Chuang Ming-Kai, Chiu Yu-Chiao, Chou Wen-Chien, Hou Hsin-An, Tseng Mei-Hsuan, Kuo Yi-Yi, Chen Yidong, Chuang Eric Y, Tien Hwei-Fang
Department of Laboratory Medicine, National Taiwan University Hospital, Taipei, Taiwan.
Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan.
Oncotarget. 2015 Nov 17;6(36):39098-110. doi: 10.18632/oncotarget.5390.
Although clinical features, cytogenetics, and mutations are widely used to predict prognosis in patients with acute myeloid leukemia (AML), further refinement of risk stratification is necessary for optimal treatment, especially in cytogenetically normal (CN) patients. We sought to generate a simple gene expression signature as a predictor of clinical outcome through analyzing the mRNA arrays of 158 de novo CN AML patients. We compared the gene expression profiles of patients with poor response to induction chemotherapy with those who responded well. Forty-six genes expressed differentially between the two groups. Among them, expression of 11 genes was significantly associated with overall survival (OS) in univariate Cox regression analysis in 104 patients who received standard intensive chemotherapy. We integrated the z-transformed expression levels of these 11 genes to generate a risk scoring system. Higher risk scores were significantly associated with shorter OS (median 17.0 months vs. not reached, P < 0.001) in ours and another 3 validation cohorts. In addition, it was an independent unfavorable prognostic factor by multivariate analysis (HR 1.116, 95% CI 1.035~1.204, P = 0.004). In conclusion, we developed a simple mRNA expression signature for prognostication in CN-AML patients. This prognostic biomarker will help refine the treatment strategies for this group of patients.
尽管临床特征、细胞遗传学和突变被广泛用于预测急性髓系白血病（AML）患者的预后，但为了实现最佳治疗，尤其是对于细胞遗传学正常（CN）的患者，有必要进一步优化风险分层。我们试图通过分析158例初发CN AML患者的mRNA阵列，生成一个简单的基因表达特征作为临床结局的预测指标。我们比较了诱导化疗反应不佳患者与反应良好患者的基因表达谱。两组之间有46个基因表达存在差异。其中，在接受标准强化化疗的104例患者中，单因素Cox回归分析显示11个基因的表达与总生存期（OS）显著相关。我们整合了这11个基因的z转换表达水平，以生成一个风险评分系统。在我们的队列以及另外3个验证队列中，较高的风险评分与较短的OS显著相关（中位数17.0个月对未达到，P < 0.001）。此外，多因素分析显示它是一个独立的不良预后因素（HR 1.116，95% CI 1.035~1.204，P = 0.004）。总之，我们为CN-AML患者开发了一个简单的mRNA表达特征用于预后评估。这种预后生物标志物将有助于优化这组患者的治疗策略。
