Identifying a high-risk cellular signature in the multiple myeloma bone marrow microenvironment_2

The multiple myeloma (MM) tumor microenvironment is thought to influence patient outcomes. To test this, we computationally enumerated relevant cell populations in 436 newly diagnosed MM patients. Unsupervised clustering identified 5 clusters, with patients in Cluster 5 having significantly worse outcomes: 13 fewer months of progression-free survival (P = 0.002) and 8 fewer months of overall survival (P = 0.040). Cell type analysis showed that patients in Cluster 5 had elevated CD8+ T cell and B cell populations, but low granulocyte levels. A granulocyte signature identified an additional 14% of patients with elevated risk but lacking International Staging System stage III or GEP-70 high- risk status To determine how the cellular content in the microenvironment contributes to patient outcome RNA extracted from whole bone marrow (WBM) biopsies were analyzed subtracting the signal derived from the malignant plasma cells to focus on the non-tumor cells. A custom algorithm was developed to define and quantitate the impact of cellular subcomponents of the bone marrow (BM) microenvironment on outcome and applied it to a series of WBM trephine biopsies.