Acute myeloid leukemia samples of samples =< 60yrs on HG-U133 plus 2. Homo sapiens

The pretreatment karyotype of leukemic blasts is currently the key determinant in therapy decision-making in acute myeloid leukemia (AML). However, approximately fifty percent of AML patients, often carrying a normal karyotype, are currently unclassifiable based these established methods. Gene expression profiling has proven to be valuable for risk stratification of AML. The gene expression profiles of AML samples of two independent cohorts (n=247 and n=214) were determined using Affymetrix U133Plus2.0 GeneChips: all Samples below 4000 are in the training cohort; all Samples higher than 4000 are in the validation cohort. Data analyses were carried out to discover and predict prognostically relevant subtypes in AML (<60 years) based on their gene expression signatures. Statistical analyses were performed to determine the prognostic significance of cases of AML with specific molecular signatures. Unsupervised cluster analyses of the gene expression signatures of both independent cohorts of AML patients confirmed that chromosomal lesions and mutations, often resulting in aberrant transcription factors, induce discriminatory patterns of gene expression. In contrast, however, mutations in signalling molecules do not establish strong molecular signatures. Consequently, prognostically important subtypes, which express mutated trancription factors were predicted with high accuracy using minimal sets of genes. We identified several novel clusters, some consisting of patients with normal karyotypes. Gene expression profiling allows classification of AML subtypes characterized by the expression of abnormal transcription factors, however, prediction of clinically relevant mutations affecting signalling molecules is impossible and thus still requires addition molecular methods. Keywords: acute myeloid leukemia, patient blood or bone marrow samples Overall design: 461 blood or bone marrow samples of acute myeloid leukemia patients were hybridized to Affymetrix HG-U133 plus 2 GeneChips. 76 additonal samples added on 7/28/2011.

急性髓系白血病（acute myeloid leukemia, AML）患者的白血病原始细胞预处理核型，目前是临床治疗决策的核心判断依据。然而，约半数AML患者（多携带正常核型）无法通过现有既定方法进行分型。基因表达谱分析（Gene expression profiling）已被证实可有效用于AML的风险分层。本研究针对两个独立队列（样本量分别为247例和214例）的AML样本开展基因表达谱分析，所用检测平台为Affymetrix U133Plus2.0基因芯片（Affymetrix U133Plus2.0 GeneChips）：样本编号低于4000的纳入训练队列（training cohort），编号高于4000的纳入验证队列（validation cohort）。研究通过数据分析，基于基因表达特征挖掘并预测了60岁以下AML患者的预后相关亚型；通过统计学分析明确了携带特定分子特征的AML病例的预后价值。对两个独立队列的AML患者基因表达特征进行无监督聚类分析（Unsupervised cluster analyses）后证实，染色体异常与突变（常导致转录因子（transcription factors）功能异常）可诱导具有区分性的基因表达模式；而信号分子（signalling molecules）突变则无法形成显著的分子特征谱。因此，携带突变转录因子的预后重要亚型，可通过最小基因集（minimal sets of genes）以高准确率实现预测。本研究鉴定出多个新的聚类亚型，其中部分包含正常核型患者。基因表达谱分析可对以异常转录因子表达为特征的AML亚型进行精准分类，但无法预测影响信号分子的临床相关突变，因此此类突变的检测仍需辅以其他分子生物学方法。关键词：急性髓系白血病，患者血液或骨髓样本。研究整体设计：共纳入461例AML患者的血液或骨髓样本，用于Affymetrix HG-U133 plus 2基因芯片（Affymetrix HG-U133 plus 2 GeneChips）杂交检测；2011年7月28日新增76例样本。