The prognostic significance of lymph node micrometastases in breast cancer is controversial. We hypothesized that the survival of patients with solely micrometastatic disease (N1mi) would be intermediate to patients with 1-3 tumor-positive lymph nodes (N1) and those with no positive lymph nodes (N0). We queried the surveillance, epidemiology and end results (SEER) database for all patients between 1992 and 2003 with invasive ductal or lobular breast cancer without distant metastases and < or = 3 axillary nodes with macroscopic disease. Patients were stratified by nodal involvement and compared using the Kaplan-Meier method. Cox proportional hazards regression was utilized to compare survival after adjusting for patient and tumor characteristics. Between 1992 and 2003, N1mi diagnoses increased from 2.3% to 7% among the 209,720 study patients (p < 0.001). In a T-stage stratified univariate analysis, N1mi patients had a worse prognosis in T2 lesions. On multivariate analysis, N1mi remained a significant prognostic indicator across all patients (p < 0.0001) with a hazard ratio of 1.35 compared to N0 disease and 0.82 compared to N1 disease. Other negative prognostic factors included male gender, estrogen-receptor negativity, progesterone-receptor negativity, lobular histology, higher grade, older age, higher T-stage, and diagnosis in an earlier time period. Nodal micrometastasis of breast cancer carries a prognosis intermediate to N0 and N1 disease, even after adjusting for tumor- and patient-related factors. Prospective study is warranted and the results of pending trials are highly anticipated. Until then adjuvant therapy trials should consider using N1mi as a stratification factor when determining nodal status.
