Abstract:Purpose This study aimed to analyze the association between the primary tumor site and clinicopathological characteristics and survival prognosis of breast cancer (BC) patients using a large population database. Methods BC patients screened in the Surveillance, Epidemiology, and End Results (SEER) database were categorized into 6 groups based on primary sites. Descriptive statistics, Kaplan-Meier curves, Cox regression models, forest plots were used to assess the effect of primary sites on overall survival (OS) and breast cancer-specific survival (BCSS). Multivariate Cox proportional analyses were conducted to calculate hazard ratios (HRs) and adjusted subgroups' hazard ratios (AHRs). Nomograms were utilized to predict OS and BCSS. Results Among 193,043 BC patients, the highest incidence was found in the upper outer quadrant (52.60%). Central portion patients are associated with more clinical features indicating a poor prognosis, and had worse OS and BCSS than other sites. Univariate and multifactorial Cox analyses showed associations between OS/BCSS and various factors. Subgroup analyses revealed differences in OS and BCSS between central portion and upper outer quadrant varied among age, T and N stage. The nomogram was established to predict the survival of central portion BC patients. Conclusions Primary tumor site is associated with clinicopathological features and prognosis of BC, may be influenced by age at diagnosis and T and N stage. Central portion BC patients have worse prognosis due to older age at diagnosis, higher T stage and higher likelihood of lymph node metastasis. Early diagnosis and treatment may help to improve survival of central portion BC.

Analysis of the Relationship Between Primary Tumor Site and Clinicopathological Characteristics and Survival Prognosis of Breast Cancer Patients Based on SEER Database