Fig. 2

Possible mechanisms of organ-specific biomarkers of irAEs. a Possible mechanisms of biomarkers associated with gastrointestinal irAEs. The main biomarkers for predicting gastrointestinal irAEs are disordered gut microbiome, CD177 and CEACAM1, peripheral blood mRNA expression (CCL3, CCR3, IL-5, IL-8 and PTGS2), IL-17 and HLA allele. a) One possible mechanism for disordered gut microbiome as a biomarker is that ICIs disrupt the stable microbial system in the gut, resulting in impaired metabolism of beneficial bacteria and a decrease in beneficial bacteria that inhibit inflammation. Another proposed hypothesis is that the microbial-derived products of intestinal disorders could trigger an in situ immune response that might eventually lead to the activation of self-reactive immune cells, causing local intestinal tissue damage. b) CD177 and CEACAM1 are involved in the activation of neutrophils and are thus closely related to immune-mediated intestinal diseases. c) Up-regulated genes such as CCL3, CCR3, IL-5, IL-8 and PTGS2 are involved in inflammatory immune response. d) IL-17, one of the up-regulated central inflammatory cytokines in IBD, was usually inhibited by CTLA-4, but the intervention of ICIs disrupted this ecological balance. e) HLA gene acts an important role in antigen expression and immune tolerance. The presence of common antigens between the tumor and colon tissue causes misleading damage to the gastrointestinal tract by the immune system. b Possible mechanisms of immune-related pneumonia. One possible hypothesis is that CD74 stimulates the release of inflammatory mediators. Additionally, there is a common antigen between the tumor and lung, causing misleading damage to the lung by the immune system. And another new hypothesis is ICIs disrupts the mechanism that inhibits the inflammatory response of Th2 cells. c Possible mechanisms of endocrine irAEs. a) Abnormal TPOAb might predict the thyroid dysfunction because ICIs enhance T cell activity against antigens present in healthy tissues and increase pre-existing autoantibody levels. b) Elevated levels of autoantibodies GNAL and ITM2B are closely related to the immune-related hypophysitis. Possible mechanism is anti-CTLA-4 could generate new reactive effector T cells in the pituitary, while anti-PD-1 / PD-L1 is more likely to make existing pituitary reactive effector T cells more active. Another hypothesis is that the pituitary endocrine cells themselves might express CTLA-4, making hypophysis a direct target for anti-CTLA-4 antibodies and causing hypophysis destruction. d Possible mechanisms of dermatologic toxicity. a) HLA gene mediates a variety of autoimmune diseases and the presence of common antigens between the tumor and skin causes dermatologic misleading damage. b) Th17 cells could induce the transcription of IL-17 cytokines. ICIs enhance the Th17-mediated immune response in susceptible patients, resulting in increased IL-17 levels and ultimately inducing dermatologic irAEs. IL-17, interleukin 17; NK cell, natural killer cell; ICIs, immune checkpoint inhibitors; Th cell, helper T cell; PD-1/L1/L2, programmed cell death protein 1/ligand 1/ligand 2; CTLA-4, cytotoxic T lymphocyte associated antigen-4; HLA, human leukocyte antigen