P02.05 A Nomogram to Predict Survival in Non-Small Cell Lung Cancer Patients Receiving Atezolizumab: An Analysis of OAK and POPLAR Cohorts

Introduction: In this study, we would like to build a novel nomogram model which could assist clinicians to select right candidates who would benefit from atezolizumab treatment. Methods: A total of 424 NSCLC patients treated with atezolizumab from the OAK trial were enrolled in the training cohort. In addition, another 144 NSCLC patients receiving atezolizumab from the POPLAR trial were included in the test cohort. Univariate and multivariate Cox regression analyses were applied to analyze the impact of clinical parameters on OS. The ROC curve was applied to assess clinical utility of the nomogram. The calibration curve and C-index was applied to perform the validation of the nomogram. Kaplan-Meier method was used to draw survival curves and compared differences by log-rank test. Results: Our results showed that race, sex, histopathology, ECOG PS, BLSLD and number of metastasis site were independent prognostic factors for NSCLC patients receiving atezolizumab. Basing on these clinical variables, we built a predictive nomogram model. Then, we classified patients into high- and low-risk group according to the risk classification system generated by the nomogram. Survival curves of the training cohort revealed that patients received atezolizumab in the high-risk group had poorer OS than those in the low-risk group (P < 0.0001). Similarly, consistent result was observed in the test cohort, showing that patients receiving atezolizumab with high-risk had relative worse OS than those with low-risk (P = 0.003). More importantly, we found that no matter in OAK trial or POPLAR trial, either in high-risk or low-risk group, NSCLC patients treated with atezolizumab had better OS compared with those receiving docetaxel (all, P < 0.05). Conclusion: Basing on inexpensive and easily available clinicopathological parameters, we built a nomogram which revealing a good performance to predict individual OS probability among patients with NSCLC receiving atezolizumab. Keywords: nomogram, NSCLC, atezolizumab