Development and validation of a blood-based genomic mutation signature to predict the clinical outcomes of atezolizumab therapy in NSCLC. (August 2022)

• Record Type: Journal Article
• Authors: Liu, Manjiao; Xia, Sijian; Zhang, Xu; Zhang, Bei; Yan, Linlin; Yang, Meijia; Ren, Yong; Guo, Hao; Zhao, Jie
• Is Part Of: Lung cancer. Volume 170(2022)
• Page Start: 148
• Page End: 155
• Publication Date: 2022-08
• Subjects: Gene mutation -- Atezolizumab -- Non-small cell lung cancer -- Survival -- Immune checkpoint inhibitor; Lungs -- Cancer -- Periodicals; Lung Neoplasms -- Abstracts; Lung Neoplasms -- Periodicals; Poumons -- Cancer -- Périodiques; Lungs -- Cancer; Periodicals; Electronic journals; Electronic journals; 616.99424
• DOI: 10.1016/j.lungcan.2022.06.016
• Languages: English
• ISSNs: 0169-5002

Abstract

Highlights: Developed and validated a blood-based genomic mutation signature (bGMS) to predict OS and PFS benefit for patients with NSCLC receiving atezolizumab therapy. BGMS is superior to blood tumor mutation burden (bTMB), LAF-bTMB, MSAF, and PD-L1 expression in predicting the OS for patients with NSCLC receiving atezolizumab therapy. Low bGMS is associated with better clinical outcomes from atezolizumab therapy than docetaxel. BGMS combined with other non-invasive clinical characteristics improved the predictive power further. Abstract: Objectives: We designed this study to develop a blood-based genomic mutation signature (bGMS) model for predicting the efficacy of atezolizumab therapy in non-small cell lung cancer (NSCLC) in a non-invasive manner. Materials and methods: Patients with NSCLC treated with atezolizumab from POPLAR and OAK clinical trials were included in our study. OAK cohort was defined as the training group, and POPLAR cohort was defined as the validation group. LASSO Cox regressions were applied to the training group to develop the gene mutation signature model to predict the overall survival (OS). Then the model was validated in the validation group. The combined impact of bGMS and other factors was explored with multivariable Cox regression. Results: A bGMS risk model including 15 genes was established to classify patients into high-bGMS and low-bGMS groups. High-bGMS patients had shorter overall survival (OS) and progression-free survival (PFS) compared with low-bGMS in both training cohort (OS 7.9 vs. 19.9 months, p < 0.0001; PFS 1.7 vs. 4 months, p = 0.011) and validation cohort (OS 8.4 vs. 18.6 months, p = 0.0019; PFS 1.5 vs. 4.4 months, p = 0.013). The bGMS was superior to the blood tumor mutation burden (bTMB), LAF-bTMB, MSAF, PD-L1 expression, and a 5-genomic mutation signature in predicting OS for patients receiving atezolizumab. In addition, low-bGMS patients receiving atezolizumab therapy had a better OS rate compared with those receiving docetaxel therapy in both training (P < 0.0001) and validation groups (P = 0.018). Multivariate Cox regression analysis showed that bGMS was an independent prognostic factor on OS and PFS for patients receiving atezolizumab. Furthermore, a nomogram was developed to combine bGMS with the clinical characteristics to improve the predictive power further. Conclusion: bGMS could predict OS benefit for patients with NSCLC receiving atezolizumab therapy. BGMS and other non-invasive clinical characteristics can be combined to develop a more accurate model.