For many tumuors, pathological subclasses exist which have to be further de
fined by genetic markers to improve therapy and follow-up strategies. In th
is study, cDNA array analyses of breast cancers have been performed to clas
sify tumuors into categories based on expression patterns. Comparing purifi
ed normal ductal epithelial cells and corresponding tumour tissues, the exp
ression of only a small fraction of genes was found to be significantly cha
nged. A subset of genes repeatedly found to be differentially expressed in
breast cancers was subsequently employed to perform a classification of 82
normal and malignant breast specimens by cluster analysis. This analysis id
entifies a subgroup of transcriptionally related tumours, designated class
A, which can be further subdivided into A1 and A2. Correlation with classic
al clinicopathological parameters revealed that subgroup A1 was characteriz
ed by a high number of node-positive tumours (14 of 16). In this subgroup t
here was a disproportionate number of patients who had already developed di
stant metastases at the time of diagnosis (25%, in this subgroup, compared
with 5%, among the rest of the samples). Taken together, the use of these d
ifferentially expressed marker genes in conjunction with sample clustering
algorithms provides a novel molecular classification of breast cancer speci
mens, which facilitates the identification of patients with a higher risk o
f recurrence. Copyright (C) 2001 John Wiley & Sons, Ltd.