Immunotherapy Side Effects May Be More Prevalent in Real World Than in Studies


A retrospective study found that clinical trials may have fewer reported side effects than the real world.

Some side effects from immunotherapies known as checkpoint inhibitors may be more prevalent than researchers realized when they tested the drugs in clinical trials, according to data presented during the 2018 Palliative and Supportive Care in Oncology Symposium, held Nov. 16 and 17 in San Diego.

A new study looked at health insurance records to find thousands of patients with non-small cell lung cancer (NSCLC) treated with PD-1 or PD-L1 inhibitors to determine the side effects they experienced. Its findings suggested that side effects such as pneumonitis (inflammation of the lungs) and hypophysitis (inflammation of the pituitary gland) may be reported more commonly in real-world settings than they were in drug trials.

“Immunotherapy continues to be well tolerated, and severe side effects are less frequent than those seen with conventional chemotherapy. Still, immunotherapy can, in rare occasions, cause other serious medical problems,” said senior study author Elizabeth Jane Cathcart-Rake, M.D., a fellow at the Mayo Clinic in Rochester, Minnesota. “It’s important to understand the full extent of cancer treatments’ side effects, and patients and providers should be aware that it can take a while to fully assess them for newer therapies.”

The retrospective study included nearly 2,800 people with NSCLC who took the immunotherapies Opdivo (nivolumab; 71.4 percent), Keytruda (pembrolizumab; 25 percent) or Tecentriq (atezolizumab; 3.6 percent) between 2015 and 2017, found in a database called OptumLabs Data Warehouse, co-founded by the Mayo Clinic in 2012. Most of the patients had received standard types of chemotherapy before taking immunotherapy.

After identifying the patients who had received any of the three checkpoint inhibitors during the designated time period, investigators looked at how often they had experienced immune-related side effects.

Side effects were more prevalent 60 days after the last immunotherapy treatment than on the day of last treatment. Because the thyroid is sensitive to immune stimuli, the researchers did not find it surprising that, at 60 days, hypothyroidism was the second-most common immune-related side effect, occurring in 7 percent of patients.

But pneumonitis, affecting 10.9 percent of patients at 60 days, was much more prevalent than expected. In the CheckMate-057 trial of Opdivo in non-small cell lung cancer, pneumonitis was reported in 3 percent of patients. It affected 5.8 percent of patients in the Keynote-024 trial of Keytruda in non-small cell lung cancer and 8.3 percent in the Keynote-042 trial of Keytruda in non-small cell lung cancer.

Nephritis, inflammation of the kidneys, was also seen at higher rates than expected. It showed up in in 5.4 percent of patients in the recent study 60 days after final treatment, but in 0.6 and 0.5 percent of patients in the Keynote-024 and Keynote-042 trials, respectively.

Furthermore, in Cathcart-Rake’s study, hypophysitis affected 2.8 percent of patients after 60 days. In Keynote-24 and Keynote-042, it affected 0.6 and 0.5 percent, respectively.

“Claims data complements trial data to allow for a broader view of true adverse event frequencies,” the authors concluded.

Going forward, researchers may also search insurance provider databases to learn when treatment autoimmune side effects typically tend to happen, according to a symposium press release. This could allow health care practitioners to intervene when side effects are most likely to occur.

One challenge in comparing clinical trial findings to what was seen in the clinic is that only about 14 percent of trials report side effects when their results are published, the study investigators pointed out.

If it is found that immune-related side effects are, in fact, more prevalent outside of clinical trials. For example, Cathcart-Rake noted that, in initial clinical trials, aromatase inhibitors for breast cancer were reported to cause joint pain in about 8 percent of patients. But patient-reported outcomes and analyses over the past two decades, she said, show that about 50 percent of patients taking these drugs report joint pain.

Related Videos
Image of a woman with brown shoulder-length hair in front of a gray background that says CURE.
Dr. Nitin Ohri in an interview with CURE
Kim Stuck in an interview with CURE
Dr. Sarah Psutka in an interview with CURE at the ASCO Annual Meeting
Kara Morris in an interview with a gray "CURE" background
Related Content