Purpose: To study the absolute number of involved nodes/the number of nodes examined or the nodal ratio (NR) in breast cancer. The primary study endpoint was to evaluate the role of supraclavicular and axillary radiotherapy (SART) according to the NR.

Methods and materials: From the Saskatchewan provincial registry of 1981-1995, the charts of 5,996 consecutive patients were retrieved to collect detailed prognostic factors. Among these patients, 1,985 were node positive. Because the NRs are more reliable the greater the number of nodes examined, we analyzed 1,255 patients with > or =10 nodes examined. Of these 1,255 patients, 667, 389, and 199 were categorized into three NR groups--low (< or =25%), medium (>25% to < or =75%), and high (>75%) nodal involvement, respectively.

Results: The NR correlated significantly with the primary tumor size (< or =2 cm, >2 to < or =5 cm, and >5 cm; p = 2.2 x 10(-16)), clinical stage group (p = 5.5 x 10(-16)), pathologic stage group (p < 2.2 x 10(-16)), and the risk of any first recurrence (p = 5.0 x 10(-15)) using chi-square tests. For a low NR, the 10-year overall survival rate with and without SART was 57% and 58% (p = 0.18), and the cause-specific survival rate was 68% and 71% (p = 0.32), respectively. For a medium NR, the 10-year overall survival rate with and without SART was 48% and 34% (p = 0.007), and the cause-specific survival rate was 57% and 43% (p = 0.002), respectively. For a high NR, the 10-year overall survival rate with and without SART was 19% and 10% (p = 0.005), and the cause-specific survival rate was 26% and 14% (p = 0.005), respectively.

Conclusion: This is the first study demonstrating that for patients with > or =10 nodes examined, SART significantly improved the survival for the median and high NR groups but not for the low NR group.