Abstract
Immune checkpoint blockade (ICB) therapy has provided clinical benefits for patients with advanced non-small-cell lung cancer (NSCLC), but the majority still do not respond. Although a few biomarkers of ICB treatment response have been developed, the predictive power of these biomarkers showed substantial variation across datasets. Therefore, predicting response to ICB therapy remains a challenge. Here, we provided a concise combinatorial strategy for predicting ICB therapy response and constructed the ICB treatment signature (ITS) in lung cancer. The prediction performance of ITS has been validated in an independent ICB treatment cohort of NSCLC, where patients with higher ITS score were significantly associated with longer progression-free survival and better response. And ITS score was more powerful than traditional biomarkers, such as TMB and PD-L1, in predicting the ICB treatment response in NSCLC. In addition, ITS scores still had predictive effects in other cancer data sets, showing strong scalability and robustness. Further research showed that a high ITS score represented comprehensive immune activation characteristics including activated immune cell infiltration, increased mutation load, and TCR diversity. In conclusion, our practice suggested that the combination of biomarkers will lead to a better prediction of ICB treatment prognosis, and the ITS score will provide NSCLC patients with better ICB treatment decisions.