Date of Award

January 2015

Document Type

Open Access Thesis

Degree Name

Master of Public Health (MPH)

Department

School of Public Health

First Advisor

Shiyi Wang

Abstract

Purpose The purpose of this study was to combine clinical pathologic variables that are associated with pathological complete response (pCR) following neoadjuvant chemotherapy into a prediction nomogram.

Methods A total of 15,553 women who underwent neoadjuvant chemotherapy for invasive breast cancer in 2010 and 2011 were identified from National Cancer Database (NCDB). Univariate analysis and multivariate logistic regression analysis were used to examine the association of patient age, race, tumor histology, tumor grade, molecular type, and clinical stage with pCR. A nomogram was then developed to predict individual patient probability of pCR to neoadjuvant chemotherapy. Internal validation was performed in terms of discrimination and calibration. The nomogram was then tested against 319 patients from Yale New-Haven Hospital.

Results The predicted probability of the nomogram is between 4% and 74% based on clinical characteristics. In multivariate analysis, high pCR rate is significantly associated with young age, white race, ductal carcinoma, poorly differentiated tumor, Her2 positive and triple negative tumor, small tumor size and less advanced nodal status (p<0.0001). The nomogram had the area under the curve (AUC) of 0.697 in the training set, 0.693 in the internal validation set, and 0.798 in the external validation set. The calibration plot showed good agreement between predicted and actual outcomes.

Conclusions We developed a nomogram that can be used to predict the individual probability of achieving pCR following neoadjuvant chemotherapy in patients with invasive breast cancer, based upon age, race and clinicopathologic characteristics.

Comments

This is an Open Access Thesis.

Share

COinS