Association between extent of axillary lymph node dissection and patient, tumor, surgeon, and hospital factors in patients with early breast cancer

Background and Objectives: Axillary lymph node dissection (ALND) in patients with breast cancer is crucial for accurate staging, provides excellent regional tumor control, and is included in the standard of care for the surgical treatment of breast cancer. However, the extent of ALND varies, and the extent of dissection and the number of lymph nodes that comprise an optimal axillary dissection are under debate. Despite conflicting evidence, several studies have shown that improved survival is correlated with more lymph nodes removed in both node-negative and node-positive patients. The purpose of this study is to determine which patient, tumor, surgeon, and hospital characteristics are associated with the number of nodes excised in early breast cancer patients. Methods: A random sample of 938 women with node-negative breast cancer was drawn from the Ontario Cancer Registry and the data supplemented with chart reviews. The extent of axillary dissection was studied by examining the number of nodes examined in relation to the patient, tumor, surgeon, and hospital factors. Results: The mean number of lymph nodes excised was 9.8 (SD = 4.8; range, 1-31), and 49% of patients had greater than or equal to 10 nodes excised. Lower patient age was associated with the excision of more lymph nodes ( greater than or equal to10 nodes: 63% of patients <40 years vs. 38% of patients greater than or equal to80 years). Surgeon academic affiliation and surgery in a teaching hospital were highly correlated with each other and were significantly associated with the excision of greater than or equal to10 nodes. The number of nodes excised was not associated with any tumor factors, nor with the breast operation performed. These results were confirmed with multivariable models. Conclusions: Even though the number of lymph nodes found in the pathologic specimen can be influenced by factors other than surgical technique (e.g., number of nodes present, specimen handling, and pathologic examination), this study shows significant variation of this variable and an association with several patient and surgeon/hospital factors. This variation and the association with survival warrant further study and effort at greater consistency. (C) 2003 Wiley-Liss, Inc.