Purpose: In early breast cancer the knowledge of the risk of axillary node involvement is important in determining local as well as systemic therapy. Because of the increased acceptance of mammography, there has been an increase in the diagnosis of small, mammographically detected tumors. The objective of this study is to determine whether mammographically detected breast cancers have a lower risk of axillary node involvement compared to those detected clinically.
Patients and methods: From our patient database of stage I and II breast cancer we identified 980 patients with tumors < or = 2 cm whom had axillary node dissection. Four hundred thirty-five (44%) patients presented with abnormal mammograms without clinically palpable tumors; 545 (56%) patients had clinically detected tumors. The median size of the mammographically detected tumors is 1.0 cm, and the median size of the clinically detected tumors is 1.5 cm. The median age of the patients with mammographically detected tumors is 61 (range: 29-87) compared to 53 (range: 27-88) in those with palpable tumors.
Results: Fourteen percent of the patients with mammographically detected tumors had positive axillary nodes compared to 26% of those with clinically detected tumors. Eight percent of patients with mammographically detected tumors had a single positive, while the clinically detected tumors 11% had a single positive node. Thirteen percent of patients with < or = 1 cm tumors and 25% with tumors 1.1 cm to 2 cm had positive axillary nodes. Because the smaller size of the mammographically detected tumors could explain the lower proportion of positive axillary nodes, we analyzed separately the < or = 1 cm tumors. In the group of < or = 1 cm tumors, 9% had positive axillary nodes iftheywere mammographically detected compared to 19% if clinically detected. Four percent had a single positive node while 5% had multiple positive nodes. If the tumors were palpable and < or = 1 cm 9% had a single positive node and 10% had multiple positive nodes. Mammo-graphicallydetected tumors < or = 1 cm had similargrade to clinically detected tumors. In multivariate analysis, method of detection remains a significant variable impacting on the risk of axillary node involvement even in tumors < or = 1 cm.
Discussion: The risk of axillary node involvement is lower in mammographically detected tumors compared to clinically detected tumors independent of tumor size or grade. Mammography detects tumors early in their metastatic progression. The majority of the axillary node-positive patients who are mammographically detected have a single positive axillary node. Method of detection may need to be considered when assessing the risk of axillary node involvement and incorporated in the staging.