Abstract
BACKGROUND/AIM: Immune-related adverse events (irAEs) occur in various organs, and sometimes multiply following treatment with immune checkpoint inhibitors (ICIs). This study aimed to determine the association between the number of irAEs and clinical outcomes. PATIENTS AND METHODS: This was a retrospective study that included patients with lung cancer, melanoma, and head and neck cancer who were treated with anti-programmed cell death (ligand) 1 (PD-1/PD-L1) monotherapy. We evaluated the association between the number of irAEs and progression-free survival (PFS) in the simple Cox regression analysis. To eliminate the immortal-time bias, an additional landmark analysis was performed. RESULTS: In total, 92, 69, and 37 patients were allocated to the no, single, and multisystem irAEs groups, respectively. The multisystem irAEs were associated with better PFS compared to the no irAE group. In contrast, at the 12-week landmark, multisystem irAEs were associated with poor PFS compared to the no irAEs group. Furthermore, the rate of treatment suspension owing to irAEs in the multisystem irAEs group (62.5%) was higher than that in the single irAE group (17.3%) at the 12-week landmark. CONCLUSION: The incidence of multisystem irAEs was associated with improved clinical outcomes in patients with lung cancer, melanoma, and head and neck cancer treated with PD-1/PD-L1 inhibitor monotherapy. However, these results may be influenced by a potential immortal-time bias. When accounting for this bias, the early development of multisystem irAEs within 12 weeks was linked to treatment suspension and poorer clinical outcomes.