Background: The extent of axillary lymph node involvement represents the foremost important prognostic parameter in primary breast cancer, and, thus, is one of the main determinants for subsequent systemic treatment. Nevertheless, the relevance of the initial axillary lymph node status on survival after disease recurrence is discussed controversially. Persisting prognostic impact after relapse would identify lymph node status as a marker for tumor biology, in contrast to a simply time-dependent phenomenon.
Method: Retrospective analysis of 813 patients with locoregional or distant recurrence of primary breast cancer, who were primarily diagnosed with their disease at the I. Frauenklinik, Ludwig-Maximilians-University, Munich, and the University Hospital in Berlin-Charlottenburg, Germany, between 1963 and 2000. To be eligible, patients were required to have been treated for resectable breast cancer free of distant disease at the time of primary diagnosis, and must have undergone systematic axillary lymph node dissection. Patients with unknown tumor size or nodal status were excluded from the study. All data were gathered contemporaneously and compared with original patients files, as well as the local cancer registry, ensuring high quality of data. The median observation time was 60 (standard deviation 44) months.
Results: At time of primary diagnosis, 273 patients (33.6%) were node-negative, while axillary lymph node metastases were detected in 540 patients (66.4%). In univariate analysis tumor size, axillary lymph node status, histopathological grading, hormone receptor status, as well as peritumoral lymphangiosis and haemangiosis carcinomatosa were significantly correlated with survival after relapse (all, P < 0.0001). Kaplan-Meier analysis estimated the median survival time after relapse in node-negative patients to be 42 months (31-52 months, 95% CI), and 20 months in patients with 1-3 axillary lymph node metastases (16-24 months, 95% CI), compared to 13 months in patients with at least 4 involved axillary nodes (12-15 months, 95% CI). Multivariate logistic regression analysis, allowing for tumor size, axillary lymph node status, histopathological grading, presence of lymphangiosis carcinomatosa, relapse site and disease-free interval confirmed all parameters, except of histopathological grading (P = 0.14), as significant, independent risk factors for cancer associated death. Subgroup analyses, accounting for site of relapse and duration of disease-free interval, confirmed primary lymph node status as independent predictor for cancer-associated death after relapse.
Conclusion: Lymph node involvement at primary diagnosis of breast cancer patients predicts an unfavorable outcome after first recurrence, independently of the site of relapse and disease-free interval. These observations support the hypothesis that primary lymph node involvement is not a merely time-dependent indicator for tumor progression, but indicates tumors with aggressive biological behavior.