Radiomics Nomogram for Predicting Locoregional Failure in Locally Advanced Non-small Cell Lung Cancer Treated with Definitive Chemoradiotherapy

Abstract

To develop and validate a radiomics nomogram for predicting locoregional failure (LRF) in patients with locally advanced non-small cell lung cancer (NSCLC) treated with definitive chemoradiotherapy (CRT). 141 patients with locally advanced NSCLC treated with definitive CRT from Jan 2014 to Dec 2017 were included and divided into testing cohort (n = 100) and validation (n = 41) cohort. Radiomic features were extracted from portal venous-phase computed tomography (CT) before treatments using 3D Slicer. The least absolute shrinkage and selection operator (LASSO) logistic regression was processed to select predictive features from the testing cohort and constructed a radiomics signature. Clinical data and the radiomics signature were analyzed using univariable and multivariate Cox regression. The radiomics nomogram was established with the radiomics signature and independent clinical factors. Harrell's C-index, calibration curves and decision curves were used to assess the performance of the radiomics nomogram. The radiomics signature, which consisted of 8 selected features, was an independent factor of LRF. And the clinical predictor of LRF was the histologic type. The radiomics nomogram combined with the radiomics signature and the histologic type showed good performance with C-indexes of 0.828 (95% confidence interval [CI]:0.748-0.908) and 0.765 (95% CI:0.668-0.862) in the testing and validation cohorts respectively. The combined radiomics nomogram resulted in better performance (p<0.001) for the estimation of LRF than the nomograms with the radiomics signature (C-index: 0.776; 95%CI: 0.686-0.866) or the histologic type (C-index: 0.631; 95%CI: 0.536-0.726) alone. Decision curve analysis demonstrated the clinical usefulness of the radiomics nomogram. The radiomics signature is an independent predictor of LRF in patients with locally advanced NSCLC. The radiomics nomogram incorporated the radiomics signature and the histologic type showed good prognostic value of LRF and might be helpful for individual treatments.