Effect of Very Small Tumor Size on Cancer-Specific Mortality in Node-Positive Breast Cancer

Abstract

Traditionally, larger tumor size and increasing lymph node (LN) involvement have been considered independent predictors of increased breast cancer-specific mortality (BCSM). We sought to characterize the interaction between tumor size and LN involvement in determination of BCSM. In particular, we evaluated whether very small tumor size may predict for increased BCSM relative to larger tumors in patients with extensive LN involvement. Using Surveillance, Epidemiology and End Results registry data, we identified 50,949 female patients diagnosed between 1990 and 2002 with nonmetastatic T1/T2 invasive breast cancer treated with surgery and axillary LN dissection. Primary study variables were tumor size, degree of LN involvement, and their corresponding interaction term. Kaplan-Meier methods, adjusted Cox proportional hazards models with interaction terms, and a linear trend test across nodal categories were performed. Median follow-up was 99 months. In multivariable analysis, there was significant interaction between tumor size and LN involvement (P < .001). Using T1aN0 as reference, T1aN2+ conferred significantly higher BCSM compared with T1bN2+ (hazard ratio [HR], 20.66 v 12.53; P = .02). A similar pattern was seen among estrogen receptor (ER) -negative patients with T1aN2+ compared with T1bN2+ (HR, 24.16 v 12.67; P = .03), but not ER-positive patients (P = .52). The effect of very small tumor size on BCSM was intermediate among N1 cancers, between that of N0 and N2+ cancers. Very small tumors with four positive LNs may predict for higher BCSM compared with larger tumors. In extensive node-positive disease, very small tumor size may be a surrogate for biologically aggressive disease. These results should be validated in future database studies.