Background and Importance The leading cause of cancer-related death remains lung cancer. Anti PD-1/PD-L1 antibodies exhibit unique immune-related adverse events (IrAEs). The assessment and comparison of different safety profiles in real clinical practice at our centres are necessary.

Aim and Objectives Evaluation and comparison of the safety in routine clinical practice of anti-PD-1/PD-L1 monoclonal antibodies (nivolumab, pembrolizumab and atezolizumab) used as monotherapy in the treatment of non-small cell-lung cancer (NSCLC).

Material and Methods Retrospective observational study that included patients with NSCLC treated with anti-PD-1/PD-L1 for 7 years in a third-level hospital. Demographic, clinical, treatment, and safety variables were collected. Data were obtained from the electronic medical record. Adverse effect (AE) incidences were calculated and compared between subgroups.

Results 44 patients were included, 18 with pembrolizumab, 17 with atezolizumab and 9 with nivolumab. 84.1% were men with stage IV in 88.6% of the cases. 70.5% had an ECOG Performance status between 0–1. All had negative mutations for targeted therapies and 75% had records of determination of PD-L1 expression, with 61.9% being high expressors (≥50%). The median duration of treatment was 108 (49.5–223.7) days. Regarding the toxicity analysis, 68.2% had a record of some AE, 70.7% grade 1–2 and 38.6% immune related. Regarding the different drugs, pembrolizumab presented more cases of AE in general and a higher incidence of IrAE (44.4%) compared with atezolizumab (29.4%). Due to toxicity, the administration of immunotherapy was delayed in 46.6% of the patients, 26.6% suspended treatment, and 16.7% required hospital admission to manage the toxicity. No statistically significant differences were observed between the different subgroups.

Conclusion and Relevance The incidence of AE in treatment with anti-PD-1/PD-L1 was similar to that available in the literature (68.2%). Approximately 30% were grade 3–4 and we observed a frequency of pneumonitis greater than 15%. The different antibodies present a similar incidence of AE, but atezolizumab seems to have a less immune related safety profile statistically non-significant than the other alternatives. It is essential to increase the sample size and follow-up time to confirm these findings.

Conflict of Interest No conflict of interest.