Idbaih Ahmed, Marie Yannick, Lucchesi Carlo, Pierron Gaëlle, Manié Elodie, Raynal Virginie, Mosseri Véronique, Hoang-Xuan Khê, Kujas Michèle, Brito Isabel, Mokhtari Karima, Sanson Marc, Barillot Emmanuel, Aurias Alain, Delattre Jean-Yves, Delattre Olivier
INSERM, Unité 830, Paris, France.
Int J Cancer. 2008 Apr 15;122(8):1778-86. doi: 10.1002/ijc.23270.
The pathological classification of gliomas constitutes a critical step of the clinical management of patients, yet it is frequently challenging. To assess the relationship between genetic abnormalities and clinicopathological characteristics, we have performed a genetic and clinical analysis of a series of gliomas. A total of 112 gliomas were analyzed by comparative genomic hybridization on a BAC array with a 1 megabase resolution. Altered regions were identified and correlation analysis enabled to retrieve significant associations and exclusions. Whole chromosomes (chrs) 1p and 19q losses with centromeric breakpoints and EGFR high level amplification were found to be mutually exclusive, permitting identification of 3 distinct, nonoverlapping groups of tumors with striking clinicopathological differences. Type A tumors with chrs 1p and 19q co-deletion exhibited an oligodendroglial phenotype and a longer patient survival. Type B tumors were characterized by EGFR amplification. They harbored a WHO high grade of malignancy and a short patient survival. Finally, type C tumors displayed none of the previous patterns but the presence of chr 7 gain, chr 9p deletion and/or chr 10 loss. It included astrocytic tumors in patients younger than in type B and whose prognosis was highly dependent upon the number of alterations. A multivariate analysis based on a Cox model shows that age, WHO grade and genomic type provide complementary prognostic informations. Finally, our results highlight the potential of a whole-genome analysis as an additional diagnostic in cases of unclear conventional genetic findings.
胶质瘤的病理分类是患者临床管理的关键步骤,但往往具有挑战性。为了评估基因异常与临床病理特征之间的关系,我们对一系列胶质瘤进行了基因和临床分析。使用分辨率为1兆碱基的BAC芯片通过比较基因组杂交对总共112例胶质瘤进行了分析。确定了改变的区域,并通过相关性分析检索出显著的关联和排除情况。发现1号染色体短臂(1p)和19号染色体长臂(19q)着丝粒断点缺失与表皮生长因子受体(EGFR)高水平扩增相互排斥,从而识别出3个不同的、不重叠的肿瘤组,它们具有显著的临床病理差异。具有1p和19q共同缺失的A型肿瘤表现出少突胶质细胞表型,患者生存期较长。B型肿瘤的特征是EGFR扩增。它们具有世界卫生组织(WHO)高度恶性等级,患者生存期短。最后,C型肿瘤未表现出上述任何模式,但存在7号染色体增加、9号染色体短臂缺失和/或10号染色体缺失。它包括比B型患者年龄更小的星形细胞瘤患者,其预后高度依赖于改变的数量。基于Cox模型的多变量分析表明,年龄、WHO分级和基因组类型提供了互补的预后信息。最后,我们的结果突出了全基因组分析作为常规基因发现不明确情况下的额外诊断方法的潜力。