Loi S et al. (2007) Definition of clinically distinct molecular subtypes in estrogen receptor-positive breast carcinomas through genomic grade. J Clin Oncol 25: 1239–1246

Distinct molecular subtypes of breast cancer, defined through microarray analysis, have been shown to have different clinical outcomes. Basal and ErbB2-type breast cancers are frequently recognized, but distinguishing the estrogen receptor (ER)-positive subtypes is problematic, not least because no simple diagnostic test exists. Loi et al. used their previously developed gene-expression grade index (GGI) to identify two distinct molecular subtypes of ER-positive breast cancer. The index works by averaging the expression level of a gene set (up to 97 genes) in order to determine the concordance between a tumor sample and histological grade.

ER-positive breast cancers were categorized as being of high or low genomic grade by use of the GGI, and classifications were compared with those made by existing molecular tools. The two ER-positive molecular subtypes identified by the GGI had similar biological pathways, but correlated closely with the luminal A and B subtypes by hierarchical cluster analysis, and showed strong agreement with the risk subgroups produced by the 21-gene recurrence score. High genomic grade was associated with a worse clinical outcome than was low genomic grade in patients treated with adjuvant tamoxifen only, as well as in those who were untreated. The GGI was a stronger prognostic variable than traditional clinical prognostic factors in univariate and multivariate analyses.

The authors conclude that the GGI is simple to use and gives highly reproducible results across multiple data sets. They note that these data reinforce the role of cell-cycle and grade-related genes in determining prognosis in patients with ER-positive breast cancers.