Axillary ultrasound in patients with clinically node-negative breast cancer: which features are predictive of disease?

2013 
Abstract Background Axillary ultrasound is used in the evaluation of breast cancer patients to identify subclinical node-positive disease. The study aim was to identify whether certain radiologic characteristics correlate with cytology and final pathology. Methods We retrospectively reviewed ultrasound images of 110 women with clinically node-negative breast cancer and suspicious axillary ultrasound to identify specific anatomic characteristics previously shown to be more commonly associated with metastatic involvement. Results were compared with cytology and final pathology. We used descriptive statistics for data summary. Results Of the 110 patients, cytology was positive in 71 (68%) and final pathology was positive in 80 (73%). The most common indication for biopsy was lymph node cortex characterized by thickening or eccentric contour ( N = 40). Loss of the fatty hilum was described in 17 patients, and 9 patients had lymph nodes with both abnormal cortical and hilar features. Of 43 patients with “suspicious” disease without specific criteria, the most common indication for biopsy was disparity in size of one or more lymph nodes compared with others. Maximum cortical thickness was greater in patients with positive cytology compared with those with negative cytology (7.6 versus 6.2 mm; P = 0.047). Ultrasound characteristics such as lymph node size, cortical morphology, contour, and hilar fat were not individually predictive of final cytology and pathology. Conclusions Axillary ultrasound is a valuable tool that accurately predicted malignant axillary disease in 73% of patients with clinically node-negative breast cancer. Elaboration of standard criteria for nodal evaluation will improve usefulness of this imaging modality in preoperative staging of the axilla.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    39
    References
    12
    Citations
    NaN
    KQI
    []