The presence or absence of estrogen and progesterone steroid hormone receptor expression (ER, PR) is an essential feature of invasive breast cancer and determines prognosis and endocrine treatment decisions. Among the four ER/PR receptor phenotypes, the ER-/PR+ is infrequent, and its clinical relevance has been controversially discussed. Thus, we investigated its clinical significance and gene expression pattern in large datasets. In a retrospective clinical study of 15,747 breast cancer patients, we determined the ER/PR subtype survival probabilities using Kaplan-Meier and Cox regression analyses. From The Cancer Genome Atlas (TCGA) breast cancer dataset, PAM50 expression signature and pathway analyses were performed to test for distinct molecular features. In our cohort, the ER-/PR+ phenotype has been observed at a frequency of 4.1 % and was associated with an improved 10-year survival for stage I cancers compared to the ER+/PR+ reference subtype (median; 95 % CI 88.1 %; 83-93 vs. 84.3 %; 82-86 %, P = 0.024) as was confirmed by multivariate analysis over the entire follow-up (HR 0.59, 95 % CI 0.38-0.92, P = 0.021). This association lacked significance when including all stages. ER-/PR+ patients treated with antihormonal agents (34.5 %) had shorter survival compared to their non-treated counterparts (Log-rank P = 0.0001). PAM50 signatures suggest a distinct configuration for the ER-/PR+ phenotype. This specific phenotype has been further separated by a set of 59 uniquely expressed genes. Our study supports the notion of the existence of an ER-/PR+ phenotype with clinical and molecular features distinct from the large group of ER+/PR+ patients.
Keywords: Breast cancer; Endocrine treatment; Estrogen receptor subtype; Progesterone receptor subtype; TCGA.