Nomogram for predicting the benefit of neoadjuvant chemoradiotherapy for esophageal cancer: A SEER-Medicare analysis.

Abstract

4073 Background: The impact of neoadjuvant chemoradiotherapy for esophageal cancer remains difficult to establish for specific patient (pt) populations. The primary aim of this study was to create a web-based prediction tool that provides individualized survival projections based on clinically relevant tumor and treatment data. Methods: Pts diagnosed with esophageal cancer between 1997 and 2005 were selected from the Surveillance, Epidemiology, and End Results (SEER) Medicare database. The covariates chosen for retrospective analysis were: sex, T and N stage, histology, total lymph nodes examined, and receipt of neoadjuvant chemotherapy (CT), radiotherapy (RT), or chemoradiotherapy (CRT). After weighting correction by treatment groups, a log logistic regression model for overall survival (OS) was selected based on goodness of fit analysis. Based on bootstrap resampling with 100 repetitions the Concordance Index (CI) was 0.703. Results: 1,128 resected esophageal pts that either did or did not receive neoadjuvant treatment were appropriate for analysis. On log logistic multivariate analysis: age, sex, T stage, N stage, number of lymph nodes harvested, receipt of neoadjuvant CRT, and receipt of chemotherapy were significantly associated with OS. All T stages greater than 1 benefitted from neoadjuvant CRT (p=<0.001). No T stage benefitted from isolated neoadjuvant CT or RT. Patients with nodal metastases benefitted from neoadjuvant CRT (p=<0.001) and CT (p=0.002). Conclusions: SEER-Medicare pts with resected esophageal cancer can be used to produce a survival prediction tool that can: 1) serve as a counseling and decision aid to pts and caregivers regarding their postoperative prognosis and 2) assist in research protocol design. Patients T2 or greater or with lymph node metastases benefitted from neoadjuvant CRT based on our data. This nomogram may underestimate the benefit of neoadjuvant CRT due to its variable downstaging effect on final pathologic stage. This web based tool is available for use at http://skynet.ohsu.edu/nomograms .