Identification of protein biomarker candidates associated with organ-specific immune-related toxicity and response to management by plasma proteomics.

OBJECTIVE: Immune-related adverse events (irAEs) are a growing challenge with checkpoint inhibitors (CPIs) and are complicated by the lack of suitable response biomarkers in many irAEs. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) based proteomics is a system-wide unbiased analysis method suited for biomarker discovery. In this study, plasma samples from patients suffering from irAEs were analysed using MS-based proteomics. METHODS AND ANALYSIS: The discovery cohort consisted of 37 melanoma patients experiencing an irAE (colitis, hepatitis and nephritis) collected around irAE management. The validation cohort consisted of 34 irAE patients (colitis, hepatitis, nephritis and pneumonitis) at management, with a control group consisting of 34 patients treated with CPIs without an irAE. The plasma samples were processed with bottom-up proteomics and analysed on an Orbitrap Astral Mass Spectrometer with short LC gradients. RESULTS: During an irAE, we found upregulation of multiple acute-phase protein reactants. The level of these was higher in patients in combination CPI therapy-also observed in the control cohort without irAE. We found multiple hepatic proteins associated with immune-related hepatitis, including fructose-biphosphate aldolase B (ALDOB), showing the widest dynamic range. Further, we found a correlation of lipocalin-2/neutrophil gelatinase-associated lipocalin (LCN2/NGAL) to nephritis and colitis with a significant correlation to the clinical toxicity grade confirmed by an immunoassay. Finally, high levels of angiotensinogen (AGT), chitinase-3-like protein 1 (CHI3L1) and lectin mannose-binding 2 (LMAN2) showed a significant association with poor prednisolone response. CONCLUSION: In conclusion, using LC-MS/MS-based plasma proteomics, we identified several irAE-related biomarker candidates to be further assessed for toxicity grading and management in the context of monitoring irAEs.