Impact of SWI/SNF complex mutations in patients with non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors: Immuno-oncology biomarker study in LC-SCRUM-Japan (LC-SCRUM-IBIS).

9530 Background: The SWI/SNF chromatin remodeling complex is reported to be involved in sensitivity and resistance to immune checkpoint inhibitor (ICI). However, their role in non-small cell lung cancer (NSCLC) remains unclear. We examined the relationship between SWI/SNF complex mutations and clinical outcomes of ICI in patients with NSCLC. Methods: Of 1017 lung cancer patients enrolled in LC-SCRUM-IBIS, 350 patients were analyzable for whole-exome sequencing (WES). WES data were used to analyze the presence of mutations in 29 major subunits of the SWI/SNF complexes. ARID1A and SMARCA4 mutations were also evaluated in a targeted NGS panel (Oncomine comprehensive assay, OCA). PD-L1 expression by 22C3, tissue tumor mutational burden (tTMB) by WES, STK11 and KEAP1 mutations by WES or OCA were also assessed. Durable clinical benefit (DCB) including CR, PR and SD > 6 mos to ICI, progression-free survival (PFS) and overall survival (OS) were compared in status of each of SWI/SNF complex mutations and other factors. Results: At least one mutation in any subunits of the SWI/SNF complex was present in 28% of NSCLC patients. The most common mutated subcomplexes were SMARCA4 (12%), BAF (7%: ARID1A, 4%), non-canonical BAF (3%), PBAF (3%), and SMARCA2 (2%). Of 101 NSCLC patients treated with PD-1/PD-L1 inhibitors, SMARCA4 mutations tended to be associated with lower DCB (16 vs 31%) and shorter median PFS (1.9 vs 3.6 m) and OS (7.4 vs 18.1m). Patients with ARID1A mutations tended to have better clinical outcomes (DCB, 40 vs 28%) compared to those without mutations. No significant associations were found between PD-L1 expression and SMARCA4 or ARID1A mutations. Patients with STK11/KEAP1 mutations had lower rate of PD-L1 expression (TPS > 50%) (18% vs 48%, P = 0.03) and worse clinical outcomes (DCB, 6 vs 33%) compared to those without mutations. There was no significant association between a tTMB status and clinical outcome. Conclusions: SMARCA4 and ARID1A mutations appear to affect clinical outcomes of ICI in NSCLC patients. These findings indicate that SWI/SNF complex mutations may serve as a predictive biomarker for ICI in NSCLC patients.