Government Owned Invention Available for License: DNA Methylation-Based Cancer Diagnostics for Accurate Tumor Classification

This technology encompasses a DNA methylation-based diagnostic platform designed to improve the accuracy and consistency of cancer classification, with demonstrated utility for tumors of the central nervous system, kidney, and hematopoietic system. By identifying disease-specific methylation signatures, the approach reduces interobserver variability and enhances diagnostic confidence. The central nervous system (CNS) classifier is built from a curated reference set of 16,567 methylation profiles and organizes tumors into 22 families and 133 clinically relevant diagnostic classes, including 21 newly developed methylation classes not represented in other existing tools. Across multiple independent validation cohorts (n = 5,875), the classifier demonstrated robust performance, and in a clinical-impact analysis of 1,204 NIH validation cases, methylation profiling materially influenced final diagnosis in 74.4% of cases by refining, increasing precision, or reclassifying. The CNS classifier was deployed as a user-facing software tool, MethylScape Analysis, which streamlines methylation-based classification workflows for CNS tumors, https://methylscape.ccr.cancer.gov/. The development of multiple specialized classifiers supports a more granular understanding of tumor biology and informed clinical decision-making.