Only 1 in 5 people treated with immune checkpoint inhibitors respond to therapy

ByJosh Friedman

Fact checked byMindy Valcarcel, MS

Add topic to email alerts

Please provide your email address to receive an email when new articles are posted on Cardiology.

Key takeaways:

Roughly 20% of individuals treated with immune checkpoint inhibitors respond.
Response rates began to plateau in 2020.

Approximately 80% of individuals who receive immune checkpoint inhibitors for advanced cancer do not respond to therapy, according to results of a cross-sectional analysis.

The number of FDA-approved immune checkpoint inhibitors (ICIs) has increased steadily since 2018. As of early this year, 11 ICIs had been approved for a combined 88 cancer indications.

However, the proportion of patients who are eligible for these agents — and who derive benefit from treatment — has leveled off over the past 5 years.

“There has been an increase in drug approvals coinciding with a plateau in eligibility and response,” Alyson Haslam, PhD, research scientist at University of California, San Francisco, told Healio. “This indicates that even though there are more approvals, it’s not translating into additional patient benefit.”

Background and methods

Immune checkpoint inhibitors (ICIs) have been a mainstay of cancer treatment since the FDA approved ipilimumab (Yervoy, Bristol Myers Squibb) in 2011.

Haslam and colleagues previously examined trends in ICI eligibility and response, publishing their first paper on the topic in 2019 in JAMA Network Open.

“The initial paper was done because there was a lot of excitement about immune checkpoint inhibitors, which were a new class of drug at the time and they were being successfully used to treat cancer,” Haslam said. “We wanted to see if that excitement was in line with actual benefit for the patient.

“Since then, there have been several checkpoint inhibitor drugs approved, and there have been many other approvals for different indications,” she added. “We realized that our estimates in the 2019 paper were very outdated.”

Haslam and colleagues used the American Cancer Society Facts & Figures report and trial data to conduct an updated analysis.

Estimated rates of ICI eligibility and treatment response among individuals with advanced cancer in the U.S. served as the main outcome.

Results

The percentage of patients with advanced or metastatic cancers eligible for ICIs increased from 1.54% in 2011 to 57.15% in 2020, then declined slightly to 56.55% in 2023.

The percentage of patients who responded to ICI treatment increased from 0.14% in 2011 to 19.49% in 2020. It has remained nearly flat since then, reaching 20.13% in 2023.

The number of approved ICI compounds has increased 83% since 2018; however, during that same period, the eligibility rate increased 27% and the response rate rose 61%.

“There have been a lot of ‘me too’ drug-approval indications,” Haslam said. “Rather than novel drugs that can actually improve upon existing drugs, [pharmaceutical companies] are copying other existing drugs with minor variations.”

More recent ICI approvals have focused on rarer tumors, which has contributed to the eligibility plateau.

“Those tumor types usually have low response rates,” Haslam said. “There have been a number of approvals, but they may not benefit large groups of people with tumor types that are very common.”

Researchers acknowledged study limitations.

For example, they used death data “as a stand-in for eligibility,” noting that ICIs often are administered to patients as later-line therapy when front-line treatments have been unsuccessful.

Investigators also noted they had to make assumptions, utilizing data from peer-reviewed publications, to refine estimates for individual tumor types.

Have ICIs reached their ‘max’?

Eligibility rates have “room for improvement,” but the focus should be on response, Haslam said.

Progress has been made over the last decade. For example, research has shown individuals with BRAF wild-type melanoma achieve worse outcomes than those with BRAF mutations, Haslam said.

However, some ICIs have tumor agnostic indications.

“Oftentimes, responses range anywhere from 0% to 100%, depending on the tumor type,” Haslam said. “In these trials, all patients are treated if they have a certain biomarker, so they all experience the side effects, but some tumor types respond better than others. In that regard, we need to get better.”

More genetic testing is needed to identify which individuals are more likely to respond.

AI could help in this regard.

In a study published in JAMA Oncology, a deep-learning model evaluated histologic images of patients with non-small cell lung cancer treated with ICIs. The model’s prediction scores independently correlated with PFS and OS.

“These findings indicate that an artificial intelligence pathology model could potentially serve as a new tool for guiding ICI treatment, refining patient selection and improving clinical outcomes in the treatment of advanced NSCLC,” Mehrdad Rakaee, PhD, associate professor of genomics and AI in pathology at Oslo University Hospital in Norway, and colleagues wrote.

Haslam agreed AI has potential, but she would prefer to see models evaluated in randomized controlled trials.

“That would be the real test,” she said.

Research is still needed to determine the true impact of ICIs, Haslam added.

She and colleagues plan to conduct another analysis with estimates of how many patients experience adverse events and how many derive a survival benefit.

“There are some successes with [checkpoint inhibitors], but there’s also a lot of times where maybe they’re overused,” Haslam said.

Other studies could evaluate whether giving ICIs earlier in treatment affects response, and the cost effectiveness of the therapies.

“[ICIs] definitely changed the landscape of cancer therapies,” Haslam said. “I would say we’re probably getting close to reaching our max on what the existing drugs can do. I think that these drugs have been tested in so many tumor types that I think we have a pretty good grasp on what they can do.”