Abstract
BACKGROUND: It is unknown whether ER(-)/PR(+) breast cancer is an independent breast cancer subtype, how it differs from other subtypes, and what its significance is regarding treatment and prognosis. This study compared ER(-)/PR(+) breast cancer with other subtypes to better understand the biological characteristics and prognosis of ER(-)/PR(+) breast cancer, to guide clinical treatment and establish a theoretical foundation. METHODS: We retrospectively analyzed data for patients diagnosed with breast cancer in the Surveillance, Epidemiology, and End Results (SEER) database. The clinicopathological characteristics of ER(-)/PR(+) breast cancer, including age, tumor size, lymph node status, HER-2 status, pathological type and histological grade, were compared with other types of breast cancer. A risk scoring system was developed based on independent risk factors influencing prognosis to predict the patient's prognosis, and a nomogram model was created to predict the patient's survival rate. Receiver operating characteristic curve (ROC) and calibration curve was used to evaluate the predictive performance of the nomogram. RESULTS: The rates of T3-4, lymph node positivity, HER-2 positivity, infiltrating non-special pathological type, and G3 were significantly higher in ER(-)/PR(+) than in ER(+)/PR(+) cancer (p <0.001). ER(-)/PR(+) was similar to biological activity of ER(-)/PR(-) type. ER(-)/PR(+)/HER-2(+) patients had a better survival prognosis than ER(-)/PR(+) HER-2(-) patients (p<0.05). The prognosis of ER-/PR+ breast cancer was significantly associated with age, HER-2 status, and T stage. CONCLUSION: ER(-)/PR(+) breast cancer is more similar to ER(-)/PR(-) breast cancer than other breast cancer subtypes, with an early age of onset, a high proportion of infiltrating non-special types, a high histological grade, and a high HER-2 positivity rate. Whether HER-2 positivity can improve the prognosis of ER(-)/PR(+)breast cancer is worth further discussion. The risk scoring system we developed can effectively distinguish between high-risk and low-risk patients. The nomogram we created had a concordance index of 0.736, and the calibration curve showed good agreement between the predicted and observed outcomes.