Immunotherapy-Related Toxicities May Be More Common Than Originally Reported

Get Permission

Immunotherapy has significantly improved the overall survival of patients with non–small cell lung cancer (NSCLC) and is generally better tolerated than traditional chemotherapies, but the results of a retrospective study suggested that immunotherapy side effects may be more common than initially reported.

According to population-level data presented at the 2018 Palliative and Supportive Care in Oncology Symposium,1 the frequency of some adverse events related to immune checkpoint inhibitors is higher than that reported in the initial trials that led to their approval by the U.S. Food and Drug Administration (FDA). For example, although hypophysitis was noted to occur in 0.6% of patients in the KEYNOTE-024 trial, 2.8% of patients identified in a large commercial insurance claims database experienced this immune-related adverse event 60 days after their last dose of immunotherapy. The authors of the study stressed that real-world data are needed to refine provider and patient expectations for outcomes beyond those observed in clinical trials.

These results mimic what we’re seeing clinically. Immune-related adverse events are occurring much more frequently than we originally thought.
— Elizabeth Jane Cathcart-Rake, MD

Tweet this quote

“The bottom line is that our patients need to understand that these therapies are wonderful medications, with a great chance of response, but they also come with potential side effects, which we’re still getting to know over time as we continue to use them,” said senior study author Elizabeth Jane Cathcart-Rake, MD, a fellow at the Mayo Clinic, Rochester, Minnesota. “It’s important not only for oncologists to be aware of these side effects, but primary care providers, palliative providers, and emergency room providers as well.”

As Dr. Cathcart-Rake explained, although the initial trials of immunotherapy reported adverse events, they were compared with traditional side effects associated with chemotherapy, such as infection risk and alopecia. Those risks are significantly lower with immunotherapy, said Dr. Cathcart-Rake, but there was little mention of immune-related adverse events in the original trials.

For this study, Dr. Cathcart-Rake and colleagues used administrative claims data from a large U.S. commercial insurance database to retrospectively identify patients with NSCLC who received programmed cell death protein 1 or programmed cell death ligand 1 (PD-L1) inhibitors between January 1, 2015, and December 31, 2017. The researchers used ICD-9 or ICD-10 codes to detect immune-related adverse events reported during the period in which the patient was receiving immunotherapy.

Focus on Hypophysitis

As Dr. Cathcart-Rake reported, of the 3,164 patients with -NSCLC identified by researchers, 2,152 (68%) received nivolumab (Opdivo), 886 (28%) received pembrolizumab (Keytruda), and 126 (4%) received atezolizumab (Tecentriq). The majority of these patients (51%) received immunotherapy in the second-line setting, and the average age at treatment initiation was approximately 68 years.

When the date of the last dose was used as a cutoff for analysis, most immune-related adverse events occurred at a higher frequency than initially reported in the clinical trials. However, Dr. Cathcart-Rake noted, this discrepancy became even more pronounced when the search was expanded to 60 days after the last dose of immunotherapy.


  • The rate of immunotherapy-related toxicities may be higher than that reported in the pivotal trials that led to the FDA approvals of immunotherapies, according to a retrospective analysis of commercial insurance claims data.
  • For example, hypophysitis was noted to occur in 0.6% of patients in the KEYNOTE-024 trial, but it was identified in 2.8% of patients in this large cohort study.

“I want to focus on hypophysitis, because I think this is a low-frequency toxicity that can be picked up by claims data a little better than a smaller trial,” said Dr. Cathcart-Rake. “If not caught early, hypophysitis is also a big cause of morbidity and potentially mortality. Even if we cut this off at the date of the last dose, the frequency of that immune-related toxicity was 1.9%.”

Data on immune-related adverse events 60 days after the last dose demonstrated a frequency of 2.8% for hypophysitis, which is more than 4 times as high as the 0.6% of patients identified in the KEYNOTE-024 trial, Dr. Cathcart-Rake stated. The analysis of claims data also showed that the rate of pneumonitis had climbed to 10.9%, up from the 3% with nivolumab initially reported.

“These results mimic what we’re seeing clinically,” said Dr. Cathcart-Rake. “Immune-related adverse events are occurring much more frequently than we originally thought.”

“Our data are different from those of prospective studies,” she emphasized. “However, this study still offers a good sense of what’s happening at a population-based level.”

Limitations of Claims Data Analysis

Not all toxicities were more common than originally reported in clinical trials, however. Hypothyroidism, for example, was seen in just 7% of patients in the retrospective study, said Dr. -Cathcart-Rake, which may speak to a limitation of claims data. “The frequency of hypothyroidism is lower than what we’re seeing clinically and what’s been reported in the trials,” she explained. “Because it’s so quickly and easily treated, I don’t think it’s being coded as a primary diagnosis for some of these patients, so I think we may be missing some of those data.”

According to Dr. Cathcart-Rake, these findings are just preliminary associations. Although a time-to-event analysis still needs to be completed, Hispanic patients were found to have a higher frequency of immune-related adverse events, and patients using PD-L1 agents in the second-line setting and later appeared to experience fewer side effects. Ultimately, the researchers plan to analyze these data in conjunction with prescribing patterns and medications to detect additional immune-related adverse events. Despite the limitations of this analysis, however, Dr. Cathcart-Rake and colleagues believe that claims data can complement trial data to provide a better sense of what patients on immunotherapies may experience.

“Our study is the first to look at adverse events based on claims data, which gives a much broader, population-based perspective on outcomes than a single trial,” said Dr. Cathcart-Rake. “Although there have been studies comparing data from multiple trials, our approach included a comprehensive look at outcomes for most insured patients.”

Does Perception of Immunotherapy Influence Patient Experience?

During the question and answer session, a member of the audience asked whether the “wonderful way in which immunotherapy is described to patients”—ie, teaching the -immune system to eliminate cancer—could influence patients’ experience of the treatment and its side effects. Allison Betof Warner, MD, PhD, a fellow in medical oncology at Memorial Sloan Kettering Cancer Center, New York, noted that her institution is starting a clinical trial about the psychology of patient decision-making with respect to treatments.

Allison Betof Warner, MD, PhD

Allison Betof Warner, MD, PhD

“The way we have treatment conversations with patients has evolved,” said Dr. Warner. “Initially, there was a lot of fear with the early trials, and then there was a lot of hope. Now, as many of us have seen the side effects, there’s a shift back to understanding that some of these toxicities are significant and potentially life-threatening. But I think that is an evolving conversation and one that we don’t have a lot of data on.”

Silvia Formenti, MD

Silvia Formenti, MD

Silvia Formenti, MD, Chair of the Department of Radiation Oncology at Weill Cornell Medicine; Associate Director of the Meyer Cancer Center; and Radiation Oncologist-in-Chief at NewYork-Presbyterian Hospital/Weill Cornell Medicine, reiterated the importance of describing treatment side effects to patients. “As much as we may recognize the toxicities and interact with them early on, some of the side effects are forever, and we really have to intervene,” said Dr. Formenti. “This is not a magic treatment…. When we unleash the immune system, it may attack normal tissue as well and induce autoimmune disease, so it’s a very complicated issue.” 

DISCLOSURE: Drs. Cathcart-Rake and Warner reported no conflicts of interest. Dr. Formenti is a consultant/advisor with Eisai, Bristol-Myers Squibb, AstraZeneca, and Merck; is on the speakers bureau of Varian Medical Systems; has received honoraria from Bristol-Myers Squibb, Varian Medical Systems, Janssen, Regeneron, GlaxoSmithKline, Eisai, Dynavax Technologies, AstraZeneca, and Merck; has received research funding from Janssen, Varian Medical Systems, Regeneron, Eisai, Merck, and Bristol-Myers Squibb; has received travel reimbursement from AstraZeneca and Merck; and has a patent with New York University.


1. Cathcart-Rake EJ, Sangaralingham LR, Shah N, et al: Immunotherapy-related toxicities: More common than originally reported? 2018 Palliative and Supportive Care in Oncology Symposium. Abstract 184. Presented November 17, 2018.