Management of axillary lymph nodes in breast cancer - A national patterns of care study of 17,151 patients

Objective To determine the rates of axillary lymph node dissection (ALND) and axillar y irradiation (AI) in patients with breast cancer and to identify the facto rs influencing them. Summary Background Data Routine performance of ALND in the treatment of breast canc er has become controversial. AI has been proposed as an alternative to ALND , and it has been suggested that AI in addition to ALND may decrease local failure in high-risk patients. Methods A joint study was conducted by the Commission on Cancer of the American Col lege of Surgeons and the American College of Radiology. A total of 17,151 p atients with stage I and II breast cancer treated at 819 institutions in 19 94 were studied. Results A total of 15,992 patients underwent ALND (93%). The mean ages of patients who did and did not undergo ALND were 60.4 and 73.0 years. Univariate analy sis demonstrated significantly decreased rates of ALND for women age 70 or older (86% vs. 97%), patients with clinical T1 a tumors (81% vs. 93%), grad e I histology (90% vs. 95%), and patients with favorable tumor types (88% v s. 94%). The ALND rate did not vary between palpable and nonpalpable tumors . Multivariate analysis of variables affecting the rate of ALND identified type of surgery, age, tumor size, histology, and payer status as significan t. A total of 889 patients received AI. Patients not undergoing ALND were m ore likely to receive AI (10% vs. 5%). A total of 1.8% of patients with no lymph node metastasis underwent AI, 8.9 % of those with one to three positive nodes underwent AI, 24.0% of those wi th four to nine positive lymph nodes underwent AI, and 29.9% of those with greater than or equal to 10 positive lymph nodes underwent AI. Multivariate analysis of variables affecting the proportion of patients who received AI and had undergone ALND identified nodal status and type of surgery as sign ificant. Conclusions Axillary lymph node dissection continues to be routinely applied in the tre atment of breast cancer, and AI remains underused in patients at high risk for local regional relapse.