An integrative multiparametric approach stratifies distinct phenotypes of blast phase chronic myelomonocytic leukemia with differential maturation status and drug sensitivity profiles

Approximately 30% of chronic myelomonocytic leukemia patients undergo transformation to a chemo-refractory blastic phase (BP-CMML). Seeking novel therapeutic approaches we profiled blast transcriptomes from 42 BP-CMMLs, observing extensive transcriptional heterogeneity and poor alignment to current AML classifications. BP-CMMLs displayed distinctive transcriptomic profiles, including enrichment for quiescence and variability in drug response signatures. Integrating clinical, immunophenotype and transcriptome parameters, RandomForest unsupervised clustering distinguished immature and mature subtypes characterized by differential expression of transcriptional modules, oncogenes, apoptotic regulators and patterns of surface marker expression. Subtypes differed in predicted response to AML drugs, validated ex vivo in primary samples. Iteratively refined stratification resolved a classification structure comprising five subtypes along a maturation spectrum, predictive of response to novel agents including consistent patterns for RTK, CDK, MTOR and MAPK inhibitors. Finally, we generated a clinically-applicable decision tree to stratify BP-CMML with high specificity and sensitivity, and potentially guide personalized drug selection for improved outcomes. RNA sequencing was performed on the available cryopreserved primary samples from BP-CMML and healthy control donors. Samples were enriched for live cells and FACS sorted for isolation of cell types of interest, followed by RNA extraction and subsequent library preparation and sequencing.

约30%的慢性粒单核细胞白血病（chronic myelomonocytic leukemia, CMML）患者会进展为化疗难治性原始细胞阶段（chemo-refractory blastic phase, BP-CMML）。为探索新型治疗策略，我们对42例BP-CMML患者的原始细胞转录组进行了测序分析，观察到广泛的转录组异质性，且与当前急性髓系白血病（acute myeloid leukemia, AML）的分型系统匹配度欠佳。BP-CMML具有独特的转录组特征，包括细胞静息状态富集以及药物反应特征谱的异质性。整合临床信息、免疫表型数据与转录组参数后，通过随机森林（RandomForest）无监督聚类区分出未成熟亚型与成熟亚型，两类亚型在转录模块、致癌基因、凋亡调控因子的表达差异以及表面标志物表达模式上均存在特征性区别。不同亚型对AML药物的预测响应存在差异，该结果在原代样本的体外实验中得到了验证。经迭代优化的分层分型分析构建了沿细胞成熟梯度分布的5个亚型分类体系，该体系可预测患者对新型治疗药物的响应，其中针对受体酪氨酸激酶（RTK）、细胞周期蛋白依赖性激酶（CDK）、雷帕霉素靶蛋白（MTOR）以及丝裂原活化蛋白激酶（MAPK）抑制剂的响应模式均具有一致性。最后，我们构建了一款临床适用的决策树模型，能够以高特异性与高灵敏度对BP-CMML进行分型，或可指导个性化用药选择以改善患者预后。本研究对来自BP-CMML患者与健康对照供体的冻存原代样本开展了RNA测序（RNA sequencing, RNA-seq）：首先对样本进行活细胞富集，再通过荧光激活细胞分选（FACS）分离目标细胞群，随后进行RNA提取、文库制备及测序。