Molecular Characterization of Bone Marrow and Peripheral Blood Compartments from Patients with Plasma Cell Leukemia in the Mmrf Commpass Study

Abstract

Plasma cell leukemia (PCL) is rare but represents an aggressive, advanced form of multiple myeloma (MM) where neoplastic plasma cells (PCs) escape the bone marrow (BM) and circulate in the peripheral blood (PB). Traditionally, PCL is defined by the presence of >20% circulating plasma cells (CPCs), however, recent studies have suggested that PCL be redefined as the presence of >5% CPCs. The Multiple Myeloma Research Foundation CoMMpass study (NCT01454297) is a longitudinal, observational clinical study with 1143 newly diagnosed MM patients. BM-derived MM samples were characterized using whole genome (WGS), exome (WES), and RNA (RNAseq) sequencing at diagnosis and each progression event. When >5% CPCs were detected by flow cytometry, PCs were enriched independently from both compartments, and T-cells were selected from the PB as a control for WGS and WES. This substudy within CoMMpass provides the largest, most comprehensively characterized dataset of matched MM and PCL samples to date, which can be leveraged to better understand the molecular drivers of PCL. At diagnosis, 813/1143 CoMMpass patients had flow cytometry data reporting the percent PCs in PB, of which 790 had <5%, 17 had 5-20%, and 6 had >20% CPCs. Survival analyses revealed that patients with 5-20% CPCs (median = 20 months) had poor overall survival (OS) outcomes compared to patients with <5% CPCs (median = 74 months, p < 0.001), and no significant difference in outcome was observed between patients with 5-20% and >20% (median = 38 months) CPCs. Patients with 1-5% CPCs (median = 50 months, HR = 2.45, 95% CI = 1.64 - 3.69, p < 0.001) also exhibited poor OS outcomes compared to patients with <1% CPCs (median = 74 months), suggesting that patients with >1% CPCs are a higher risk population, even if they do not meet the PCL threshold. Using a cutoff of >5% CPCs, 23/813 (2.8%) patients presented with primary PCL (pPCL) at diagnosis. Of these patients, 7 (30%) were hyperdiploid (HRD), of whom 1 had a CCND1 and 1 had a MYC translocation; while 16 (70%) were nonhyperdiploid (NHRD), all of whom had a canonical immunoglobulin translocation (6 CCND1, 5 WHSC1, 3 MAF, 1 MAFA, and 1 MAFB). Of 124 patients with serial sample collections, 5 (4%) patients without pPCL had >5% CPCs at progression, and thus relapsed with secondary PCL (sPCL). Of the 5 sPCL patients, 2 (40%) were NHRD with a CCND1 or MAF translocation; while 3 (60%) were HRD, 1 with a WHSC1 translocation. Median time to diagnosis of sPCL was 22 months (range = 2 - 31 months), and patients with sPCL (median = 22 months) and pPCL (median = 30 months) exhibited poor OS outcomes as compared to MM patients (74 months, p < 0.001). Sequencing data was available for 15 pPCL and 5 sPCL samples. For 12 patients with WES, WGS, and RNAseq performed on their PCL tumor sample, an integrated analysis identified recurrent, complete loss-of-function (LOF) events in only CDKN2C/FAF1, SETD2, and TRAF3. Five pPCL patients had complete LOF of a gene involved in G1/S cell cycle control, including CDKN2C, CDKN2A, CDKN1C, and ATM. These LOF events were not observed in NHRD t(11;14) PCL patients, suggesting that CCND1 overexpression and LOF of genes involved in G1/S cell cycle control may represent independent drivers of PCL. Comparing WES and WGS data between matched MM and PCL tumor samples revealed a high degree of similarity in mutation and copy number profile. However, differential expression analysis performed for 13 patients with RNAseq data comparing their MM and PCL tumors revealed 27 up- and 39 downregulated genes (padj < 0.01, FDR = 0.1) in PCL versus MM. Pathway analysis revealed an enrichment (p < 0.001) for genes involved in adhesion and diapedesis, including upregulation of ITGB2, PF4, and PPBP, and downregulation of CCL8, CXCL12, MMP19, and VCAM1. The most significantly downregulated gene in PCL (log2FC = -6.98) was VCAM1, which plays a role in cell adhesion, and where loss of expression (TPM < 0.01) was observed across all PCL samples. Upregulation of four S100 genes including S100A8, S100A9, S100A12, and S100P, which have been implicated in tumor growth, metastasis, and immune evasion, was also observed in PCL. Interestingly, a S100A9 inhibitor has been developed and may represent a novel treatment option for PCL patients. In summary, PCL was found to be associated with molecular events dysregulating G1/S cell cycle control coupled with subtle changes in transcription that likely occur in a subclonal population of the MM tumor. Disclosures Lonial: Genentech: Consultancy; GSK: Consultancy; BMS: Consultancy; Janssen: Consultancy, Research Funding; Karyopharm: Consultancy; Takeda: Consultancy, Research Funding; Celgene Corporation: Consultancy, Research Funding; Amgen: Consultancy.