Immune checkpoint inhibitors improve survival of patients with advanced melanoma. Combination anti-PD1 and anti-CTLA4 therapy improves the response rate compared to single agent anti-PD1, at the expense of far greater toxicity. While efforts have been made to identify biomarkers to predict treatment response, there are no biomarkers to predict for immunotherapy toxicity.
198 plasma samples from 99 patients (discovery cohort – 50 patients, validation cohort – 49 patients) receiving combination anti-PD1 (nivolumab or pembrolizumab) and anti-CTLA4 (ipilimumab) were studied. The expression of 65 cytokines, measured at baseline (PRE) and early during treatment (EDT: week 1-3), was correlated with immune-related toxicity, defined as toxicity that warranted discontinuation of treatment and administration of high dose steroids (≥ 50mg PO prednisolone equivalent), within 6 months of starting treatment (hypophysitis requiring steroid replacement was excluded).
Toxicity rates were comparable between the discovery and validation cohorts, 21/50 (42%) and 22/49 (45%) respectively. The median expression of all 65 cytokines was significantly higher in patients with immune-related toxicity, compared to patients without toxicity in both the discovery cohort (PRE; p = 0.01, EDT; p < 0.01) and validation cohort (PRE; p = 0.03, EDT; p = 0.02). In addition, the median expression of six pro-inflammatory cytokines associated with autoimmune disease (IFN, IL-2, IL-17A, IL-1α, IL-1β, IL12) was significantly higher at baseline and EDT in patients with toxicity; discovery (PRE and EDT, p < 0.01) and validation (PRE, p < 0.01; EDT, p = 0.01). Expression of FGF-2, IL-12P70, IL13, IL-1 β, IL-2, MIL-1β and VEGF-A was also significantly higher in patients with toxicity at PRE and EDT, in both the discovery and validation cohorts, compared to patients without toxicity.
Elevated levels of circulating cytokines at baseline and EDT predict development of immune-related toxicity. Pre-treatment measurement of circulating inflammatory cytokines may have an important role in treatment selection for metastatic melanoma patients.
Clinical trial identification
Legal entity responsible for the study
National Health and Medical Research Council.
A.M. Menzies: Consultant or advisory role: BMS, MSD, Novartis, Roche, Pierre-Fabre; Honoraria: BMS, MSD, Novartis, Roche. M.S. Carlino: Consultant or advisory role: MSD, BMS, Amgen, Novartis; Honoraria: BMS, Novartis, MSD A. Guminski: Consultant or advisory role: BMS, Roche, Sun Pharma, Merck KgA Travel support: BMS, Sun Pharma. R. Kefford: Consultant or advisory role: BMS, Merck; Honoraria: BMS Conference attendance: BMS, Amgen. G.V. Long: Consultant or advisory role: BMS, Novartis, Roche, Amgen, Pierre Fabre, MSD, Array; Honoraria: BMS, MSD, Roche, Novartis, Incycte. All other authors have declared no conflicts of interest.