Abstract
Abstract
Introduction: Detection of sentinel node (SN) metastases is traditionally followed by completion axillary clearance (AC) with attendant increased morbidity and cost. AC often detects no further disease, and the ACOSOG Z-11 trial reported no difference in outcomes with or without AC in patients (pts) with involved SN. The rationale for completion AC is to inform adjuvant therapy decisions or improve locoregional control. The burden of nodal disease is only one factor influencing adjuvant therapy recommendations, but is likely to be a major factor in regional recurrence should AC be omitted. Guidelines recommending AC if the SN contains a macro or micrometastasis are being followed selectively with clinicians using clinical judgement and nomograms to decide management. From 2006, axillary ultrasound (AUS) was introduced as part of our standard workup of pts with screen-detected cancers. Fine needle aspiration cytology (FNAC) was carried out if the AUS was abnormal. Hypothesis We hypothesised that pre-operative (AUS) could help to differentiate those patients where AC will have a therapeutic role from those where minimal residual disease is likely.
Methods: A review of prospective data from a mammographic screening centre in Melbourne, Australia. Imaging and pathology results were reviewed, and MSKCC nomogram scores calculated. AUS was classified as abnormal if: cortical thickness >2.3mm, less than 2.3mm but eccentric, lymph-node enlargement beyond 5mm, or less but larger than adjacent ones.
Results: Five hundred and eleven clinically node-negative pts were diagnosed with invasive breast cancer from February 2006 - October 2009. The mean age was 60 (range 40-86) and 471 (78%) were 50-69 years. Seventy-nine percent were T1 and 80% of ductal type; axillary staging was carried out in 503. All had AUS - 418 (83%) were normal, and 85(17%) were abnormal. Patients with abnormal AUS and positive FNA underwent AC. Overall, 128 of 503 pts (25%) were node positive. If the AUS was abnormal, then 49/85 (58%) had positive nodes, 41/49 (84%) of these with macrometastases. In comparison, if the AUS was normal, then only 79/418(19%) had positive nodes, with 39/79(49%) being macrometastases. The likelihood of nodal metastases was 32% where AUS was abnormal but FNAC non-confirmatory. Of patients with nodal metastasis, if AUS was abnormal, 28% had one node involved, 22% had two, and 50% had more than 2 nodes involved, compared to 54%, 21% and 25% if AUS was normal.
In pts with macrometastasis in the SN and a normal AUS, none were found to have macroscopic disease in non-sentinel axillary nodes if they had also had a MSKCC Nomogram prediction of less than 25%. The same was true for pts with micrometastasis or isolated tumour cells only in the SN, regardless of nomogram prediction.
Of the pts in the cohort with 4 or more involved axillary nodes (n=23), 57% had abnormal AUS and positive FNA, 13% had abnormal AUS but not FNA, and 30% had normal AUS. The MSKCC nomogram score for these latter patients was estimated to be 50-88%
Conclusions: AUS prior to breast biopsy can identify most cases with a large burden of nodal disease. A normal pre-biopsy AUS in the setting of a positive sentinel node may be a useful addition to nomograms to assess the need for completion AC.
Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr PD06-05.