OAB-008: Identification of high-risk Multiple Myeloma with a plasma cell Leukemia-like transcriptomic profile

Abstract

<h3>Background</h3> Primary plasma cell leukemia (pPCL) is an aggressive subtype of multiple myeloma that is characterized by ≥20% circulating tumor cells (CTCs), whereas CTC levels in newly diagnosed multiple myeloma (NDMM) are <20%. Currently, there is no molecular marker for pPCL. It therefore remains to be elucidated if certain NDMM tumors molecularly resemble pPCL and if this has any prognostic significance in the context of conventional high-risk (HR) markers in NDMM. Aims: (1) To construct and validate a transcriptomic classifier for PCL-like disease. (2) To test its value as independent prognostic marker in NDMM. <h3>Methods</h3> Transcriptomic data were generated of CD138-enriched bone marrow (BM) plasma cells from both NDMM patients enrolled in the Cassiopeia (NCT02541383) and HO143 (EudraCT 2016-002600-90) trials and pPCL patients from the EMN12/HO129 (EudraCT 2013-005157-75) trial. NDMM CTC levels were determined with the EuroFlow NGF protocol. HR FISH was defined as the presence of either t(4;14), t(14;16) or del17p.154 NDMM and 29 pPCL patients were divided into a discovery and validation cohort to construct and test a classifier for PCL-like disease. Subsequently, data from 8 additional NDMM cohorts were used to assess the association of PCL-like status with progression-free survival (PFS) and overall survival (OS) in Cox proportional hazards (PH) models including conventional HR markers. <h3>Results</h3> Baseline CTC levels were determined in 297 NDMM and 51 pPCL patients. CTCs could be detected in 87% of NDMM patients, with a limit of detection <10-5. CTC levels were positively associated with tumor burden (BM plasmacytosis). In the discovery cohort, 1700 genes correlated with high CTC levels (FDR <0.05), independently of tumor burden. These included genes involved in cell adhesion, cell migration and proliferation. After optimization by leave-one-out cross-validation, a classifier for PCL-like disease was built with a selection of 54 genes. This showed a sensitivity of 93% to identify pPCL in the validation cohort, but also classified 10% of NDMM tumors as PCL-like. In a cohort of 2139 NDMM patients, Cox PH regression analyses confirmed a significant and independent association of PCL-like status with PFS and OS, in the context of either R-ISS stage, ISS stage, HR FISH, SKY92 HR or UAMS70 HR status. PCL-like status in combination with R-ISS, age and treatment status showed a hazard ratio of 1.64 (95% CI, 1.30-2.07, p<0.0001) for PFS and 1.89 (95% CI, 1.42-2.50, p<0.0001) for OS. PCL-like status conferred a median OS of only 13.2 months among NDMM patients with R-ISS III. <h3>Conclusions</h3> (1) pPCL cannot only be identified clinically, but also molecularly. (2) PCL-like status is a novel marker for HR disease in NDMM that identifies patients with a tumor transcriptome similar to pPCL and has independent prognostic value in the context of conventional HR markers. (3) PCL-like status could help detect NDMM patients with early stage or borderline pPCL.