Enabling blood-based mass spectrometry MRD detection in multiple myeloma without archived serum samples.

7556 Background: Minimal residual disease (MRD) assessment in multiple myeloma (MM) relies almost exclusively on bone marrow (BM) biopsy, an invasive, costly, and spatially limited procedure that is poorly suited for frequent longitudinal monitoring and may miss focal or extramedullary disease. Blood-based MRD strategies offer a less invasive, lower-cost alternative that could enable more frequent surveillance and earlier detection of disease progression. Mass spectrometry (MS)-based detection of monoclonal immunoglobulin peptides in peripheral blood is a promising MRD approach; however, current methods typically require prior identification of circulating M-protein amino acid sequences from archived serum, limiting applicability in patients without banked samples. We sought to determine whether patient-specific immunoglobulin V(D)J sequences could be reconstructed from BM-derived DNA or RNA, or from peripheral blood RNA, to enable blood-based MRD detection. Methods: We analyzed RNA-seq and whole exome sequencing (WES) from BM plasma cells and RNA-seq from peripheral blood mononuclear cells in 75 patients with MM, all of whom had abnormal plasma cells identified by flow cytometry. Immunoglobulin heavy chain (IGH) V(D)J sequences were reconstructed using MiXCR, and the dominant productive clonotype in the BM was selected per patient. Variable-region sequences containing complementarity-determining region 3 (CDR3) and adjacent framework regions were extracted when available. Peripheral blood RNA-seq was evaluated for clonotype recovery without BM sampling. In silico proteolytic digestion was performed to identify peptides compatible with standard MS detection. Results: Across 75 patients, dominant IGH clonotypes were reconstructed from 100% of BM sequencing data. From these clonotypes, a mean of 7 RNA-seq–derived peptides and 6 WES-derived peptides per patient were identified, with a mean of 2 peptides shared between platforms. Fifty patients (67%) generated ≥10 MS-compatible peptides, and 33 patients (44%) had at least one peptide shared between RNA-seq and WES. Among 51 patients with paired tumor and peripheral blood RNA-seq, dominant clonotypes were detectable in peripheral blood by CDR3 matching in 30 patients (58.8%). Conclusions: Patient-specific immunoglobulin clonotypes can be reliably reconstructed from prior BM sequencing data, yielding MS-compatible peptide targets for downstream analysis. In selected patients, clonotype discovery is feasible directly from peripheral blood, potentially reducing the need for invasive BM sampling. This approach overcomes a key limitation of existing MS-based MRD assays by enabling peptide target discovery in patients without archived serum but with available BM sequencing. Additional studies are required to validate MS detection of these peptides and establish a scalable, minimally invasive blood-based MRD platform.