Comparison of marking techniques for target lymph nodes in 2,596 patients with node-positive breast cancer treated with neoadjuvant chemotherapy: Results from the prospective multicenter AXSANA/EUBREAST-03/AGO-B-053 study (NCT04373655).

503 Background: Surgical axillary staging in patients with node-positive (cN+) breast cancer scheduled for neoadjuvant chemotherapy (NACT) varies significantly, and includes axillary lymph node dissection (ALND), sentinel lymph node biopsy (SLNB), targeted axillary dissection (TAD), and target lymph node biopsy (TLNB). SLNB/TAD/TLNB aim at reducing surgical morbidity without loosing staging accuracy. Comparative data on marking techniques for TAD/TLNB are limited. Here, different marking techniques from the largest available international prospective cohort are critically evaluated. Methods: AXSANA is an ongoing cohort study investigating oncological and patient-reported outcomes after different axillary procedures in cN+ breast cancer treated with NACT. In the present analysis, the subgroup of patients receiving marking of their TLN is selected, and detection and removal rates are analyzed. The entire dataset is continuously and systematically monitored for data quality assurance. Results: 6,129 patients from 291 sites in 26 countries were included between June 2020 and January 6 th , 2025. Of these, 2,596 had ≥ 1 TLN marked before NACT and had completed surgery at time of analysis. The mean number of suspicious nodes at diagnosis was 1.9 (≥ 4 in 13.4%). 2,484 patients (95.7%) received a minimally invasive biopsy of ≥ 1 node. TLN marking was performed using a clip in 2,003 patients (77.2%), a magnetic seed in 287 (11.1%), carbon ink in 192 (7.4%), radar marker in 119 (4.6%), radioactive seed in 18 (0.7%), radiofrequency identification device (RFID) in 12 (0.5%) or other methods in 2 (0.1%). > 1 type of marker was placed in 36 patients (1.4%). 1 TLN was marked in 2,427 patients (93.5%), followed by 2 TLNs in 138 (5.3%) and ≥ 3 in 27 patients (1%). The mean number of marked TLNs was highest if carbon ink was used (mean 1.21), followed by clip (1.07), magnetic seed (1.06) and radar marker (1.04); no patient received > 1 radioactive seed/RFID. 1,895 patients (73.0%) achieved ycN0 status. Targeted removal of the TLN was planned in 2,100 patients (80.9%): 2,076 (80.0%) were scheduled for a TAD and 24 (0.9%) for a TLNB. TLN was detected and removed during TAD/TLNB in 1,915 patients (91.2%). TLN detection rate was highest in patients whose TLNs were marked with probe-guided techniques (96.6%; radioactive seed: 100%, magnetic seed: 96.9%, radar marker: 96.1%, RFID: 90%), followed by carbon (94.9%) and clip (89.6%; p < 0.001). TAD/TLNB removed a median number of 3 nodes (mean 4.1, SD 2.77; carbon: median 4, mean 4.29, SD 3.52, probe-guided: median 3, mean 3.82, SD 2.63, clip: median 3, mean 4.15, SD 2.75). Conclusions: This large prospective analysis of patients with initially cN+ breast cancer receiving NACT demonstrates that probe-guided markers provide superior TLN detection rates. Clinical trial information: NCT04373655 .