SWI/SNF Complex Mutations Serves As The Potential Biomarker Of Recurrence-Free Survival And Efficacy In Non-Small Cell Lung Cancer

Background:The SWItch/Sucrose Nonfermentable complex mutations(SWI/SNF-mut)is the most common chromatin remodeling complex mutation in non-small cell lung cancer(NSCLC).In recent years,the clinical significance of SWI/SNF-mut in NSCLC has attracted extensive attention.Chemotherapy is the cornerstone treatment in advanced NSCLC,but there is still no efficacy predictive markers.Immune checkpoint inhibitors(ICIs)represented by PD-1/PD-L1 inhibitors bring long-term survival benefits for advanced NSCLC,but PD-L1 expression and TMB only have predictive significance for part of patients.There is still an unmet clinical need for prediction markers of efficacy and prognosis in NSCLC.This is a large retrospective study to explore the impact of SWI/SNF-mut on clinical outcomes of chemotherapy and immunotherapy treatment.Methods:We retrospectively collected data of patients who detected by targeted next-generation sequencing(NGS)at Cancer Hospital Chinese Academy of Medical Sciences.Patients with advanced NSCLC who received chemotherapy as first-line treatment were included in chemotherapy cohort and those received PD-1/PD-L1 inhibitors were included in ICIs cohort.Two reported MSKCC cohorts received immunotherapy alone used as the validation for ICIs cohort.We analyzed the relationship between SWI/SNF alterations and clinical outcomes in each cohort and explore the mechanism of SWI/SNF complex mutation status on treatment by analyzing TCGA transcriptome data.Results:In total,1162 patients were included,of which 230 patients(19.8%)were identified as SWI/SNF-mut.SWI/SNF-mut were more common in older age,male sex,smokers,positive PD-L1 expression,high TMB,advanced stage,bone and liver metastasis(p<0.05).In chemotherapy cohort(N=111),patients with SWI/SNF-mut(N=33)had a shorter progression-free survival(PFS)than those with SWI/SNF-wt(4.0 m vs.9.3 m,hazard ratio[HR]=3.05,95%confidence intervals[CI]:1.89-4.98,p<0.001).In ICIs cohort(N=146),there was no significant difference between SWI/SNF-wt and SWI/SNF-mut or different SWI/SNF mutation types in the efficacy of ICIs,while patients with co-mutations of SWI/SNF and KRAS(SWI/SNFmutKRASmut,N=18)had a prolonged PFS to PD(L)-1 inhibitors,especially in the monotherapy treatment subgroup(8.6 m vs.1.9 m,HR=0.31,95%CI:0.11-0.83,p=0.032).Moreover,SWI/SNFmutKRASmut was the independent factor for PFS after ICIs treatment(HR=0.14,95%CI:0.03-0.79,P=0.027).Consistently,no significant differences were observed betweent SWI/SNF-mut and SWI/SNF-wt patients for PFS,and patients with SWI/SNFmutKRASmut presented longer PFS than those with non-SWI/SNFmutKRASmut to ICIs(NR vs.6.3m,HR=0.36,95%CI:0.15-0.82,p=0.016)in the validation cohort.Gene expression ontology and immune cell enrichment analyses revealed that co-mutations of SWI/SNF and KRAS can improve the tumour microenvironment by regulating the cytokines and infiltration of immune cells.Conclusions:SWI/SNF-mut was associated with clinicopathological features in NSCLC tumors.In advanced NSCLC,patients with SWI/SNF-mut had a worse prognosis than those with SWI/SNF-wt after chemotherapy,while patients with SWI/SNFmutKRASmut may benefit more from ICIs.The gene expression and immune cell infiltration of SWI/SNFmutKRASmut tumors showed the characteristics of an activated immune microenvironment.Background:ARID1A is an important transcriptional regulator that plays an important role in cell development,differentiation,and replication and repair for DNA.However,there has been no consensus on the effect of ARID1A-mut on the prognosis of non-small cell lung cancer(NSCLC).In particular,there is few studies about the efficacy of targeted therapy for NSCLC.As a result,we aim to analyze the effect of ARID 1 A-mut on the prognosis of NSCLC and epidermal growth factor receptor tyrosine kinase inhibitors(EGFR-TKIs)to provide a certain value for the prognosis evaluation and precise treatment in NSCLC patients.Methods:The clinical data and whole exome sequencing(WES)data of MSKCC and TRACERx cohorts containing 705 NSCLC cases underwent radical resection which were downloaded from cBioPortal were included in the training cohort.We retrospectively collected data of patients with stage Ⅰ-Ⅲ NSCLC underwent radical resectionat in our hospital were included in the validation cohort,and those harboring EGFR sensitive mutations treated with EGFR-TKIs alone as first-line treatment were included in the targeted cohort.We analyzed the association between ARID1A-mut and clinical outcomes in the above cohorts and used the training cohort to establish a nomogram model to predict recurrence-free survival(RFS).Validation cohort was used for external validation.The model is verified internally and externally by index of concordance(C-index)and calibration chart.Transcriptome data of pan-cancer cohort and NSCLC(lung adenocarcinoma and lung squamous cell carcinoma)from The Cancer Genome Atlas(TCGA)were obtained by cBioPortal to analyze the potential mechanism of relationships between ARID1A-mut and clinical outcomes.Results:Both the training cohort and validati on cohort showed that ARID1 A-mut was associated with younger age,smoke exposure,advanced stage,lymph node metastasis,low differentiation and high SUV value(P<0.05).The RFS of patients with ARID1A-mut was significantly shorter(32.0 m vs.53.0m,hazard ratio[HR]=0.38,95%Confidence intervals[CI]:0.20-0.71],p=0.002)than those with ARID1A wild type(ARID1A-wt),which was consistent either in the group with adjuvant therapy(8.4m vs Not reach[NR],HR=6.14,95%CI:1.37-27.61,p=0.018)or without adjuvant therapy(2.2m vs NR,HR=12.58,95%CI:2.64-59.87,p=0.001),Moreover,multivariate COX regression analysis showed that ARID1A was an independent prognostic factor for RFS in NSCLC patients(HR=2.23,95%CI:1.19-4.20,p=0.013).Based on univariate and multivariate Cox regression analysis,three independent prognostic factors containing ARID1A,pathological type and postoperative pathological stage(p<0.05)were used to construct Nomogram model.The C index in the training cohort and validation cohort were 0.801(95%CI:0.753-0.849)and 0.668(95%CI:0.611-0.725),respectively.The calibration curves for the probabilities of 12-,36-and 48-month RFS revealed satisfactory consistency in internal and external validations.In addition,patients were divided into two groups according to risk stratification(high risk and low risk).There were significant differences in RFS between the two groups in both the trainging and external validation cohorts(p<0.001).In the targeted cohort,patients with ARID1A-mut had a significantly longer PFS than those with ARID1A-wt to EGFR-TKIs(20.6 m vs 11.2 m,HR=0.47,95%CI:0.27-0.94,p=0.023),which was consistent with the subgroups of first/second generation TKIs and third generation TKIs.In multivariate COX regression analysis,ARID1A-mut was also an independent prognostic factor for PFS(HR=0.49,95%CI:0.25-0.98,p=0.047).Transcriptome data analysis of pan-cancer cohort and NSCLC showed that ARID1A-mut tumor were significantly enriched in functions related to cell proliferation,and replication than ARID1A-wt tumors.In addition,ARID1A-mut tumors were also significantly enriched in tumor metabolism-related functions.Conclusions:ARID1A-mut was associated with clinicopathological features and was an independent prognostic factor for RFS in NSCLC.We constructed a nomogram model based on ARID1A mutation status and common clinicopathological factors,which was a reliable and feasible prediction tool for postoperative prognosis in NSCLC.Furthermore,we found ARID1A-mut patients seem to benefit more from EGFR-TkIs treatment.The effect of ARID1A-mut on clinical outcome may be related to tumor proliferation and tumor metabolism functions.