Molecular Characterization of Breast Cancer with High-Resolution Oligonucleotide Comparative Genomic Hybridization Array
Molecular Characterization of Breast Cancer with High-Resolution Oligonucleotide Comparative Genomic Hybridization Array
复制标题
DOI：
10.1158/1078-0432.ccr-08-1791
复制
发表时间：
2009-01-15
影响因子：
11.5
通讯作者：
Pusztai, Lajos
中科院分区：
文献类型：
--
作者：
Andre, Fabrice;Job, Bastien;Pusztai, Lajos
目的：利用高分辨率的寡核苷酸比较基因组杂交(CGH)阵列和匹配的基因表达阵列数据来识别异常基因，并根据基因拷贝数异常对乳腺癌进行分类。实验设计：从106例接受术前化疗的II-III期乳腺癌患者的预处理细针活检标本中提取DNA。使用Agilent Human 4 x 44K阵列进行CGH。用Affymetrix U133A基因芯片产生的基因表达数据也可以在103名患者中获得。结果：单个肿瘤的拷贝数异常平均为76个(范围1-318)。在20%的样本中，分别有11个和37个明显的最小公共区域被获得或丢失。确定了几个潜在的治疗靶点，包括在10%的病例中显示高水平扩增的FGFR1。DNA拷贝数与mRNA表达水平密切相关。DNA拷贝数偏差的非负矩阵因式分解(NMF)聚类揭示了该数据集中的三个不同的分子类别。NMFI级的特点是三阴性癌的发生率很高(%)，收益为6p21。在三阴性肿瘤中，VEGFA、E2F3和NOTCH4也有29%到34%的增加。52%的NMF II级NMF中有ERBB2基因的获得，III级NMF中雌激素受体阳性的肿瘤发生率高(73%)，术前化疗的病理完全缓解率低(3%)。结论：本研究发现了可用于乳腺癌分类的异常基因，可能成为肿瘤分子亚型的新的治疗靶点。
Purpose: We used high-resolution oligonucleotide comparative genomic hybridization (CGH) arrays and matching gene expression array data to identify dysregulated genes and to classify breast cancers according to gene copy number anomalies.Experimental Design: DNA was extracted from 106 pretreatment fine needle aspirations of stage II-III breast cancers that received preoperative chemotherapy. CGH was done using Agilent Human 4 x 44K arrays. Gene expression data generated with Affymetrix U133A gene chips was also available on 103 patients. All P values were adjusted for multiple comparisons.Results: The average number of copy number abnormalities in individual tumors was 76 (range 1-318). Eleven and 37 distinct minimal common regions were gained or lost in >20% of samples, respectively. Several potential therapeutic targets were identified, including FGFR1 that showed high-level amplification in 10% of cases. Close correlation between DNA copy number and mRNA expression levels was detected. Nonnegative matrix factorization (NMF) clustering of DNA copy number aberrations revealed three distinct molecular classes in this data set. NMF class I was characterized by a high rate of triple-negative cancers (64%) and gains of 6p21. VEGFA, E2F3, and NOTCH4 were also gained in 29% to 34% of triple-negative tumors. A gain of ERBB2 gene was observed in 52% of NMF class II and class III was characterized by a high rate of estrogen receptor - positive tumors (73%) and a low rate of pathologic complete response to preoperative chemotherapy (3%).Conclusion: The present study identified dysregulated genes that could classify breast cancer and may represent novel therapeutic targets for molecular subsets of cancers.