Impact of PD-L1 on first-line osimertinib outcomes in EGFR-mutant NSCLC: real-world data from the AURORA25 study and meta-analysis.

[INTRODUCTION] The prognostic significance of PD-L1 expression in EGFR-mutant non-small cell lung cancer (NSCLC) treated with first-line osimertinib remains uncertain. This study evaluated its association with survival in a real-world cohort, supported by meta-analysis.

[MATERIALS] Retrospective multicentre study using AURORA cohort (ACTRN12625000038493). Primary endpoint was real-world progression-free survival (rwPFS) comparing PD-L1 high (≥50 %) vs low (0-49 %). Secondary endpoints included overall survival (OS), osimertinib duration and response, and analyses at ≥1 %/≥25 %/75 % cut-points. Kaplan-Meier and Cox models were applied. Meta-analysis of studies to 18April2025 reporting first-line osimertinib outcomes by PD-L1 expression was performed.

[RESULTS] Among 216 patients from 15 centres, median rwPFS was 24.8 months (95 %CI 19.7-29.6). Higher PD-L1 (≥25 %, ≥50 %, ≥75 % versus below each threshold) was associated with shorter rwPFS. The adjusted HR for PD-L1 ≥50 % (n = 35/216, 16 %) vs 0-49 % (n = 181, 84 %) was 3.03 (95 %CI 1.85-4.96, p < 0.001). Median OS was 56.9 months (95 %CI 23.3-57.4): 40.4 with PD-L1 ≥50 and 57.0 with 0-49 %. PD-L1 correlated with shorter OS in unadjusted analyses, with ≥75 % remaining significant after adjustment (HR 2.09, 95 %CI 1.01-4.33, p = 0.046). Meta-analysis of six studies confirmed shorter rwPFS (HR 2.32, 95 %CI 1.16-4.64, p = 0.0178) and OS (HR 2.38, 95 %CI 1.16-4.86, p = 0.0176) for PD-L1 ≥50 % vs 0-49 %.

[CONCLUSIONS] PD-L1 ≥50 % was associated with over twofold risk of progression or death in EGFR-mutant NSCLC on first-line osimertinib. Supported by meta-analysis, results suggest PD-L1 expression is a negative prognostic factor, and these patients may benefit from intensified first-line strategies with prospective evaluation.