Axillary lymph node clearance in patients with positive sentinel lymph node biopsy

Abstract

The use of adjuvant radiotherapy is standard practice following breast conserving surgery and mastectomy in selected patients. Prospective clinical trials are currently being designed to assess the effect of omitting axillary lymph node clearance (ALNC) in selected patients. The aim of this study was to identify the percentage of patients understaged and not considered for postmastectomy radiotherapy (PMRT) and/or supraclavicular fossa radiotherapy (SCFRT) with positive sentinel lymph node (SLN) macrometastasis if the proposed prospective trial inclusion/exclusion protocols are followed. A total of 38 women who were found negative for axillary metastases preoperatively but positive at SLN biopsy and who had ALNC were analysed. PMRT or SCFRT was offered to patients if ≥4 positive lymph nodes (including sentinel nodes) were positive for macrometastasis and/or a tumour size of ≥5cm was detected. Fisher's exact test was used to determine the statistical significance of omitting ALNC. The mean age of the 38 patients was 55 years. A fifth (21.1%) of patients had T1, 76.3% had T2 and 2.6% had T3 disease. The percentage of positive SLNs was 52.6% (1 node), 34.2% (2 nodes) and 13.1% (3 nodes). The number of positive nodes at clearance was 0-3. If the inclusion criteria for trials that consider omitting ALNC are followed (eg POSNOC trial), 23.7% of patients (p=0.0001) with ≥4 positive nodes (including SLNs) would not be offered SCFRT and PMRT. Similarly, if multicentric disease were to be excluded from the trial criteria, the proportion of undertreated patients would reduce by 15.7%. Our study has shown a significant risk of missing patients for PMRT or SCFRT if no ALNC is offered in the presence of SLN macrometastasis. Tumour multicentricity is an important factor in predicting high axillary nodal involvement. Consequently, exclusion of T2 tumours with multicentric involvement in trials considering omitting ALNC may be more appropriate.