The number of involved axillary lymph nodes (LNs) found pathologically is regarded as a significant prognostic factor in early-stage breast cancer (EBC). Recently, there is speculation that LN ratio (LNR) may be a better surrogate at predicting cancer-specific outcome than number of involved LNs. This study investigated prognostic value of LNR, using predetermined cutoff values.Methods
Data included all women diagnosed with node-positive EBC between January 1, 2001, and December 31, 2010 (N = 553). Retrospective evaluation for clinical, demographic, and pathologic data was performed. Most had axillary node clearance (ANC) (548/553; 99.1%). Cohorts were divided by LNR risk groups (low: ≤ 0.20; intermediate: 0.21-0.65; high: >0.65). Proportional hazard modeling was undertaken to evaluate whether LNR was associated with overall survival (OS).Results
Median follow-up was 59.8 months. LNR distribution was as follows: low, 303/553 (54.8%); intermediate, 160/553 (28.9%); high, 90/553 (16.3%). Kaplan-Meier estimates for OS were stratified by LNR: low-risk group had better outcome for OS (P < .001). Overall 5- and 10-year OS was 63% and 58%, respectively. Number of positive LNs correlated with 10-year OS (66%, 48%, and 48% for patients with N1, N2, and N3 stage, respectively; P < .001). LNR also correlated with 5-year OS (69%, 48%, and 41% for low-, intermediate-, and high-risk groups, respectively; P < .001). Significantly, LNR on multivariate analysis also formed a prognostic model when combined with age, estrogen receptor status, PgR status and, HER2 status (P < .001).Conclusion
The Findings support LNR as a predictor for OS in EBC. LNR should be considered an independent prognostic variable to current prognostic instruments already in use.Micro-Abstract
Lymph node ratio (LNR) is considered to have prognostic significance in patients with solid tumors. To better understand its application, we reviewed 553 patients with node-positive early-stage breast cancer (EBC), concentrating on clinical, pathologic, treatment, and outcome data.