Erschienen in:
08.05.2017 | Breast Oncology
Preoperative Prediction of Node-Negative Disease After Neoadjuvant Chemotherapy in Patients Presenting with Node-Negative or Node-Positive Breast Cancer
verfasst von:
Brittany L. Murphy, MD, MS, Tanya L. Hoskin, MS, Courtney Day N. (Heins), BS, Elizabeth B. Habermann, PhD, MPH, Judy C. Boughey, MD
Erschienen in:
Annals of Surgical Oncology
|
Ausgabe 9/2017
Einloggen, um Zugang zu erhalten
Abstract
Background
Axillary node status after neoadjuvant chemotherapy (NAC) influences the axillary surgical staging procedure as well as recommendations regarding reconstruction and radiation.
Objective
Our aim was to construct a clinical preoperative prediction model to identify the likelihood of patients being node negative after NAC.
Methods
Using the National Cancer Database (NCDB) from January 2010 to December 2012, we identified cT1–T4c, N0–N3 breast cancer patients treated with NAC. The effects of patient and tumor factors on pathologic node status were assessed by multivariable logistic regression separately for clinically node negative (cN0) and clinically node positive (cN+) disease, and two models were constructed. Model performance was validated in a cohort of NAC patients treated at our institution (January 2013–July 2016), and model discrimination was assessed by estimating the area under the curve (AUC).
Results
Of 16,153 NCDB patients, 6659 (41%) were cN0 and 9494 (59%) were cN+. Factors associated with pathologic nodal status and included in the models were patient age, tumor grade, biologic subtype, histology, clinical tumor category, and, in cN+ patients only, clinical nodal category. The validation dataset included 194 cN0 and 180 cN+ patients. The cN0 model demonstrated good discrimination, with an AUC of 0.73 (95% confidence interval [CI] 0.72–0.74) in the NCDB and 0.77 (95% CI 0.68–0.85) in the external validation, while the cN+ patient model AUC was 0.71 (95% CI 0.70–0.72) in the NCDB and 0.74 (95% CI 0.67–0.82) in the external validation.
Conclusions
We constructed two models that showed good discrimination for predicting ypN0 status following NAC in cN0 and cN+ patients. These clinically useful models can guide surgical planning after NAC.