Abstract PO3-22-07: Predicting Additional Axillary Metastases in Breast Cancer Patients with Positive Targeted Axillary Dissection Lymph Nodes after Neoadjuvant Treatment

Abstract Background: Neoadjuvant chemotherapy (NACT) is increasingly used for axillary downstaging in clinically node-positive breast cancer patients, and a considerable proportion achieves axillary pathological complete response (ax-pCR). After NACT, axillary staging can be done by targeted axillary dissection (TAD). In case of metastases at TAD, axillary lymph node dissection (ALND) is offered regardless of metastases size. This contrasts primary surgery, where small sentinel node metastases (ypN0(i) and ypN1mi) and ≤2 positive sentinel nodes do not confer ALND, although a proportion of patients with small metastases have additional metastatic lymph nodes (LNs) in the axilla. So far, the residual metastatic burden in the axilla when TAD LNs are positive after NACT is unknown. If subgroups of patients with low residual metastatic burden in the axilla (non-TAD LNs) could be identified, these subgroups may be offered de-escalated axillary treatment. Therefore, we investigated the risk of residual metastatic burden in the axilla when the TAD LNs harbored metastases. Methods: We retrospectively retrieved DBCG data on patients staged by TAD after NACT in Denmark between 1.1.2016-31.8.2021. We registered: age, breast biopsy date, type of surgery, type of axillary surgery, count of LNs, sentinel nodes, and marked lymph nodes with and without metastases, including metastasis size, breast tumor histology and receptor subtype, breast tumor size at diagnosis and in the surgical specimen, malignancy grade and type of neoadjuvant treatment. We excluded patients with inflammatory breast cancer, < 4/>8 cycles of NACT, or a non-standard NACT regimen. The primary outcome was risk factors for having high ( >3), low (1-3), or no residual metastatic burden in the axilla when the TAD LNs harbored metastases. We modeled risk factors for both high and low residual metastatic burden in the axilla using multivariable logistic regression and constructed risk models based on the regression coefficients. Results: We identified 1626 patients receiving NACT and TAD in the inclusion period. After excluding ineligible patients and patients who achieved ax-pCR with no subsequent ALND (46%), the study included 383 patients with positive LNs at TAD for further analysis: thereof 188, 127, and 68 with 0, 1-3 and >3 positive non-TAD LNs, respectively. In the adjusted logistic regression analysis, we found that breast pCR (OR= 0.06, 95% CI < .01-0.41, p < .001) and a low proportion of positive TAD LNs (0-66% vs >66%) (OR=0.32, 95% CI 0.17-0.58, p = < .001) were associated with low risk of high residual metastatic burden in the axilla. Patients with one or both low-risk factors present had an 8% (14 of 176 patients) risk of high residual metastatic burden in the axilla. The predictive value of the model for having < 3 non-TAD LN metastases was 92%. When analyzing the 315 patients with ≤3 positive non-TAD LNs, the adjusted logistic regression analysis of 1-3 vs 0 positive non-TAD LNs showed that ypN0(i) in the TAD LN (OR=0.14, 95% CI 0.04-0.53, p = 0.002), small tumor size at diagnosis (20-49 mm vs ≥ 50 mm) (OR = 0.29, 95% CI 0.14-0.60, p = 0.002), breast pCR (OR= 0.38, 95% CI 0.15-0.98, p = 0.04) and low proportion of positive TAD LNs (33-66% vs >66%) (OR= 0.46, 95% CI 0.27-0.77, p = 0.01) were associated with no residual metastases in the axilla. Using these risk factors, 19% (11/58) of the patients in the lowest risk quartile had further metastatic spread to the axilla. Conclusion: Based on an extensive breast cancer registry, we find that breast pCR, low proportion of positive TAD LNs, small metastases, and small tumor size are associated with low risk of residual metastatic LNs in the axilla when the TAD LNs are positive after NACT. With these risk factors, we propose two models to identify patients with low non-TAD residual metastatic burden and patients with a high likelihood of no further metastases. The models can guide breast surgeons in de-escalating axillary treatment in these groups. Citation Format: Frederikke Munck, Maj-Britt Jensen, Ilse Vejborg, Maria Gerlach, Maja Maraldo, Niels Kroman, Tove Tvedskov. Predicting Additional Axillary Metastases in Breast Cancer Patients with Positive Targeted Axillary Dissection Lymph Nodes after Neoadjuvant Treatment [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO3-22-07.