RESUMEN
<p><b>OBJECTIVE</b>To screen molecular markers in early breast cancer and establish gene subtyping-based diagnostic criteria for predicting the prognosis of early breast cancers.</p><p><b>METHODS</b>Tumor tissue specimens were obtained from 8 patients with early breast cancer for analysis of the differentially expressed genes using Agilent custom 8×15 000 chips in combination with the prognostic data of the patients. Another 42 tumor tissue specimens were used to validate the differential genes by real-time fluorescent quantitative PCR.</p><p><b>RESULTS</b>Gene microarray analysis identified 132 differentially expressed genes between the patients with favorable and poor prognosis, and 44 of these genes were significantly up-regulated (by over two folds) and 88 down-regulated in patients with poor prognoses.</p><p><b>CONCLUSION</b>The gene expression profiles differ in early breast cancer tissues of the same pathological type but with different clinical stages and prognoses, and CD44, MKI67, NTRK2, Nek2, C16orf60, TOP2A, ANCCA, and RRM2 genes can be used as the prognostic markers for early breast cancer.</p>