Paper

Dissecting the Heterogeneity of Luminal Subtype Breast Cancer Using Gene Component Analysis


Authors:
Eric Y. Chuang; Shin-Hsiu Tu; Chi-Cheng Huang
Abstract
Breast cancer is a heterogeneous disease in terms of molecular aberrations. Microarray technology in the past decade has redefined breast cancer as a group of distinct disease based on gene expression profiles of certain intrinsic genes. Luminal subtype breast cancers, most of which are also estrogen receptor (ER) positive clinically, constitute the majority of human breast cancers and better prognosis is reported compared with ER negative breast cancers such as basal-like or HER2-enriched subtype. Significant difference in survival and therapeutic response, however, is observed between Luminal A and Luminal B subtype of breast cancers. The aim of the study is to use gene component analysis for the classification of Luminal A and Luminal B breast cancer and candidate genes constituting the gene component classifiers may reveal therapeutic targets as potential biomarkers in further breast cancer treatment. A total of 169 breast cancer microarray experiments from Taiwan and Mainland China with Han Chinese ethnic origin were analyzed and 80 out of which were consistently designated into Luminal A or Luminal B subtype by Hu306 and PAM50 intrinsic signatures under gene-centring. Partial least square followed by discriminative analysis (PLS-DA) algorithm from the Top 40 of the 2576 filtered genes with the p<10-4 t-statistic selection criteria delivered the best predictive model with parsimony. Unsupervised hierarchical clustering from the Top 40 genes also revealed well segregation of Luminal A and Luminal B breast cancers with only three errors. From our study, gene component analysis could discriminate Luminal A and Luminal B breast cancers and several proliferation-related genes over-expressed in the Luminal B subtype with compromised prognosis were identified.
Keywords
Luminal Breast Cancer; Microarray; Gene Expression; Gene Component Analysis; Han Chinese
StartPage
21
EndPage
27
Doi
Download | Back to Issue| Archive