Background and importance Nivolumab was authorised in Spain in 2015. It is a human immunoglobulin monoclonal antibody that binds to PD-1 receptor and blocks its interaction with PD-L1 and PD-L2. It is indicated in adjuvant or metastatic melanoma (MC), metastatic non-small cell lung cancer (NSCLC), renal cell carcinoma (RC) and squamous cell carcinoma of the head and neck (HNC), among others. Its recent commercialisation means there are no data on adverse events (AE) following long term treatment in routine clinical practice.

Aim and objectives To assess the tolerability of nivolumab, to identify and calculate the prevalence of AE related to nivolumab and to compare its frequency with the that described on the data sheet (DS).

Material and methods A descriptive, retrospective, observational study was carried out from March 2016 to September 2019 in a tertiary hospital. It included all patients treated with nivolumab since it was commercialised. Medical records and blood tests of all treated patients were reviewed from the start of nivolumab treatment. Information was collected from the applications Abucasis, Mizar, Farmis-Oncofarm and Gestlab. Variables collected were sex, age, diagnosis, number of nivolumab doses, AE, if the patient died, nivolumab start/end date and reason for stopping treatment. AE were classified according the prevalence described on the DS: very common (>10%), common (1–10%), not common (0.1–1%), rare (0.01–0.1%) and very rare (<0.01%).

Results A total of 48 patients were included, 77% were men and median age was 63 years. The main diagnoses were NSCLC (40%), followed by RC (29%), MC (21%) and HNC (10%). The median dose received was 7. If the number of doses was calculated according to diagnosis, RC patients received 11 doses, 9 for MC, 4 for LC and 4 for HNC. During the study period, 84.2% of LC patients, 60% of HNC, 20% of MC and 50% of RC died.

Regarding AE, very common (>10%) ones were an increase in lactate dehydrogenase (25%), hypothyroidism (14.6%), eruption (10.4%) and increases in gamma-glutamyl transferase and glutamic-oxaloacetic transaminase (10.4%). The remaining AE were classified as common according its frequency (1–10%): pneumonitis (6.3%), nephritis (4.2%), hepatitis (4.2%), increase in alkaline phosphatase (6.3%), diarrhoea (2.1%), colitis (2.1%), liver failure (2.1%) and arthritis (2.1%). Comparing AE frequency obtained with those reported on the DS, we found that the prevalence of hypothyroidism, colitis, hepatitis, nephritis and arthritis was higher in routine clinical practice than expected.

We found that 77% of patients interrupted nivolumab due to progression of disease (78.4%), AE (16.2%) or ending treatment (5.4%).

Conclusion and relevance Relevant AE that occurred during the study period were hypothyroidism, pneumonitis, hepatitis, nephritis and colitis. Their prevalence was higher than expected and they caused interruption of treatment. The increased prevalence of AE in routine clinical practice highlights the need for strict monitoring of analytical parameters to detect AE as early as possible.

References and/or acknowledgements No conflict of interest.