Soluble BCMA Defines Spatial Phenotypes and Prognosis in Newly Diagnosed Multiple Myeloma

In multiple myeloma (MM), defining tumor burden is difficult because systemic bone marrow disease and focal lesions often show different levels of activity. We analyzed 352 newly diagnosed patients. We used bone marrow plasma cell percentage (BMPC) quantified from whole slide images, MRI-total diffusion volume (TDV), PET-metabolic tumor volume (TMTV), and serum sBCMA levels. In patients with complete datasets, Principal Component Analysis (PCA) and unsupervised clustering were used to classify patients into four phenotypes based on two factors: physical tumor volume (based on BMPC, TDV and TMTV) and sBCMA levels. Among all markers evaluated, sBCMA showed the strongest and most consistent prognostic association, independent of imaging-defined tumor volume. The four phenotypes were: Dual-High (high volume / high sBCMA), Sero-dominant (low volume / high sBCMA), Macro-dominant (high volume / low sBCMA), and Dual-Low (low volume / low sBCMA). Elevated sBCMA, rather than tumor volume, was the primary determinant of early treatment failure; the Sero-dominant and Dual-High groups showed similar 12-month progression-free survival despite markedly different tumor volumes. The Macro-dominant phenotype represented a state of "spatial sequestration," characterized by extensive imaging-defined tumor burden that remained anatomically confined without hematogenous dissemination or extramedullary spread. In this phenotype, the prognostic impact of Gain(1q21) was not significant, whereas it was strongly associated with inferior survival in the Sero-dominant phenotype. MM outcomes are determined by the interaction between physical tumor volume and sBCMA levels. This multimodal classification provides a rational framework for risk stratification.