Single cell analysis of multiple myeloma.

8026 Background: Multiple myeloma (MM) remains mostly incurable despite advances in treatments in the last decade. Major clinical challenges include: capturing inter- and intra- patient heterogeneity, stratification of asymptomatic patients, sensitive and functional identification of residual disease to tailor therapeutic strategies, as well as performing accurate liquid biopsies to mitigate the need for recurrent bone marrow (BM) sampling. Methods: To capture BM and circulating plasma cells (PC), we performed single cell RNA-seq (scRNA-seq). Index sorting enabled to link cell surface markers intensity on each cell with its expression profile. Analysis of the transcriptome together with the specific B cell receptor (BCR) using k-nearest-neighbor graphs characterizes the inter and intra tumor heterogeneity of myeloma patients in both the BM and blood at single cell resolution. Results: Analysis of more than 30,000 PC sorted from 35 patients and 12 control subjects, created a detailed map of 43 transcriptionally homogeneous subpopulations. These were based on cluster-specific expression patterns of more than 1500 genes. PC from healthy donors clustered together and displayed 4 states representing either long-lived PC or cycling short-lived PC. In contrast, each patient presented a unique PC transcriptional state, with a specific BCR clonotype. We uncover novel overexpressed genes, previously unimplicated in MM, such as LAMP5. We identify intra-tumor heterogeneity in 10/35 patients, for example, a patient with a FRZB and SMAD1 gene signature in 68% of PCs and an antimicrobial gene DEFB1 in the rest of the PC, which share the same BCR clonotype. We define new markers to differentiate normal circulating PC from tumor cells (CTC) in the blood and show that CTC mirror the BM state. Further, we accurately characterize the heterogeneity within MGUS patients into PC with a normal transcriptional state in contrast to small clones with a defined malignant program. Finally, we identify small cancer clones in the setting of minimal residual disease (MRD) following high-dose therapy. Conclusions: We show the power of scRNA-seq as a new, high resolution tool, to identify new pathways and potential targets in myeloma, stratify patients based on unique molecular programs, and identify residual disease in both the BM and blood.