Predictive value of immune-related adverse events during pembrolizumab treatment in non-small cell lung cancer

This is the authors' abstract. We don't add key points for this paper.

OBJECTIVES: Several studies have reported the role of immune-related adverse events as a predictor of clinical benefit, but few have properly described these findings in advanced or metastatic non-small cell lung cancer treated with pembrolizumab. This study aimed to evaluate the association between immune-related adverse events development and clinical outcomes in the aforementioned group of patients. METHODS: We conducted a retrospective study in patients with advanced or metastatic non-small cell lung cancer treated with pembrolizumab. Overall response rate, progression-free survival and overall survival were evaluated according to the appearance, subtype and number of immune-related adverse events developed. We report the results of the immune-related adverse events analysis and the potential correlation between immune-related adverse events and clinical outcomes. Univariate and multivariate analyses were performed to evaluate this relationship. Median progression-free survival was statistically significant in favour of patients with at least one immune-related adverse event (p=0.015). Median overall survival was not reached in patients with ≥1 immune-related adverse events, compared with 8 months (95% CI 0.6 to 15.4 months) in those without immune-related adverse events. Patients who developed ≥2 immune-related adverse events had longer median progression-free survival (11 vs 4 months, not statistically significant) and overall survival (not reached vs 11, p=0.022) compared with those with ≤1 immune-related adverse events. CONCLUSIONS: Obtained data showed that patients with immune-related adverse events occurrence had significantly better overall response rate and longer progression-free survival and overall survival. This study highlights the role of immune-related adverse events as a predictor of survival in a real-life setting.