Abstract.A high number of uninvolved axillary nodes was found by some authors to be associated with poor survival in node-negative breast cancer. We searched for confirmation and extend the investigation to node-positive cases, using population data from the SEER Program. Patients selected were women aged 40-69 years, diagnosed 1988-1997, T1-T2 breast cancer, undergoing axillary dissection with 4-35 nodes examined. Survivals were estimated by the product-limit method and were computed on pooled data. Results in nodenegative patients (n=37,519) showed a 5-year overall survival from 92% (95% confidence interval: 88-95%) with 4 uninvolved nodes, to 93% (87-98%) with 34 uninvolved nodes. In node-positive patients (n=16,978), the 5-year survival increased from 50% (44-56%) with 0 uninvolved nodes, to 91% (82-100%) with 30 uninvolved nodes. Survival graphs indicated an improvement or a plateau with higher number of uninvolved nodes. The graphs also suggested that the ratio of involved and uninvolved nodes might be correlated with survival. We conclude that there was no evidence of poor outcome associated with a high number of uninvolved nodes. The incidental finding that ratio-based characterization of node involvement might be a prognostic factor will be further investigated.
IntroductionTumor draining lymph nodes can undergo hyperplasia, resulting in increases in the number and size of detectable lymph nodes. A high number of tumor free axillary lymph nodes might thus represent the effect of lymphangiogenic and lymph node development factors playing a role in tumor metastasis (1). In support of that hypothesis, Camp et al reported that axillary dissection yielding more than 20 nodes in node-negative breast cancer was associated with a poor survival (1) when compared with node-negative breast cancer patients having less than 20 nodes examined.The Surveillance, Epidemiology, and End Results (SEER) provides detailed data on nodal extent for cases diagnosed since 1988 (2). The objective of the present study is to examine the pattern of survival, whether or not there is supportive evidence in the SEER registries data that high numbers of uninvolved nodes (nneg) are indicators of poor outcome.If lymphogenic factors are involved in node-negative (N0) breast cancer, then it should be expected that these factors are also involved in node-positive (N + ) cases. Therefore, consistency requires that the role of nneg should be examined in both N0 and N + cases, taking into account the number of positive nodes (npos).Outline of this study and results of a multivariate analysis by proportional hazard models have been reported previously (3). The mortality hazard ratio associated with the number of uninvolved nodes was 0.990 (95% confidence interval: 0.983-0.996) in N0, and 0.970 (0.963-0.976) in N + , i.e. 1% risk reduction and 3% risk reduction per uninvolved node, respectively. These results argued against the role of uninvolved nodes as indicators of poor prognosis. The present report will focus on the visual display of data. W...