The purpose of this retrospective study was to examine the prognostic value of expression of luminal epithelial antigen (LEA.135) for recurrence and overall survival of patients with primary invasive breast carcinoma by both univariate and multivariate analyses. The possible prognostic value of LEA.135 was also compared with some widely utilized prognostic biomarkers such as c-erbB 2, topoisomerase II.alpha (TPII.alpha), MIB 1, estrogen receptor (ER) and progesterone receptor (PR), as well as age of the patients and clinicopathologic parameters. The study was carried out by immunohistochemical methods on formalin-fixed/paraffin-embedded tissue sections in a series of 225 patients with median follow-up of 8.5 years. Prognostic significance of the biomarkers was determined by two-sided p value. In this series of patients, among the age and clinicopathologic parameters, only age, was significantly associated with a decreased overall survival (logrank p = 0.027). Among the prognostic biomarkers, TPII a expression at high (> 50% positive cells) or moderate (6-50% positive cells) level was associated with an increased rate of recurrence (logrank p < 0.001). However, the association of TPII.alpha expression with a decreased overall survival failed to reach a statistically significance. Expression of c-erbB 2 showed a trend of being associated with an increased probability of recurrence, but the association did not reach statistical significance. The remaining biomarkers were not associated with either the probability of recurrence or overall survival. LEA.135 expression was observed in 163 (72.4%) of the 225 patients. The patients with high (> 50% positive cells) or moderate (6-50% positive cells) level of LEA.135-positive cancer cells showed a significantly decreased probability of recurrence (logrank p < 0.001) and an increased overall survival (logrank p < 0.001) compared with those with LEA.135-negative cancer cells. The association remained significant by multivariate analysis for recurrence (likelihood ratio test p < 0.001) and overall survival (likelihood ratio test p < 0.001) when assessed with other prognostic parameters. Furthermore, the combination of LEA.135 with other prognostic biomarkers stratified four subgroups of patients with distinct clinical outcome. The subgroup of patients who were LEA.135+/TPII.alpha- showed the lowest probability of recurrence and the longest overall survival compared with those who were LEA.135-/TPII.alpha+ (logrank p < 0.001). Interestingly, the patients whose cancer cells were LEA.135+/TPII.alpha+, LEA.135+ MIB.1+ or LEA.135+/c-erbB 2+ experienced a decreased probability of recurrence and an increased overall survival compared with those with LEA.135-/TPII.alpha+, LEA.135- MIB.1+ or LEA.135-/c-erbB 2+ (logrank p < 0.001). The results demonstrated that LEA.135 is an independent and favorable prognostic biomarker for patients with primary invasive breast carcinoma, that the loss of LEA.135 expression is associated with aggressive phenotype of cancer cells during the breast cancer progression, and that its continued expression seems to override the adverse effects of expression of an oncogene or cell proliferation-associated molecules.