Indications to immunotherapy in cancer treatment continue to expand, thus there are more and more questions about clinical aspects of using different checkpoint inhibitors. Despite similar mechanism of action between widely used antibodies to PD-1 (nivolumab, pembrolizumab, prolgolimab) and PD-L1 (durvalumab, avelumab, atezolizumab), inhibitors are different due to features of monoclonal antibodies structure they are based on. For instance, toxicity leading to discontinuation of treatment occurs more frequently with anti-PD-L1 drugs than PD-1 inhibitors. On the contrary, the average incidence of any grade IRAEs was higher in patients treated with anti-PD-1 drugs. The revealed differences in the toxicity of the ana-lyzed groups of drugs could be associated with the type of action of the drug. The feature of the PD-L1 inhibitors is more frequent occurrence of antibody-dependent cellular cytotoxicity reactions. However, PD-1 inhibitors showed a statistically significant survival benefit, according to a meta-analysis comparing anti-PD-1 and anti-PD-L1 groups. Besides data on differences in the efficacy and toxicity profiles of these agents, in this article we also analyze different issues in the structure of drug molecules, in particular, the role of LALA mutation in anti-PD-1 inhibitors. Understanding the key distinctive points of check-point inhibitors (CPI) in the future may allow to solve the problem of rechallenge and reintroduction after management of severe IRAEs.