LAUSR

9043 Background: Treatment outcome of advanced EGFR exon20 insertion (ex20ins) patients with 1st- or 2nd-line chemotherapy and EGFR TKIs remains limited. Methods: Retrospective real-world analysis of clinical, molecular, and treatment outcome of a web-based patient registry and hospital chart review. Results: Between 9/13/18-10/22/18, 165 EGFR ex20ins NSCLC patients treated in 99 hospitals from 26 different regions in China were registered. Data cutoff was 12/20/18. EGFR ex20ins were determined by commercial NGS (84%) or PCR (16%). 41 different molecular subtypes of EGFR ex20ins were identified. Ala767_Val769dupASV was the most common (23.0%) followed by Ser768_Asp770dupSVD (17.6%). 61% were females, 77% never-smokers, and CNS mets occurred in 22.4% of patients at time of diagnosis. Median PFS was significantly longer in patients who received 1st-line platinum-based chemotherapy (N = 105) (6.4m; 95% CI:5.7-7.1) vs. EGFR TKI (all generations, N = 23) (2.9m; 95%CI:1.5-4.3; P < 0.001) or vs. EGFR TKI (1G, N = 14) (2.6m; 95%CI:0.8-4.4; P < 0.001). Median PFS was numerically longer in patients who received 2nd-line paclitaxel-based chemotherapy (N = 34) (4.1m; 95%CI:3.3-4.9) versus patients who received 2nd-line EGFR TKIs (N = 18) (2.0m; 95%CI:0.8-3.2; P = 0.342). Patients with CNS mets (N = 21) had shorter median PFS with 1st-line chemotherapy than those without CNS mets (N = 84) (5.5m; 95%CI:0.7-10.3 vs. 6.4m; 95%CI:5.8-7.0; P = 0.694). Patients with CNS mets had numerically shorter median PFS with 1st-line EGFR TKI (N = 7) than those without CNS mets (N = 16) (2.0m; 95%CI:0.8-3.2 vs. 2.9m; 95%CI:2.1-3.7; P = 0.077). Conclusions: Chemotherapy is superior to current approved EGFR TKIs as 1st- or 2nd-line treatment of EGFR ex20ins. Patients with CNS mets had numerically poorer PFS with chemotherapy and EGFR TKI treatment. Targeted therapy against EGFR ex20ins with CNS activity is needed.