e20535 Background: ZFHX4 mutation as a novel predictive biomarker for immune checkpoint blockade in non-small cell lung cancer. Methods: 165 NSCLC patients treated with ICIs from independent public cohorts(Rizvi2015; Hellmann2018; Miao 2018) were included to assess immunotherapeutic predictive performance of ZFHX4 mutation. The genomic, mRNA profile and its corresponding survival data were retrieved from The Cancer Genome Atlas (TCGA) database for prognostic analysis. Results: In public cohorts 43 (26.06%) NSCLC patients harbored ZFHX4 mutations. Longer PFS was detected in NSCLC patients with ZFHX4 mutation (ZFHX4-mut) compared with those with wildtype ZFHX4 (ZFHX4-wt) (median PFS 22.1 months vs 4.2 months, HR 0.37; 95% CI 0.22 to 0.61; p < 0.001). Higher ORR and DCR were also observed in patients with ZFHX4-mut compared with ZFHX4-wt (ORR: 62.8% vs 21.3%, p＜0.001; DCR: 86.0% vs 60.7%, p＜0.001, respectively). In the multivariable Cox proportional hazards regression model adjusted by PD- L1 expression and tumor mutational burden (TMB) and other confounding factors, ZFHX4-mut remained an independent predictor (HR 0.43; 95% CI 0.22 to 0.83; p = 0.01). ZFHX4-mut was associated with significantly higher TMB compared to ZFHX4-wt (P＜0.001). According to ZFHX4 status and TMB level, we stratified patients into four subgroups: ZFHX4mut-TMB high, ZFHX4mut-TMB low, ZFHX4wt-TMB high, and ZFHX4wt-TMB low. As expected, ZFHX4mut-TMB high patients achieved the longest PFS 23 months (HR 0.21 95% CI 0.10 to 0.43; p < 0.01) among all groups. The combination of ZFHX4 mutation and TMB-H had a better predictive effect. Conclusions: Our results suggested that ZFHX4 mutation might be a potential predictor for ICIs therapy in NSCLC. Prospective studies to further understand the mechanism are warranted.