Nomogram to Predict the Overall Survival for Patients Underwent Preoperative Radiation Therapy with Chemotherapy

M. Ma,X-S. Gao
DOI: https://doi.org/10.1093/annonc/mdx660.024
IF: 51.769
2017-01-01
Annals of Oncology
Abstract:Background: Currently there is no reliable nomogram to predict the overall survival for patients with esophageal squamous carcinoma who underwent preoperative radiation therapy with chemotherapy. The purpose of this study is to examine SEER database and to develop the above nomogram through the correlation of pretreatment clinical features. Methods: We analyzed esophageal squamous cancer patients between 2010 to 2014 in SEER database. Patients with stage T1-4N0-3M0 and underwent preoperative radiation therapy with chemotherapy and surgery were analyzed. A total of 326 patients were eligible for analysis. All patients had complete clinical data. Age, sex, location of cancer, ethnicity, T stage and N stage were recorded as preoperative predictors. Thesepredictorswere used in multivariable logistic regression analysis based nomograms to estimate the 3-year and 4 year OS, respectively. The predictive accuracy and discriminative ability of the nomogram were determined by concordance index (C-index) and calibration curve. Analyses were performed with open-source statistical software R (version 3.2.1). Results: 41% of patients were male, 72.8% were the white race; 6.4%, 41.9% and 47.4% of the patients had ESCC located in the upper, middle and distal esophagus, respectively; 4% had overlapping lesions. 64% of the patients had T3 disease and 52.3% were in N1 stage. 3 year and 4 year OS of all patients were 53% and 47%. Multivariate analyses demonstrated that age (p = 0.004), sex (p = 0.024), ethnicity (p = 0.05) and N stage (p < 0.001) were independent risk factors of OS for these patients. Nomograms were developed using independent risk factors and also T stage which has clinical significant for the predicted probabilities of 3 year and 4 year OS. The calibration curve for probabilities showed good agreement between prediction by nomogram and actual observation. The C-index of the nomograms for OS was 0.7. The risk of having one of the indications of adjuvant radiation therapy increased with increases in predictors. Conclusions: Using clinic-pathological information, we produced nomograms which may accurately predict the OS for ESCC patients after preoperative radiation therapy with chemotherapy and surgical resection. Legal entity responsible for the study: N/A Funding: None Disclosure: All authors have declared no conflicts of interest.
What problem does this paper attempt to address?