Abstract:Background: Brain metastases (BM) are the most common intracranial tumors affecting about 8-10% of all cancer patients. Morphology and immunohistochemical staining are two common approaches used to identify the primary sites of BM samples, but morphology fails to identify poorly differentiated tumors and IHC markers usually lack specificity. About 2% to 14% of BM patients still present with unknown primary sites. A 90-gene assay, proposed in our previous study, is an RNA-based gene expression test to identify the tissue of origin in poorly differentiated and undifferentiated tumors. This study aims to evaluate the performance of the 90-gene assay in determining the primary sites for BM samples. Methods: The sequence-based gene expression profiles of 708 primary brain tumors (PBT) collected from The Cancer Genome Atlas database were performed by a 90-gene expression signature, with a similarity score for each of 21 tumor types. We used Optimal Binning algorithm to generate a threshold for separating PBT from BM. Eighteen PBT samples from Fudan University Shanghai Cancer Center were analyzed to substantiate reliability of the threshold. In addition, the performance of the 90-gene assay for identifying the tissue of origin was validated in a cohort of 48 BM samples with known origin from The First Affiliated Hospital, Zhejiang University. For each BM sample, the tumor type with the highest similarity score was considered tissue of origin. When a sample was diagnosed as PBT but the similarity score below the threshold, the second prediction was considered as primary site. Results: A threshold of the similarity score, 70, was identified to discriminate PBT from BM (PBT: ≥ 70, BM: < 70) with an accuracy of 99% (703/708). Eighteen PBT and 44 BM were performed by the 90-gene assay. The results of 18 PBT samples matched reference diagnosis with a concordance rate of 100% and all similarity scores were above 70. Of 44 BM samples, the 90-gene assay accurately predicted primary sites in 89% (39/44, 95%CI: 0.75-0.96) of the cases. Conclusions: The 90-gene assay showed promising discriminatory ability to separate PBT from BM and identify the primary site of BM. Our findings demonstrated the potential that 90-gene assay can serve as a powerful tool for accurately identifying the tissue of origin for BM samples. Legal entity responsible for the study: Yulong Zheng. Funding: Has not received any funding. Disclosure: Y. Sun, C. Chen, L. Chen, J. Zhu: Employment: Canhelp Genomics Co. Ltd. Q. Xu: Employment, stock ownership: Canhelp Genomics Co. Ltd. All other authors have declared no conflicts of interest.

Genome Methylation Accurately Predicts Neuroendocrine Tumor Origin: An Online Tool

Quantitative Analysis of Plasma Cell-Free DNA and Its DNA Integrity and Hypomethylation Status as Biomarkers for Tumor Burden and Disease Progression in Patients with Metastatic Neuroendocrine Neoplasias

Validation of a 90-Gene Assay for Tissue Origin Diagnosis of Brain Metastases.

A 92-gene cancer classifier predicts the site of origin for neuroendocrine tumors

Unclassifiable CNS tumors in DNA methylation-based classification: clinical challenges and prognostic impact

Construction of methylation driver gene-related prognostic signature and development of a new prognostic stratification strategy in neuroblastoma

DNA methylation-based classification of central nervous system tumours

DNA Methylation Profiling Enables Accurate Classification of Nonductal Primary Pancreatic Neoplasms

Abstract 4331: Machine learning-based classification of tissue origin of cancer using methylation profiles

Molecular identification of bronchopulmonary neuroendocrine tumours and neuroendocrine genotype in lung neoplasia using the NETest liquid biopsy

Decoding the DNA methylome of central nervous system tumors: An emerging modality for integrated diagnosis

Diagnostic classification based on DNA methylation profiles using sequential machine learning approaches.

Genome-Wide DNA Methylation Profiling as a Prognostic Marker in Pituitary Adenomas—A Pilot Study

Machine-learning operations streamlined clinical workflows of DNA methylation-based CNS tumor classification.

DNA methylation-based epigenetic signatures predict somatic genomic alterations in gliomas

Biom-19. Methylation Profiling Of Plasma Cell-Free Dna In Pediatric Brain Tumor Patients

Prognostic evaluation of the novel blueprint of DNA methylation sites by integrating bulk RNA-sequencing and methylation modification data in endometrial cancer

Exploring prognostic DNA methylation genes in bladder cancer: a comprehensive analysis

Prediction of brain metastasis development with DNA methylation signatures

Detection of glioma and prognostic subtypes by non-invasive circulating cell-free DNA methylation markers

Abstract 886: Integrated deep learning model for predicting DNA methylation and tumor types from histopathology in central nervous system tumors