Some immune-related adverse events (irAEs) associated with immune checkpoint inhibitor (ICI) therapies may prove to be higher than the data reported in the trials that led to the drugs’ initial FDA approval, according to study findings reported at the Palliative and Supportive Care in Oncology Symposium that the American Society of Clinical Oncology sponsored in November.
The study assessed administrative claims data from the OptumLabs Data Warehouse, a large US commercial payer database that contains de-identified clinical data from more than 150 million individuals in the United States. The investigators sought to correlate new medical claims submitted by patients with non–small cell lung cancer (NSCLC) who received nivolumab (Opdivo), pembrolizumab (Keytruda), and atezolizumab (Tecentriq) between 2015 and 2017 with the irAE data reported in clinical trials. In the study 68% of patients (n = 3164) received nivolumab versus 28% for pembrolizumab and 4% for atezolizumab. Most were treated with ICIs in first- or second-line therapy (33.1% vs 51.3%).
“We’re seeing higher and higher frequencies reported in these trials as far as irAEs, and that mirrors what we are seeing clinically as well,” said Elizabeth Jane Cathcart-Rake, MD, senior study author and an oncology fellow at Mayo Clinic, Rochester, Minnesota, in an interview with OncologyLive®
In their study, Cathcart-Rake and colleagues looked at the frequencies of irAEs at the date of the last dose of immunotherapy and then 60 days after the last dose for patients in the payer database. They compared these data with rates reported in pivotal clinical trials.
In clinical trials, the rates of pneumonitis were reported at 3% for nivolumab in CheckMate-057, 5.8% for pembrolizumab in KEYNOTE-024, and 8.3% for pembrolizumab in KEYNOTE-042. For hypothyroidism in those trials, the rates were 7% for nivolumab, and 9.1% and 12.1% for pembrolizumab, respectively, in the KEYNOTE trials. For hypophysitis, the rate was not reported for nivolumab in CheckMate-057 versus 0.6% and O.5% for pembrolizumab, respectively, in the KEYNOTE studies.
The analysis of the claims data showed overall rates with ICI therapy of 4.9% for pneumonitis, 5.9% for hypothyroidism, and 1.9% for hypophysitis at the date of the last ICI dose. At 60 days after the last dose, the rates for those irAEs were 10.9% for pneumonitis, 7.0% hypothyroidism, and 2.8% for hypophysitis.
These data show that some irAEs, such as pneumonitis, may be higher in clinical settings than in trials and others, such as hypopysitis may be relatively overlooked in early studies, Cathcart-Rake said. “[Hypophysitis] is a low frequency toxicity that can be picked up by claims data a little better than by a smaller trial. It’s also a big cause of morbidity and potentially mortality if not caught early,” she said.
Based on these findings, it’s important to counsel patients that understanding of ICIs is not yet complete. “Our patients need to know that these are wonderful medications with a great chance of response, but they also come with a chance of potential side effects that we’re still getting to know over time as we use these medications.” Also, clinicians and especially emergency department providers need to know more about irAEs so that they can intervene quickly and effectively and avoid possible misdiagnoses, Cathcart-Rake said.
Figure. Frequency of IRAEs Following ICI Therapy in Clinical Practice
Looking forward, she said that real-world studies may help determine the risk-factors for less prevalent irAEs, such as hypophysitis and myocarditis. Prospective studies of real-world frequency of irAEs also are an important next step for researchers as these allow them to examine large numbers of patients who are not excluded based on pre-existing conditions, Cathcart-Rake said. The findings can better inform patients of the likelihood of irAEs outside of the clinical trial setting.
“Can we tease apart what are the risk factors for getting these irAEs? Are they more likely in one population or another? Are they more likely at a specific time point? I think we are going to be able to do those analyses. It just may take us a little bit of time,” she said.
The mounting evidence indicating that the incidence of irAEs is higher than initially thought mirrors the earlier development of understanding of aromatase inhibitors (AIs) in breast cancer, Cathcart-Rake said. Although initial clinical trial results of AIs in this setting reported joint pain in about 8% of patients, “current findings, based on patient-reported outcomes and more comprehensive analyses over the past 2 decades, show that about 50% of patients taking aromatase inhibitors report joint pain.”
Cathcart-Rake E, Sangaralingham LR, Shah N, et al. Immunotherapy-related toxicities: more common than originally reported? Presented at: Palliative and Supportive Care in Oncology Symposium; November 17, 2018; San Diego, CA. Abstract 184. meetinglibrary.asco.org/record/167317/abstract.