New CT response criteria in non-small cell lung cancer: proposal and application in EGFR tyrosine kinase inhibitor therapy

Abstract

We aimed to devise new CT response criteria (new response criteria, NRC) in patients with non-small cell lung cancer (NSCLC) and to evaluate the efficacy of the criteria for stratifying patient responses and predicting patient survival compared to that of the traditional size-based criteria RECIST version 1.1. Our institutional review board approved this study with a waiver of informed consent. We enrolled 80 NSCLC patients as an experimental arm and treated them with epidermal growth factor receptor tyrosine kinase inhibitors (EGFR TKIs). Two blinded, independent radiologists assessed CT images for tumor response using the NRC, which were also validated in a separate arm (75 NSCLC patients). Tumor responses evaluated by RECIST 1.1 and the new criteria were compared from each other and correlated with patient survival. For statistical analyses, Kaplan-Meier method and kappa statistics were used. In the experimental arm (n=80), interobserver agreements for the assessment of patient response were excellent for both RECIST and NRC. Sixteen RECIST nonresponding patients achieved a designation of partial response according to NRC. In the validation arm (n=75), patients of optimal response (partial response) with the new criteria had median overall survival of 18.4 months compared with 8.5 months in patients with poor response (P=.04). However, RECIST failed to show survival difference between the two response groups. In NSCLC patients treated with EGFR-TKIs, new CT criteria reflecting additional morphological characteristics of target lesions are reproducible and have statistically significant association with overall survival.