Abstract
BACKGROUND: Osimertinib has been the standard of care in epidermal growth factor receptor (EGFR)-mutant non-small-cell lung cancer (NSCLC). We evaluated outcomes between osimertinib and first/second-generation (1G/2G) EGFR-tyrosine kinase inhibitors (TKIs) as first-line (1L), and investigated how T790M status and sequential osimertinib after 1G/2G EGFR-TKI failure affected overall survival (OS). MATERIALS AND METHODS: We retrospectively evaluated the outcomes of patients with advanced NSCLC harboring exon 19 deletion or L858R mutation who received osimertinib and 1G/2G EGFR-TKIs as 1L treatment from January 2015 to March 2021. In the exploratory analysis, we analyzed the outcomes among three groups: osimertinib as 1L (1L-Osi), 1L 1G/2G EGFR-TKIs followed by osimertinib (2L-Osi), and 1L 1G/2G EGFR-TKIs without osimertinib (No-Osi). Propensity score matching (PSM) and 12-month landmark analysis were used to mitigate selection bias and immortal time bias. RESULTS: Of 485 patients, 213 and 272 received 1L osimertinib and 1L 1G/2G EGFR-TKIs. All 2L-Osi patients had T790M mutations after 1G/2G EGFR-TKI failure. OS did not differ according to 1L EGFR-TKIs [osimertinib versus 1G/2G EGFR-TKIs; 33.7 versus 41.8 months; hazard ratio (HR) 0.92; 95% confidence interval (CI) 0.65-1.29]. In the 12-month landmark analysis, the median OS was 34.4 months [95% CI 21.3 months-not reached (NR)] in 1L-Osi, 63.8 months (95% CI 46.0 months-NR) in 2L-Osi, and 22.5 months (95% CI 19.0-35.3 months) in No-Osi. After PSM, similar results were observed. CONCLUSIONS: There was no significant difference in OS between osimertinib and 1G/2G EGFR-TKIs as 1L treatment in patients with EGFR-mutant NSCLC. However, 2L osimertinib following 1L 1G/2G EGFR-TKIs in patients who would acquire T790M mutation has been linked to a better prognosis compared to 1L osimertinib.