Predictive Biomarkers of Immune Checkpoint Inhibitors-Related Toxicities.

The emergence and continuous development of immune checkpoint inhibitors (ICIs) therapy brings a revolution in cancer therapy history, but the major hurdle associated with their usage is the concomitant ICIs-related toxicities that present a challenge for oncologists. The toxicities may involve non-specific symptoms of multiple systems as for the unique mechanism of formation, which are not easily distinguishable from traditional toxicities. A few of these adverse events are self-limiting and readily manageable, but others may limit treatment, cause interruption and need to be treated with methylprednisolone or tumor necrosis factor-α (TNF-α) antibody infliximab, and even directly threaten life. Early accurate recognition and adequate management are critical to the patient's prognosis and overall survival (OS). Several biomarkers such as the expression of programmed cell death ligand 1 (PD-L1), tumor mutation burden (TMB), and microsatellite instability-high (MSI-H)/mismatch repair-deficient (dMMR) have been proved to be the predictors for anti-tumor efficacy of ICIs, but there is a gap in clinical needs for effective biomarkers that predict toxicities and help filter out the patients who may benefit most from these costly therapies while avoiding major risks of toxicities. Here, we summarize several types of risk factors correlated with ICIs-related toxicities to provide a reference for oncologists to predict the occurrence of ICIs-related toxicities resulting in a timely process in clinical practice.