Peripheral blood immune cell and cytokine profiling of patients receiving corticosteroids for immune checkpoint inhibitor-related adverse events: Steroid dose and duration

e14694 Background: Immune checkpoint inhibitors (ICI) have dramatically extended survival in numerous malignancies, but many patients treated with ICIs experience severe immune-related adverse events (irAEs) that can affect virtually any organ. Corticosteroids (steroids) tapered over weeks to months are usually effective but may affect ICI anti-tumor activity. Tapers are often imprecise, resulting in over- or undertreatment of some patients. There are no known blood biomarkers to guide steroid taper management. We therefore profiled peripheral blood immune cells and cytokines at key timepoints to identify biomarkers of response to steroids for irAE treatment. Methods: We collected blood samples before ICI, during irAE, at three steroid dose levels (high, intermediate, and low), and post-steroids from patients treated with ICI therapy at Yale Cancer Center. We measured 48 serum cytokines with a multiplexed immunoassay and immunophenotyped 40 markers on peripheral blood mononuclear cells with mass cytometry (CyTOF). We correlated changes in serum cytokine concentration across timepoints with clinical features. We applied a cluster analysis to the CyTOF data to identify immune cell types and their change in frequency across timepoints. Results: We evaluated 12 patients treated with anti-PD-1 + anti-CTLA-4 therapy who experienced at least 1 severe irAE (11 patients with melanoma, 1 patient with renal cell carcinoma). Five patients had colitis, 6 had hepatitis, and 1 had hepatitis and enteritis. Concentrations of TNFα, IFNγ-mediated cytokines (IL-12p40, IL-18, IL-27), and T cell homing chemokines (CXCL9, CXCL10) increased during irAEs and decreased during steroid therapy without significant change across steroid dose levels (p < 0.05). We observed the largest changes in patients who experienced another irAE within 3 months (n = 5). IL-6 was notably higher in patients who required steroid taper escalation for recurrent symptoms before or soon after completing steroid therapy (n = 8, p = 0.02). Compared to other T cell subtypes, the frequency of CD4 T effector memory (Tem) cells increased during irAE and decreased upon completing steroid therapy but did not return to baseline (p < 0.01). Conclusions: In patients treated with anti-CTLA-4 + anti-PD-1 therapy, IL-6 and IFNγ-mediated cytokines were elevated during irAEs and suppressed at multiple steroid dose levels despite symptom recurrence during the steroid taper. Increased CD4 Tem cell frequency during irAEs may reflect circulating pathogenic T cells, and the delayed decrease and incomplete return to baseline could be due to the severity of irAEs represented in this study. Specific cytokines may predict the occurrence of multiple irAEs and the need for a longer steroid taper. Additional longitudinal analyses of steroid effects on irAEs are needed to validate these findings.