Corticosteroids for Immune-Related Adverse Events and Checkpoint Inhibitor Efficacy: Analysis of Six Clinical Trials.

Purpose: Retrospective studies suggest that immunosuppressive treatment of immune-related adverse events (irAEs) impairs survival in patients with melanoma who received immune checkpoint inhibitors. Here, we study this association across tumor types using data from six international phase II/III registrational trials.
Methods: A post hoc analysis was performed on individual patient data from the anti-programmed cell death-1 (anti-PD-1) + anti-cytotoxic T lymphocyte-associated protein-4 (anti-CTLA-4) treatment arms of six clinical trials (CheckMate-067, -142, -214, -648, -743, and -9LA). Among patients who received systemic immunosuppression for treatment-related adverse events (trAEs), associations of peak and cumulative corticosteroid dose, and use of second-line immunosuppression with overall survival (OS) and progression-free survival (PFS) were assessed using multilevel Cox regression with adjustment for age and sex.
Results: Of the 1,959 patients who received anti-PD-1 + anti-CTLA-4 therapy, 834 patients who were treated with immunosuppression for trAEs were included. Eight hundred and thirty-two patients (100%) received corticosteroids and 81 patients (10%) received second-line immunosuppressants. High corticosteroid peak dose was associated with worse PFS: adjusted hazard ratio (HR _adj ), 1.15 (95% CI, 1.02 to 1.29) for 1 versus 0.5 mg/kg prednisolone and HR _adj , 1.43 (95% CI, 1.05 to 1.96) for 2 versus 0.5 mg/kg. Similar effects were observed for OS: HR _adj , 1.21 (95% CI, 1.06 to 1.39) and HR _adj , 1.66 (95% CI, 1.17 to 2.37) for 1 and 2 versus 0.5 mg/kg, respectively. Cumulative corticosteroid dose was not associated with survival. HR _adj of use of second-line immunosuppression was 1.23 (95% CI, 0.90 to 1.68) for PFS and 1.25 (95% CI, 0.88 to 1.77) for OS.
Conclusion: Higher corticosteroid peak dose for trAEs is associated with worse survival across tumor types, while cumulative dose is not. Too few patients received second-line immunosuppressants to confirm or reject an association with survival. These data argue for a reconsideration of irAE management approaches, starting with lower corticosteroid dose whenever feasible.