Meta-analysis of immune-related adverse events in phase 3 clinical trials assessing immune checkpoint inhibitors for lung cancer

The introduction in clinical practice of the immune checkpoint inhibitors (ICIs) radically changed the treatment algorithm of lung cancers. To characterize the toxicity of ICIs (atezolizumab, durvalumab, nivolumab, pembrolizumab) is important for personalizing treatment. We performed a systematic review and meta-analysis of phase III randomized controlled trials assessing ICIs, from inception until April 23rd, 2020. We extracted the data from the ICI arm of each trial for indirect comparisons to estimate relative risk for immune-related adverse events (irAEs), severe (grade ≥3) irAEs, drug discontinuation due to irAEs or toxic death. Sixteen trials included a total of 6226 subjects randomized to the experimental immunotherapy arm. Immunotherapy was administered in monotherapy (8 trials), in combination with chemotherapy (6 trials) or other ICI (2 trials). Any grade irAEs and severe irAEs for ICI were 37.1% and 18.5%, respectively. Discontinuations due to any grade irAEs and severe irAEs were 13.8% and 9.2%, respectively; toxic deaths were 2.9% in the immunotherapy arm. Pooled data on any, severe and organ-specific irAEs showed that immunotherapy has a significantly lower risk of irAEs compared to immuno-chemotherapy, especially when analysis was restricted to monoimmunotherapy, like drug discontinuation and toxic death (all p < 0.05). Detailed comparisons between different ICIs provided treatment-related risk profiles for organ-specific irAEs. Our findings contribute to clarifying frequency and features of immune-related toxicities between different ICIs in lung cancer patients, including any grade irAEs, severe irAEs, drug discontinuation and toxic deaths, and may be useful to inform the selection of treatment.