Chaoxing Liu,1– 3,* Jiabin Ding,1– 3,* Jinbiao Xu,1– 3,* Chen Fang,1– 3 GuoHua Zhang,2 Chao Shi,1,3 Feng Qiu1,3 

1Department of Oncology, Gaoxin Branch of the First Affiliated Hospital of Nanchang University, Nanchang, Jiangxi, People’s Republic of China; 2Nanchang Key Laboratory of Tumor Gene Diagnosis and Innovative Treatment Research, Gaoxin Branch of the First Affiliated Hospital of Nanchang University, Nanchang, Jiangxi, People’s Republic of China; 3Department of Oncology, the First Affiliated Hospital of Nanchang University, Nanchang, Jiangxi, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Feng Qiu, Email ndyfy01149@ncu.edu.cn; Chao Shi, Email ndyfy4540@ncu.edu.cn

Background: Hormone receptor-positive breast cancer (HR-positive BC), the most prevalent subtype, typically has a favorable prognosis. However, treatment decision-making and survival prediction remain challenging due to the limitations of traditional staging systems like AJCC. Improved prognostic tools are needed to enhance individualized risk stratification.
Materials and Methods: Clinical information from the Surveillance, Epidemiology, and End Results (SEER) database and the First Affiliated Hospital of Nanchang University were analyzed to evaluate outcomes across HR-positive BC subtypes. Patients were divided into training and validation cohorts. A prognostic nomogram was developed using factors identified by univariate and multivariate Cox regression analyses and evaluated through C-index, Receiver Operating Characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).
Results: The study included 156,378 patients (training) and 67,016 (validation) for breast cancer-specific survival (BCSS) and 165,047 (training) and 70,732 (validation) for overall survival (OS), along with 232 external validation cases. Multivariate Cox regression analysis revealed that the ER-positive/PR-negative (HR=2.317 (2.219– 2.419)) and ER-negative/PR-positive (HR=3.498 (3.143– 3.894)) subtypes had worse prognosis than ER-positive/PR-positive patients. The prognosis of ER-negative/PR-positive subtype (HR=1.511 (1.686– 1.351)) was also worse than that of ER-positive/PR-negative subtype. A nomogram integrating age, race, tumor size, grade, histology, bone, brain, lung, and liver metastases, tumor stage, HER2, marital status, positive lymph node numbers, and radiation therapy. The nomogram had a good C-index values and area under curve values for predicting OS and BCSS in both the training and validation set. Moreover, the DCA revealed that the nomogram performed better than the AJCC (TNM) staging system in predicting the three- and five-year OS and BCSS in both the groups.
Conclusion: This study introduces and validates a novel prognostic nomogram for HR-positive BC, providing enhanced risk stratification, particularly in regions with limited access to comprehensive genetic testing. Further validation through multicenter clinical studies is recommended to confirm its clinical utility.

Keywords: breast cancer, hormone receptor, nomogram and prognosis