A clinical variable-based nomogram could predict the survival for advanced NSCLC patients receiving second-line atezolizumab

Abstract

Objective: A nomogram model based on clinical variables was conducted to predict the survival in patients with non-small cell lung cancer (NSCLC) receiving second-line atezolizumab. Methods: Four hundred and twenty-four patients with NSCLC receiving atezolizumab from OAK study were regarded as the training cohort. Next, a nomogram model based on clinical variables in the training cohort was established to predict the survival of patients receiving atezolizumab. The concordance index, area under curve (AUC), and calibration plots were used to assess the performance of the nomogram model. In addition, 144 patients with NSCLC receiving atezolizumab from POPLAR study were regarded as the test cohort to validate the nomogram model. Using Kaplan–Meier and log-rank test, we compared the survival difference between the high- and low-risk groups, atezolizumab and docetaxel treatment groups, respectively. Results: We successfully constructed a nomogram model based on different variable scores for predicting the survival in NSCLC patients receiving atezolizumab using the training cohort. According to risk score, patients receiving atezolizumab were divided into the high- and low-risk groups. Importantly, in the training cohort, patients had worse overall survival (OS) in high-risk group compared with the low-risk group (median survival: 252.3 vs. 556.9 days; p < 0.0001). As expected, in the test cohort, the high-risk patients also showed a worse OS (median survival: 288.8 vs. 529.3 days, p = 0.0003). In addition, all the patients from the training and test cohorts could be found the OS benefit from atezolizumab compared with docetaxel (all, p < 0.05). Conclusions: The clinical variable-based nomogram model could predict the survival benefit for NSCLC patients receiving second-line atezolizumab.