Targeted axillary dissection in the primary surgery setting for node-positive breast cancer: A systematic review

Targeted axillary dissection (TAD) improves the accuracy of axillary staging in patients undergoing neoadjuvant chemotherapy but its role in patients with low volume nodal disease undergoing primary surgery is unclear. This study aimed to summarise the evidence of upfront TAD in the primary surgical setting in patients with node-positive breast cancer.

Searches were performed in PubMed, Embase, and Scopus to identify studies reporting the outcomes of upfront TAD in patients with node-positive breast cancer. Primary outcomes were to report procedural technique, the rate of clip/marker retrieval, concordance with sentinel nodes and the presence of residual nodal disease in the axillary clearance specimen.

Six observational studies (four prospective cohorts and two retrospective) were included involving 66 patients undergoing marking of pathologically involved nodes prior to excision. A single suspicious node was marked in 95.7% (45/47). All localisation markers were detected (13/13), but the tattooed node was not received in 5.3% (1/19), US-visible clips could not be identified in 25% (6/24) and nickle-titanium clips were not retrieved in 23.8% (5/21). Concordance between the marked node and sentinel nodes was 92.1% (35/38) and overall the FNR of SLNB was 0% (0/21). Four studies mandated ALND for all patients whereas two studies considered ALND omission following TAD. Across two studies, 31.6% (6/19) had axillary metastasis outside of the TAD specimen (marked node and SLNs). No complications were reported in the only study reporting complications (0/13). No studies evaluated adjuvant treatment decision or oncological outcomes following upfront TAD.

The current evidence supporting upfront TAD is extremely limited. Consensus work and further observational studies are needed to establish eligibility criteria. Well-designed trials will then be required to determine the oncological safety of this approach.