Non-Invasive Predictive Biomarkers for Immune-Related Adverse Events Due to Immune Checkpoint Inhibitors

Simple Summary Immune-related adverse events (irAEs) are one of the most common complications after cancer treatment caused by immune checkpoint inhibitors (ICIs). We performed an extensive review of the potential tests to predict irAEs in patients who receive ICIs. Approximately 40% of patients who receive ICIs experience irAEs. Despite this, only thyroid function tests are currently in mainstream use for predicting who will experience the adverse effects (namely, thyroid function abnormalities) from ICI treatment. A significant amount of research has paved the way for further potential tests that can be used in the clinic, but none have been meaningfully implemented in clinical practice. Nonetheless, further research on identifying these tests and incorporating them into clinical practice to help predict irAEs for patients with cancer will be significant in the field of ICI therapy.Abstract Immune-related adverse events (irAEs) are the most common complication of immune checkpoint inhibitor (ICI) therapy. With the widespread use of ICIs in patients with solid tumors, up to 40% of the patients develop irAEs within five months of treatment, and 11% develop severe irAEs requiring interventions. A predictive test for irAEs would be a crucial tool for monitoring for complications during and after ICI therapy. We performed an extensive review of potential predictive biomarkers for irAEs in patients who received ICI therapy. Currently, only thyroid-stimulating hormone is utilized in common clinical practice. This is due to the unavailability of commercial tests and unclear predictive values from various studies. Given the lack of single strong predictive biomarkers, some novel approaches using composite scores using genomic, transcriptomics, cytokine levels, or clinical parameters appear appealing. Still, these have yet to be validated and incorporated into clinical practice. Further research conducted to validate the models before implementing them into real-world settings will be of the utmost importance for irAE prediction.