Breast cancer-specific mortality in small-sized tumor with node-positive breast cancer: a nation-wide study in Korean breast cancer society

Tumor size and number of lymph node (LN) metastases are well known as the most important prognostic factors of breast cancer. We hypothesized that very small breast cancers with LN metastasis represent a progressive biologic behavior and evaluated tumor size stratified by LN metastasis. Data between 1990 and 2010 were obtained retrospectively from the Korean Breast Cancer Society Registry with inclusion criteria of female, non-metastatic, unilateral, and T1/2 breast cancer. We collected the following variables: age at surgery, tumor size, number of LN metastases, nuclear grade (NG), lymphovascular invasion (LVI), estrogen receptor status, progesterone receptor status, and epidermal growth factor receptor-2 status. Patient characteristics were compared by means of independent t-tests for continuous variables and the Chi-square or Fisher's exact test for categorical variables. Kaplan-Meier curves, with corresponding results of log-rank tests, were constructed for breast cancer-specific survival (BCSS). Five- and eight-year breast cancer-specific mortality (BCSM) was obtained in groups of 300 patients, followed by smoothing according to the confidence interval using the lowess method. We identified 39,826 breast cancer patients who met the inclusion criteria. Among them, 1433 (3.6 %) patients died due to breast cancer. The median follow-up duration was 63.4 (3-255) months. In the multivariate analysis, age at surgery, NG, LVI, subtype, and tumor size-nodal interactions were independently associated with BCSM. The N1 group had lower BCSS for T1a than T1b. The N2+ group also had lower BCSS for T1b than T1c or T2. In the N1 group of tumors smaller than 10 mm, 5- and 8-year BCSM decreased with larger tumor size. Patients with very small tumors with LN metastasis have decreased BCSM according to increase tumor size. Small tumors with LN metastasis could have aggressive biological behavior.