Examination of PD-1/PD-L1 inhibitor use and adverse events in a real-world setting.

Abstract

e19294 Background: Immune checkpoint blockade through PD-1 and PD-L1 inhibition is an effective treatment for multiple cancers. This study used administrative claims to examine potential adverse events (AEs) associated with real-world PD-1 or PD-L1 inhibitor use. Methods: Adult patients newly initiating a PD-1 (pembrolizumab or nivolumab) or PD-L1 (atezolizumab, avelumab, or durvalumab) inhibitor from September 1, 2016 to August 30, 2018 were selected in the MarketScan Commercial and Medicare Supplemental Database. Patients were grouped by use of PD-1 or PD-L1 inhibitor; the study period consisted of 90 days baseline and 60 days follow-up around drug initiation. Patients who used both PD-1 and PD-L1 inhibitors during follow-up were excluded. Clinical characteristics were examined during baseline, while AEs were investigated over follow-up. Results: A total of 6,430 patients qualified for the analysis. The majority of the sample (N = 5,956; 93%) received PD-1 inhibitors. Compared to the PD-1 cohort, the PD-L1 cohort was older (64±10 vs. 61±12 yrs) and more likely to be male (61% vs. 56%), p < 0.05. PD-L1 patients were significantly more likely to have history of chronic pulmonary disease (28% vs. 23%) or myocardial infarction (4% vs. 3%) but less likely to have liver disease (2% vs. 0.6%) compared to PD-1 patients, p < 0.05. Lung cancer was the most common diagnosis in both groups (PD-1: 47%; PD-L1: 62%, p < 0.001). The PD-L1 cohort was more likely to have evidence of bladder cancer (36% vs. 5%), while the PD-1 cohort was more likely to have a melanoma (19% vs. 0.8%) or renal cell carcinoma diagnosis (10% vs. 7%), p < 0.05. Over half of the PD-1 (65%) and PD-L1 (61%) patients had metastatic cancer diagnosis during the study period. Incident AEs occurring in > 5% of the sample included dyspnea (PD-1: 13%; PD-L1: 14%), nausea/vomiting (PD-1: 11%; PD-L1: 8%, p < 0.05), anemia (PD-1: 11%; PD-L1: 12%), fatigue (PD-1: 10%; PD-L1: 12%), abdominal pain (PD-1: 7%; PD-L1: 7%), cough (PD-1: 7%; PD-L1: 10%, p < 0.05), back pain (PD-1: 7%; PD-L1: 10%, p < 0.05), constipation (PD-1: 6%; PD-L1: 8%), arthralgia (PD-1: 6%; PD-L1: 6%), pyrexia (PD-1: 5%; PD-L1: 8%, p < 0.01), and edema (PD-1: 6%; PD-L1: 5%). Conclusions: This study assessed real-world AEs associated with PD-1/PD-L1 inhibitor use in the 60 days following first treatment. Results showed AEs are common soon after starting therapy. Although, PD-1 and PD-L1 inhibitors target the same pathway, slightly different AE profiles exist for the two classes. More longitudinal analyses of real-world AEs are needed to better understand potential impacts of prolonged therapy.