Multiple myeloma (MM) remains an incurable plasma cell (PC) malignancy. Although it is known that MM tumor cells display extensive intratumoral genetic heterogeneity, an integrated map of the tumor proteomic landscape has not been comprehensively evaluated. We evaluated 49 primary tumor samples from newly diagnosed or relapsed/refractory MM patients by mass cytometry (CyTOF) using 34 antibody targets to characterize the integrated landscape of single-cell cell surface and intracellular signaling proteins. We identified 13 phenotypic meta-clusters across all samples. The abundance of each phenotypic meta-cluster was compared to patient age, sex, treatment response, tumor genetic abnormalities and overall survival. Relative abundance of several of these phenotypic meta-clusters were associated with disease subtypes and clinical behavior. Increased abundance of phenotypic meta-cluster 1, characterized by elevated CD45 and reduced BCL-2 expression, was significantly associated with a favorable treatment response and improved overall survival independent of tumor genetic abnormalities or patient demographic variables. We validated this association using an unrelated gene expression dataset. This study represents the first, large-scale, single-cell protein atlas of primary MM tumors and demonstrates that subclonal protein profiling may be an important determinant of clinical behavior and outcome.
多发性骨髓瘤(MM)仍是一种无法治愈的浆细胞恶性肿瘤。尽管已知MM肿瘤细胞存在显著的瘤内遗传异质性,但其肿瘤蛋白质组景观的综合图谱尚未得到全面评估。我们通过质谱流式细胞技术(CyTOF)分析了49例新诊断或复发/难治性MM患者的原代肿瘤样本,使用34种抗体靶标来绘制单细胞表面及细胞内信号蛋白的综合图谱。在所有样本中识别出13种表型超集,并将各表型超集的丰度与患者年龄、性别、治疗反应、肿瘤遗传异常及总生存期进行比对。其中若干表型超集的相对丰度与疾病亚型和临床行为相关。特征为CD45高表达、BCL-2低表达的表型超集1的丰度增加,与良好治疗反应及改善的总生存期显著相关,且该关联独立于肿瘤遗传异常或患者人口统计学变量。我们通过独立基因表达数据集验证了这一关联。本研究首次构建了大规模原代MM肿瘤的单细胞蛋白质图谱,并证明亚克隆蛋白质谱分析可能是临床行为和预后的重要决定因素。