Validation of a nomogram predicting nonsentinel lymphnode metastases among patients with breast cancer after neoadjuvant chemotherapy (NACT): A transSENTINA substudy.

Liedtke Cornelia , Görlich Dennis, Bauerfeind Ingo , Fehm Tanja, Fleige Barbara, Helms Gisela, Lebeau Annette, Staebler Annette, Von Minckwitz Gunter, Untch Michael, Kühn Thorsten

Zusammenfassung

Background: The optimal timing for sentinel lymph node biopsy (SLNB) in the setting of NACT is still unclear. Recent studies such as SENTINA suggest that performing SLNB in patients with a cN1 status before but converting to a cN0 status through NACT may result in an unacceptable false-negative rate. There is a need to predict non-SLN status and tailor axillary surgery after NACT. We validated the Jeruss-Nomogramm to predict non-SLN lymph node metastases after primary surgery in patients after NACT. Methods: The SENTINA trial was described in length last year. The Jeruss-nomogram includes lymphovascular invasion after NACT , detection method, multicentricity after NACT, clinical initial nodal involvement, and pathologic tumor size after NACT. It was applied to patients converting froom cN1 to cN0 status through NACT (arm C) but with histologically positive SLN (ypN+). The performance of the nomograms was evaluated by the area under the receiver operator curve (AUC). Results: This subgroup comprised 592 patients with 153 patients having a positive SLN after NACT and had all available data to run the nomogram. Our calculations result in an AUC-value of 0.659 (CI: 0.571 - 0.746, p=0.001) showing a significantly better performance than by random choice (AUC=0.5). For this analysis, we selected a cutpoint of 90% showing reasonable sensitivity (0.81) but low specificity (0.438). A subgroup of 75 patients from SENTINA having complete data for all nomograms shows a better performance of the Jeruss nomogram (0.709, CI: 0.587-0.832) which was still lower compared to e.g. the MSKCC nomogram with the best predictive value among the established nomograms (AUC=0.821, CI: 0.724-0.917). Conclusions: Validation of the Jeruss nomogram in the post-neoadjuvant setting yielded AUC values lower than in the original validation cohort lower than those of nomograms from patients with primary surgery (MSKCC) suggesting predictive inferiority of the prior nomogram compared to others. Our results show that nomograms predicting non-SLN status need to be optimized to avoid full axillary dissection in patients after NACT

Projekte, aus denen die Publikation entstanden ist