Real-world clinical outcomes in a US Asian population with stage IV NSCLC treated with osimertinib (Osi) stratified by EGFR subtype

e21128 Background: EGFR mutations are observed in 40-50% of Asian populations with NSCLC compared to 10-15% in Caucasian populations, with EGFR exon 19 deletion (ex19del) and exon 21 L858R as the most common activating mutations. The FLAURA trial demonstrated PFS and OS benefit of osi, a 3rd generation tyrosine kinase inhibitor (TKI), compared to earlier generation TKIs with a median PFS (mPFS) of 19 months (mos) and overall survival (OS) of 38.6 mos. Subgroup analysis, however, suggested this survival benefit was not seen in Asian patients (pts) or in pts with the L858R mutation. Further analysis of the 322 Asian pts enrolled at Asian sites demonstrated a mPFS of 16.5 mos (HR 0.54, p < 0.0001) with osi compared to 11.0 mos with SOC EGFR TKI, but no difference on two-year OS (2y OS) (HR 0.65, 95% CI 0.42-1.02). We evaluated the 2y OS in Asian pts treated in the U.S. with stage IV NSCLC with first-line (1L) osi. We further explored the 2y OS of Asian pts who received an earlier generation EGFR TKI prior to osi. Methods: We conducted a single institution IRB-approved retrospective study of Asian pts treated at NYU Perlmutter Cancer Center with stage IV NSCLC with EGFR ex19del or L858R mutations treated with osi from 1/1/2010 – 2/1/2021. 2y OS was compared based on EGFR mutations and receipt of 1L osi vs an earlier generation EGFR TKI prior to osi (TKI-pretreated). Log-rank test was used to calculate the P values of Kaplan-Meier curves subgroups. Cox proportional-hazards model was used to investigate the association between the survival time of pts and multiple predictor variables including age, sex, EGFR subtypes (L858R or ex19del) and smoking status. Results: Of 117 total cases: 85 (73%) were female, 32 (36%) were male. 57 (49%) had ex19del and 60 (51%) had L858R. 94 (80%) were nonsmokers. Age range at diagnosis with stage IV disease was 31-96 years. The analysis revealed no significant difference in 2y OS observed between the 1L osi vs. TKI-pretreated groups (HR 1.90, 95%CI 0.93-3.89). However, the subgroup analysis showed that the 2y OS was inferior in the 1L osi group compared to the TKI-pretreated group in pts with ex19del (HR 2.99, 95%CI 1.12 -7.96), but not in those with L858R (HR 1.29, 95%CI 0.43-3.86). No significant effect on 2y OS was observed from age (p = 0.19), sex (p = 0.50), EGFR subtype (p = 0.19) or smoking status (p = 0.67). Conclusions: In accordance with the FLAURA data, our real-world analysis revealed that Asian pts treated with 1L osi was comparable to pts treated with 1st- or 2nd- generation TKIs. Furthermore, our data showed that the use of 1L osi had a lower 2y OS in Asian pts with ex19del compared to those pre-treated with TKIs, but this was not the case with L858R. Although small numbers, the results suggest that ex19del and L858R should be considered differently when considering 1L treatment options and further investigation into alternative options, such as combination therapies, warrants exploration.