103P Time-dependent censoring imbalance and robustness of PFS and OS estimates in trials supporting EMA approval of first-line anti–PD-1/PD-L1 therapy for PD-L1–high advanced NSCLC

In first-line advanced non-small cell lung cancer (NSCLC) with PD-L1 ≥50%, immune checkpoint inhibitors (ICIs) improve progression-free survival (PFS) and overall survival (OS). Standard survival analyses assume non-informative censoring, an assumption that may be violated in open-label immunotherapy trials. We assessed the magnitude, timing, and impact of censoring imbalance on survival estimates in pivotal trials supporting EMA approval.