Nanopore sequencing as a game-changer in the field of MB classification.

EPEN-27. GAINING A CLUE FROM 1Q: A HIGH-RISK CHROMOSOMAL GAIN IN PEDIATRIC POSTERIOR FOSSA EPENDYMOMA

Abstract BACKGROUND Ependymomas are neuroepithelial tumours that can be classified into distinct molecular subgroups, each exhibiting a unique clinical outcome. Therapeutic efforts against the most aggressive and abundant subgroup of pediatric ependymoma (posterior fossa group A, PFA) have traditionally been hampered by a lack of known recurrent driver mutations. However, approximately 25% of PFA tumours exhibit gains of the chromosome arm 1q which is associated with an extremely poor prognosis despite aggressive treatment. By interrogating the multi-omic landscape of PFAs harbouring this copy number variation, we aimed to uncover genetic dependencies from which actionable therapeutics could be derived. METHODS Transcriptomic (bulk/single-cell RNA sequencing), proteomic (LC-MS/MS) and phosphoproteomic analyses were conducted on patient tissues and primary patient-derived cell lines to gain insight into the pathways that drive 1q gain pathogenesis. Chromatin immunoprecipitation sequencing of H3K27me3 and H3K27ac marked histones was carried out to profile the epigenetic landscape of these tumours, and whole genome CRISPR knockout screens were performed to reveal the contribution of essential genes unique to tumours harboring 1q gains. RESULTS Through the comparison of 1q gain and balanced PFA samples, our multi-omic analyses identified the up-regulation and essentiality of known glioma oncogenic drivers previously uncharacterized in PFA. The investigation of essential candidate genes showed a convergence on pathways related to ciliogenesis, with genetic vulnerabilities informing in vitro drug screening. Candidate genes were further validated by in vivo knockdown using established xenograft mouse models for pre-clinical testing. CONCLUSIONS This combined approach has allowed for the derivation of a functional cancer signature underlying 1q gains in PFA and presents novel therapeutic targets for this high-risk subgroup. PFA ependymoma is a deadly malignancy whose complex biology has so far eluded therapy, and new insights from this project will help to identify the first effective therapies for this vulnerable patient population.

EPEN-09. TARGETING EPHB2/ABL1 RESTORES ANTI-TUMOR IMMUNITY IN A PRECLINICAL MODEL OF EPENDYMOMA

Abstract BACKGROUND Ependymoma (EPN) is a childhood brain cancer that is often resistant to cytotoxic therapies. Molecular profiling has led to a better understanding of unique EPN subtypes and revealed a critical role of EPHB2 in driving disease. With this in mind, new treatments for EPN are desperately needed and should be developed in a molecular subtype-specific fashion. METHODS We developed a syngeneic mouse model of EPHB2-driven genetically engineered EPN tumor cells. We performed transcriptomic profiling and analysis of EPHB2-driven murine EPN tumors and human EPN datasets to reveal multiple protein kinases as potential druggable targets. After identifying the tyrosine kinase inhibitor, Dasatinib, as a potentially effective FDA-approved agent, we measured changes in the murine microenvironment during EPN growth and after Dasatinib treatment. RESULTS We determined that Dasatinib inhibited the growth of EPN both in vitro and in vivo, through blocking EPHB2 and ABL1 signaling. Furthermore, we identified an increased frequency of immunosuppressive M2-like tumor- associated macrophages (TAMs), which proportionally increased with tumor size during tumor progression. However, treatment with Dasatinib reprogrammed the EPN immune microenvironment by polarizing TAMs toward an anti-tumor M1-like phenotype and increasing CD8 T cell activation. In addition, Dasatinib treatment induced complete regression of established EPN tumors in 78% of the animals and protected survivors against tumor recurrence. Depletion of CD8 T cells compromised the durability of response and reduced overall survival. CONCLUSION In conclusion, these data indicate that Dasatinib may be an effective therapy for EPHB2-driven molecular subgroup of EPN by activating the anti-tumor immune response and support further investigation of Dasatinib in clinical trials.

MDB-98. HISTOPATHOLOGY IS A ROBUST TOOL TO DIFFERENTIATE BETWEEN SHH AND WNT MEDULLOBLASTOMA VERSUS NON-SHH/NON-WNT MEDULLOBLASTOMA: COMPARATIVE ANALYSIS OF HISTOPATHOLOGY, CHROMOSOMAL MICROARRAY, NANOSTRING BASED 22-GENE ASSAY, AND DNA METHYLATION FOR MEDULLOBLASTOMA SUBGROUP ASSIGNMENT ON “HEAD START” 4 CLINICAL TRIAL

Abstract BACKGROUND “Head Start-4” (HS-4) is a prospective, randomized clinic trial that tailors treatment based on medulloblastoma molecular subgroups (WNT, SHH, Group 3, and Group 4) and response to induction chemotherapy, and compares the efficacy of one versus three (tandem) cycles of myeloablative therapy. Here we compare different methodologies used to distinguish between WNT/SHH medulloblastoma (low-risk arm) and non-SHH/non-WNT medulloblastoma (high-risk arm) during the course of the trial. METHODS When HS-4 trial began enrolling patients in 2015, in the absence of a CAP-CLIA certified test for methylation and gene expression profile, we utilized histopathology/immunochemistry (HP/IHC) and chromosomal microarray (CMA) via OncoScanTM (Thermo Fisher) to classify medulloblastoma samples into either WNT, SHH, or non-WNT/non-SHH (Subgroups 3 and 4) at the time of diagnosis. Retrospectively, we performed both NanoString based 22-gene assay and DNA methylation profiling on all patient samples. RESULTS We have HP/IHC, CMA, NanoString and methylation profiling for 54 infants and young children with medulloblastoma enrolled on HS-4. While indeterminate result occurred with CMA in three cases and NanoString in two cases, HP/IHC successfully assigned samples to SHH/WNT and non-SHH/non-WNT arms of the study in all 54 cases. We have HP/IHC, CMA and DNA methylation profiling for additional 33 patients. While pathology/IHC was indeterminate in two cases (one WNT and one group 3 MB), remaining cases accurately categorized medulloblastoma methylation subtype as WNT/SHH and non-WNT/non-SHH. CONCLUSION Due to the long turnaround time, waiting for medulloblastoma molecular subtype confirmation by DNA methylation array may delay treatment initiation. HS-4 data displays robust prediction of WNT/SHH versus non-WNT/non-SHH molecular subtypes by HP/IHC in comparison to DNA methylation profiling and provides a platform to initiate treatment based on HP/IHC while awaiting molecular information. In addition, these data support relying on histopathology for medulloblastoma subtypes in resource poor countries where methylation profiling is not readily available.

MDB-05. PERSONALIZED, RISK ADAPTED THERAPY FOR PEDIATRIC PATIENTS WITH MEDULLOBLASTOMA – A SINGLE CENTER 8-YEAR RETROSPECTIVE REVIEW

Abstract BACKGROUND Clinically relevant risk stratification and optimal therapies for pediatric medulloblastoma molecular subgroups are still being investigated. METHODS We reviewed host/disease characteristics, patient/tumor specific risk adapted therapy and outcomes of patients treated for medulloblastoma at UC Davis since 2016. RESULTS Fifteen patients (age 1-19 years, 12 male, 3 with Li Fraumeni Syndrome (LFS), 2 with spinal cord disease) were included in our review. Tumor histologies included: Classic (n=8), Desmoplastic-Nodular (n=6), Anaplastic (n=1). Tumor genomic profiles were SHHP53WT(n=5), SHHP53 mutated (n=4), Group4 (n=3), NWNSHH (n=2), WNT (n=1). Gross total resection was achieved in 9 patients. Patients with post resection residual tumor, high risk genomic features or disseminated disease were stratified as high risk. We treated all radiation eligible patients (n=13) per COG ACNS0332 maintenance chemotherapy regardless of risk group with 12/13 receiving radiosensitization (Vincristine/Carboplatin) and either 23.4 Gy/36 Gy CSI (two patients treated per Headstart3D2 and ACNS0334 respectively). 2 patients with spinal cord disease were treated with intrathecal topotecan and liposomal cytarabine respectively. During ACNS0332 maintenance bacteremia, weight loss and hypomagnesemia were frequent complications. 5 patients required chemotherapy dose reductions (4 Vincristine/1 Cisplatin). Two patients with LFS developed glioblastoma/undifferentiated pleomorphic sarcoma following ACNS0332 therapy. Two patients (with WNT and NWNSHH subgroup tumors) experienced medulloblastoma relapse. These patients were treated using LITT/Intrathecal chemotherapy and ACNS0821 chemotherapy. Both are alive and one patient is NED off therapy. All patients who have not developed secondary malignancies (n=13) are alive. (follow up 7-72 months). All attend school with support, 10 have hearing impairment and 5 receive endocrine supplements. CONCLUSIONS Utilizing patient specific risk adapted therapy we have achieved favorable clinical outcomes in our patient cohort. Therapy deintensification for patients with WNT medulloblastomas should be approached with caution. Patients with LFS might require radiation sparing therapeutic protocols for longer term benefit.

IMMU-11. CLINICAL EXPERIENCE WITH THE CONCEPT OF MULTIPHASE COMBINED TREATMENT FOR ADULTS WITH GBM

Abstract BACKGROUND We published survival data of 50 adults with GBM treated with individualized multimodal immunotherapy (IMI) as part of a multiphase combined treatment (Van Gool et al, https://doi:org/10.1016/bs.mcb.2023.06.001). Analysis was repeated with new observation data. METHODS Patient definition for retrospective analysis: start IMI between 27/05/2015 and 31/10/2023, >18y, newly diagnosed, IDH1 wild-type, MGMT promoter-methylation status documented, survival known, no second malignancy. RESULTS Forty (14 female, 26 male) GBM patients were MGMT-promoter-unmethylated (unmeth), while 31 (15 female, 16 male) were MGMT-promoter-methylated (meth). Comparing the molecular subgroups, both respective median age; 48 years (range 18-65) and 54 years (range 26-72), and respective Karnofsky performance index; 70 (range 50-100) and 80 (range 60-100), were equal. Patient distribution (< complete resection, complete resection, not documented) was equal: 17/15/8 (unmeth) versus 17/11/3 (meth). Both groups were treated similarly: in median 37 (range 0-147) sessions of modulated electrohyperthermia, in median 37 (range 5-147) bolus injections of Newcastle Disease Virus, in median 2 (range 0-6) IO-Vac® vaccines with a median total number of 30.7x10e6 DCs (range 0-158x10e6). The median overall survival (mOS) for the total group was 27 months. The 2-year overall survival (OS) was 57.9 % and the 3-year OS 37.1 %. The mOS for unmeth patients was 22.1 month, with a 2-year OS of 42.7%. The mOS for meth patients was 37.7 months, with a 2-year OS of 75.5%. Sampling five control arms of contemporary RCTs as external control arm (Liau et al, JAMA Oncol 2023) showed for unmeth and meth patients a mOS of 14.6 resp. 21.3 months with a 2-year OS of 21% resp. 42%, which reflects the current prognosis for these patients. CONCLUSIONS Our analysis confirms that the integration of IMI during and after standard of care improves the OS. Similar multiphase combined treatment strategy, including IMI, should be considered to treat childhood GBM.

Single-Cell Chromatin Accessibility Analysis Reveals Subgroup-Specific TF-NTR Regulatory Circuits in Medulloblastoma.

Medulloblastoma (MB) stands as one of the prevalent malignant brain tumors among pediatric patients. Despite its prevalence, the intricate interplay between the regulatory program driving malignancy in MB cells and their interactions with the microenvironment remains insufficiently understood. Leveraging the capabilities of single-cell Assay for Transposase-Accessible Chromatin sequencing (scATAC-seq), the chromatin accessibility landscape is unveiled across 59,015 distinct MB cells. This expansive dataset encompasses cells belonging to discrete molecular subgroups, namely SHH, WNT, Group3, and Group4. Within these chromatin accessibility profiles, specific regulatory elements tied to individual subgroups are uncovered, shedding light on the distinct activities of transcription factors (TFs) that likely orchestrate the tumorigenesis process. Moreover, it is found that certain neurotransmitter receptors (NTRs) are subgroup-specific and can predict MB subgroup classification when combined with their associated transcription factors. Notably, targeting essential NTRs within tumors influences both the in vitro sphere-forming capability and the in vivo tumorigenic capacity of MB cells. These findings collectively provide fresh insights into comprehending the regulatory networks and cellular dynamics within MBs. Furthermore, the significance of the TF-NTR regulatory circuits is underscored as prospective biomarkers and viable therapeutic targets.

Role of epigenetic regulation on catecholamine synthesis in pheochromocytoma and paraganglioma.

Pheochromocytomas and paragangliomas (PPGLs) typically secrete catecholamines and their metabolites (metanephrines [MN] and normetanephrine [NMN]). Catecholamines are synthesized by several enzymes: phenylalanine hydroxylase (encoded by PAH), tyrosine hydroxylase (TH), aromatic L-amino acid decarboxylase (DDC), dopamine β-hydroxylase (DBH), and phenylethanolamine N-methyltransferase (PNMT). MN/NMN secretion varies between anatomical and molecular subgroups. The aim of this study was to assess the correlation between DNA methylation of catecholamine synthesis genes and MN/NMN secretion. Gene promoter methylation of PAH, TH, AADC, DBH, and PNMT were extracted and calculated based on publicly available data. Comparisons and correlation analysis were performed between MN±NMN (MN/NMN), NMN only, and neither/unknown secretion patterns. Methylation levels and MN/NMN patterns were compared by three genetic alteration subgroups: pseudohypoxia (PH), kinase signaling (KS), and others. A total of 178 cases were included. Methylation of PAH CpGs negatively correlated with probability for MN/NMN secretion (p<.05 for all CpGs) and positively with NMN-only secretion. NMN-only secreting tumors had significantly higher promoter methylation of PAH, DBH, and PNMT compared with MN/NMN-secreting tumors. MN/NMN-secreting PPGLs had mainly KS alterations (52.1%), whereas NMN-only PPGLs had PH alterations (41.9%). PPGLs in the PH versus KS group had gene promoter hypermethylation of PAH (p=.002), DBH (p=.02), and PNMT (p=.003). Promoter methylation of genes encoding catecholamine synthesis enzymes is strongly and inversely correlated with MN/NMN patterns in PPGLs. KS and PH-related tumors have distinct methylation patterns. These results imply that methylation is a key regulatory mechanism of catecholamine synthesis in PPGLs.

Comprehensive genomic profiling from C-CAT database unveiled over 80% presence of oncogenic drivers in anaplastic thyroid carcinoma including BRAF, RAS family, NF1, and FGFR1.

Anaplastic thyroid carcinoma (ATC) is considered a very aggressive carcinoma and has been difficult to treat with therapeutic strategies. This study examines the landscape of genomic alteration in ATC, including the BRAF V600E mutation, and its clinical implications. A retrospective observational study was conducted using collected at the Center for Cancer Genomics and Advanced Therapeutics (C-CAT) in Japan, utilizing comprehensive genomic profiling data from 102 ATC cases. Additionally, AACR-GENIE data from 267 cases were analysed for validation. Statistical methods, including the conditional Kendall tau statistic and χ2 tests, were employed for survival analysis and gene mutation comparisons. Among 102 ATCs, BRAF, RAS, and other driver mutations were found in 83 cases (81.2%). The prevalence of BRAF V600E mutations was as high as 60%. Co-mutation analysis identified different genomic profiles in the BRAF, RAS, and wild-type groups. Despite the diverse molecular backgrounds, no significant differences in clinical variables and overall survival were observed. The analysis considering left-side amputation suggested that RAS mutations had a poorer prognosis. In the BRAF/RAS wild-type group, FGFR1 and NF1 were identified as driver mutations, with an accumulation of copy number variations and less TERT promoter mutations. This molecular subgrouping was also supported by the AACR-GENIE data. Comprehensive genomic analysis of ATC in Japan revealed distinct molecular subgroups, highlighting the importance of BRAF V600E mutations, particularly V600E, as potential therapeutic targets and suggest the relevance of tailor-made therapeutic strategies based on genomic profiling.

An analysis of PIK3CA hotspot mutations and response to neoadjuvant therapy in breast cancer patients from four prospective clinical trials.

The PI3K signaling pathway is frequently dysregulated in breast cancer (BC), and mutations in PIK3CA, are relevant for therapy resistance in HER2pos BC. Mutations in exons 9 or exon 20 may have different impact on response to neoadjuvant chemotherapy-based treatment regimens. We investigated PIK3CA mutations in 1691 early BC patients, randomized in four neoadjuvant multicenter trials: GeparQuattro (NCT00288002), GeparQuinto (NCT00567554), GeparSixto (NCT01426880) and GeparSepto (NCT01583426). The role of different PIK3CA exons and hotspots for pathological complete response (pCR) after neoadjuvant chemotherapy (NACT) and patient survival was evaluated for distinct molecular subgroups and anti-HER2 treatment procedures. A total of 302 patients (17.9%) of the full cohort of 1691 patients had a tumor with a PIK3CA mutation, with a different prevalence in molecular subgroups: luminal/HER2neg 95 of 404 patients (23.5%), HER2pos 170 of 819 patients (20.8%) and TNBC 37 of 431 patients (7.9%). We identified mutations in PIK3CA exon 20 to be linked with worse response to anti-HER2 treatment (OR=0.507, 95%CI 0.320-0.802, p=0.004), especially in HR positive HER2 positive BC (OR=0.445, 95%CI 0.237-0.837, p=0.012). In contrast, exon 9 hotspot mutations p.E452K and p.E545K revealed no noteworthy differences in response therapy response. Luminal/HER2neg patients show a trend to have worse treatment response when PIK3CA was mutated. Interestingly, patients with residual disease after neoadjuvant treatment, have better survival when PIK3CA was mutated. PIK3CA hotspot mutation p.H1047R are associated with worse pCR rates after NACT in HER2pos BC, while hotspot mutations in exon 9 seems to have less impact.

Precise Molecular Subtyping Reveals Heterogeneity of Lung Adenocarcinoma Based on DNA Methylation

Background: Due to the high heterogeneity of lung adenocarcinoma (LUAD), which restricts the effectiveness of therapy, precise molecular subgrouping of LUAD is of great significance. Clinical research has demonstrated the significant potential of DNA methylation as a classification indicator for human malignancies. Methods: WGML framework (which was developed based on weighted gene correlation network analysis (WGCNA), Gene Ontology (GO), and machine learning) was developed to precisely subgroup molecular subtypes of LUAD. This framework included two parts: the WG algorithm and the machine learning part. The WG algorithm part was an original algorithm used to obtain a crucial module, which was characterized by weighted correlation network analysis, functional annotation, and mathematical algorithms. The machine learning part utilized the Boruta algorithm, random forest algorithm, and Gradient Boosting Regression Tree algorithm to select feature genes. Then, based on the results of the WGML framework, subtypes were computed by the hierarchical clustering algorithm. A series of analyses, including dimensionality reduction methods, survival analysis, clinical stage analysis, immune infiltration analysis, tumor environment analysis, immune checkpoints analysis, TIDE analysis, CYT analysis, somatic mutation analysis, and drug sensitivity analysis, were utilized to demonstrate the effectiveness of subgrouping. GEO datasets were used to externally validate the results. Meanwhile, another subgrouping method of LUAD from another study was employed to compare with the WGML framework. Result: By importing DNA methylation data into the WGML framework, nine genes were obtained to further subgroup LUAD. Three subtypes, the Carcinogenesis subtype, Immune-infiltration subtype, and Chemoresistance subtype, were identified. The dimensionality reduction method exhibited great distinctness between subtypes. A series of analyses were employed to exhibit the difference among the three subtypes and to demonstrate the accuracy of the definition of subtypes. Besides, the WGML framework was compared with a LUAD subgrouping method from another research, which demonstrated that WGML had better efficiency for subgrouping LUAD. Conclusion: This study provides a novel LUAD subgrouping framework named WGML for the accurate subgrouping of lung adenocarcinoma. result: By importing DNA methylation data into WGML framework, nine genes were obtained to further subgroup LUAD. Three subtypes, Carcinogenesis subtype, Immune-infiltration subtype, and Chemoresistance subtype were identified. Dimensionality reduction method exhibited great distinctness between subtypes. A series analyses were employed to exhibit the difference of three subtypes and to demonstrate accuracy of definition of subtypes. Besides, WGML framework was compared with a LUAD subgrouping method from another research, which demonstrated WGML had better efficiency for subgrouping LUAD.

Identifying molecular subgroups of patients with preeclampsia through bioinformatics.

Preeclampsia (PE) is a pregnancy-related disorder associated with serious complications. Its molecular mechanisms remain undefined; hence, we aimed to identify molecular subgroups of patients with PE using bioinformatics to aid treatment strategies. R software was used to analyze gene expression data of 130 patients with PE and 138 healthy individuals from the Gene Expression Omnibus database. Patients with PE were divided into two molecular subgroups using the unsupervised clustering learning method. Clinical feature analysis of subgroups using weighted gene co-expression network analysis showed that the patients in subgroup I were primarily characterized by early onset of PE, severe symptoms at disease onset, and induced labor as the main delivery method. Patients in subgroup II primarily exhibited late PE onset, relatively mild symptoms, and natural delivery as the main delivery method. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses revealed that the significant enrichment of calcium ion channels in subgroup II indicated the potential efficacy of calcium antagonists and magnesium sulfate therapy. In conclusion, the establishment of PE molecular subgroups can aid in diagnosing and treating PE.

Molecular profiling after short-term preoperative endocrine therapy to identify oestrogen receptor positive (ER+)/HER2+ early breast tumours with favourable prognosis.

560 Background: The impeded anti-proliferative response of Estrogen Receptor positive (ER+) HER2+ breast cancer (BC) to endocrine therapy (ET) has been considered negated by treatment with anti-HER2 therapies but residual disease after anti-HER2 therapy remains at risk of recurrence. Studying the molecular changes (MolC) of ER+HER2+ BC in response to ET will provide clinical insights to treatment options. Methods: POETIC was a phase III trial of post-menopausal patients with ER+ BC randomized 2:1 to 2-weeks of peri-operative aromatase inhibitors (POAI) vs control, followed by standard-of-care. Paired pre-treatment (B) and on-treatment (2wk) samples from 313 ER+HER2+ BC (213 POAI/100 controls) were gene expression profiled (BC 360 codeset; NanoString). Association of MolC with Early biological response to AI was assessed by residual Ki672wk (low ≤10%; high &gt;10%) by T-test, multiple testing corrected by Benjamini &amp; Hochberg (FDR); with time to recurrence (TTR) was estimated using multivariable Cox regression models adjusted for post-surgery clinicopathological variables and age as adjuvant treatment surrogate. Results: In POAI tumours, immunity-related signatures and mammary stemness were significantly upregulated in responders while proliferation, DNA-damage repair (DDR), TP53mutational status and ER-signaling were downregulated (FDR&lt;0.05). Controls had exclusive downregulation of PDL1 and upregulation of hypoxia (FDR &lt; 0.05). We previously identified 5 new molecular subgroups based on baseline gene expression (1- Immune high, ESR1 low; 2- ECM, ESR1 low, highest ERBB2; 3- DDR deficiency, 4- Endocrine signalling high and 5- Endocrine and PI3K/MAPK/RAS signalling high) which were associated with different response to AI and differential outcome. Endocrine-signalling, PI3K/MAPK/RAS signalling, tumour-immunity and chemokines were upregulated at a higher magnitude in AI sensitive subgroups (3 and 4) while the molecular subgroup with early resistance to AI and poorer outcome at baseline (2) did not show significant changes. Intrinsic subtype (IS) shifting was significantly more prevalent in treated (37%, 79/213) than in controls (14%, 14/100) (p &lt;0.001). In POAI most Luminal B (LumB) shifted to Luminal A (LumA; 79%, 59/75), driven by a reduction in proliferation. LumA2wk was associated with better outcome compared to LumB2wk (HR 0.2; CI95% 0.06-0.72, p=0.01). IS2wk provides additional information predicting TTR, than ISB (AIC value = 217.2 vs 221.6). Conclusions: We provide a comprehensive picture on how ER+HER2+ BC gene expression profile changes in response to ET. Most LumB BC shift to LumA, being LumA2wk IS significantly associated with better outcome. There is a clinical utility of peri-operative ET for ER+HER2+ BC, guiding who would have good prognosis for adjuvant ET and who may need additional therapy. Clinical trial information: NCT02338310 .

38MO Progression-free survival (PFS) in primary advanced or recurrent endometrial cancer (pA/rEC) in the overall and mismatch repair proficient (MMR/MSS) populations and in histological and molecular subgroups: Results from part 2 of the RUBY trial

38MO Progression-free survival (PFS) in primary advanced or recurrent endometrial cancer (pA/rEC) in the overall and mismatch repair proficient (MMR/MSS) populations and in histological and molecular subgroups: Results from part 2 of the RUBY trial