Model To Predict PFS Of Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitors In Chinese Han Lung Adenocarcinoma Population

Purpose: To explore factors that may predict clinical benefit time(defined as progression-free survival(PFS)in this study)both in unselected and selected epidermal growth factor receptor(EGFR)-activating Chinese lung adenocarcinoma patients treated with first-generation EGFR-TKIs.Then to establish a Prognostic Index(PI)model of PFS using these factors to further analyze the implications of the results.Methods: We retrospectively collected clinicopathologic data on 208 patients who received either gefitinib,erlotinib or icotinib,along with the patients' EGFR mutation status and levels of six serum tumor markers(CEA,NSE,CA125,SCC,CYFRA21-1 and LDH).Then,data on 128 of these patients who harbored EGFR sensitive mutations(19 deletions or 21L858R)was independently analyzed.ADx EGFR Mutation Detection Kit(Amoy Diagnostics,Xiamen,China)was used to determine the patients' EGFR mutation status,and serum tumor markers were detected by radioimmunoassay.All patients received 1 of the 3 EGFR-TKIs in 28-day cycles.Gefitinib and erlotinib were administered in dosages of 250 mg and 150 mg once daily,respectively,while icotinib was administered in a dosage of 125 mg 3 times daily.PFS was defined as the duration of using EGFR-TKIs.SPSS 13.0 for windows was used to perform statistical analysis.Kaplan-Meier and Cox regression methods were applied to identify independent factors associated with PFS and to generate a PI model.The study cutoff date was December 1,2015.Results: At the study cutoff date,189(90.9%)of the patients discontinued EGFR-TKI therapy.The median PFS of the 208 patients was 12.4 months(95% CI 11.0-13.8 months).Univariate analysis suggested that gender,smoking history,surgical history,tumor location,T-stage,N-stage,treatment line,EGFR mutation status,PS,and pretreatment CEA levels could affect PFS.A non-smoking history,first-line treatment,and a high pretreatment serum level of CEA were found to be significant predictors of a longer PFS using Cox multivariate analysis precluding EGFR mutation status.Predictive model can be established as: PI = 0.9*Smo+0.405*L+0.354*CEA.The median PFS of the 128 patients who harbored an EGFR sensitive mutation was 14.9 months(95% CI,13.2-16.5 months).Univariate analysis shown that gender,smoking history,surgical history,tumor location,T-stage,treatment line,EGFR mutation status,PS,and pretreatment CA125 levels could affect PFS.A nonsmoking history and first-line EGFR-TKIs treatment were found to be independent predictive factors of a longer PFS with EGFR-TKIs therapy using Cox multivariate analysis.Predictive model can be established as: PI = 1.063*Smo+0.434*L according to the results of Cox regression.Further analysis using the PI model indicated the PFS differences of three groups: nonsmoking and first-line therapy,non-smoking and non-first-line therapy,smoking regardless of treatment timing.Conclusions: In Chinese lung adenocarcinoma patients treated with EGFRTKIs,a non-smoking history,first-line EGFR-TKIs treatment and a high serum level of CEA were independent predictors of a longer PFS along with an EGFR-activating mutation.Nevertheless,when taking EGFR mutation status into account,a non-smoking history and a first-line EGFR-TKIs treatment timing are independent predictors of a longer PFS in EGFR-mutant lung adenocarcinoma patients.PFS is longer for those who are never smokers and receive first-line EGFR-TKIs,compared with other groups.