Abstract P2-01-21: Preoperative axillary ultrasound guided needle sampling in breast cancer: Comparing the sensitivity of fine needle aspiration cytology and core needle biopsy

Abstract

Abstract Background: Preoperative axillary ultrasound (US) combined with selective US-guided needle sampling (UNS) can be used to identify lymph node metastases. This can inform decisions about neoadjuvant chemotherapy and allow a patient to proceed immediately to axillary lymph node dissection (ALND) thus avoiding an extra sentinel node biopsy (SNB) procedure. We acknowledge the landmark ACOSOG Z0011 trial showing a subgroup of patients (T1-2) undergoing breast conserving surgery and whole-breast radiotherapy in which ALND can safely be omitted if they have minimal nodal disease burden. For these patients the utility of UNS may be limited if the surgeon has modified their practice according to the trial. For patients not fitting the Z0011 trial criteria, preoperative UNS remains important. Previous studies comparing the sensitivity of axillary US-guided fine needle aspiration cytology (FNA) and core needle biopsy (CNB) have been small and a meta-analysis has not shown a difference in sensitivity1. Our aim was to directly compare the sensitivity of the two techniques. Method: Patients with macrometastatic nodal involvement that were treated at a tertiary referral centre between January 2013 and December 2014 were retrospectively identified from pathology records. Preoperative UNS had been performed by one of eight Consultant Radiologists with the sampling method being according to each individual radiologist's preference. The result of the first UNS performed on each patient was compared to post-operative histopathology results. Patients who had undergone previous axillary surgery or any part of their investigations/treatment at another unit were excluded. Results: A total of 101 CNBs and 181 FNAs were performed in 282 patients. There were 78 true positive CNBs and 96 true positive FNAs. US-guided CNB was therefore more sensitive than US-guided FNA (77.2% vs. 53.0%, p=<0.001). Two non-diagnostic CNBs and eight non-diagnostic FNAs were performed. Five patients in the CNB group were correctly identified preoperatively as having isolated tumour cells (ITCs) or micrometastatic disease only in their axillary lymph nodes and were therefore triaged to SNB rather than ALND. A single haematoma requiring non-operative management was recorded in the CNB group. Conclusion: US-guided CNB of the axilla is more sensitive than US-guided FNA and is a safe technique in experienced hands. We also highlight the additional potential benefit (whilst accepting the possibility of sampling error) of CNB over FNA in assisting the multidisciplinary planning of axillary surgery in patients who are found to have ITCs or micrometastatic disease only during their preoperative axillary staging.