Abstract
Multiple myeloma (MM) is monitored using bone marrow-based minimal residual disease (MRD) assays, which provide high sensitivity but require repeated invasive sampling and are limited to a single anatomical site. Cell-free DNA (cfDNA) offers a minimally invasive alternative, yet most cfDNA approaches rely on prior tumor profiling to identify patient-specific genomic alterations for subsequent tracking in plasma. Here, we developed MATADOR, a myeloma-specific, tumor-agnostic framework that leverages cfDNA fragmentation patterns at a fixed set of myeloma-specific chromatin-accessible regulatory regions. These regions were defined using single-nucleus ATAC-seq data from an independent MM cohort and used to quantify plasma cfDNA fragmentation at these loci without patient-specific tumor profiling. Using these locus-specific features alone, MATADOR distinguished clinically measurable from unmeasurable disease in a held-out test set with an area under the receiver operating characteristic curve (AUROC) of 0.94. These findings establish chromatin-informed cfDNA fragmentomics as a minimally invasive approach for disease assessment in MM without individualized tumor profiling, enabling blood-based assessment of disease beyond the constraints of bone marrow sampling.