A decade before he became the US Surgeon General, Vivek Murthy, MD, published a 2004 study in JAMA about participants in cancer clinical trials. Among other findings, Murthy and his coauthors concluded that from 2000 to 2002, people aged 65 years or older were “strikingly underrepresented” in a set of National Cancer Institute–funded drug trials. Fifteen years later, a JAMA Oncology study suggests that not much has changed for older people with cancer. In fact, things appear to be getting worse.
For the new study, researcher Ethan Ludmir, MD, a radiation oncology resident at MD Anderson Cancer Center in Houston, Texas, and colleagues considered the same cancers as Murthy: breast, colorectal, lung, and prostate—the 4 types that strike and kill the most people every year. But they cast a wider net, looking at all 302 phase 3 randomized, multigroup clinical trials for these cancers that initiated enrollment from 1994 to 2015.
The median age of the more than 260 000 trial participants turned out to be an average of 6.5 years younger than that of the broader cancer population. Because cancer surveillance may undercount older patients, the age disparities might be larger, the researchers said. Worse yet, despite having been recognized for decades, the age gap is widening instead of shrinking.
“We’re not making any progress and, if anything, there’s a suggestion that we may be losing some ground on this issue,” said Heidi Klepin, MD, a geriatric oncologist at the Wake Forest School of Medicine in Winston-Salem, North Carolina, who was not involved with the study.
Medication Without Representation
Today, most people who are diagnosed with cancer are aged 65 years or older. With a fast-growing older population, it’s projected that the incidence among seniors will increase 67% between 2010 and 2030.
Age disparities between trial participants and the cancer population are a “huge problem” in oncology, according to Klepin. “For the majority of patients that we see, we are missing key information on the benefits and harms of any given new therapy,” she said.
When new drugs or indications are approved, the evidence gap forces oncologists to extrapolate treatment plans for older patients based on data from younger, healthier cohorts, who may have fewer comorbidities, higher medication tolerances, and lower risks of adverse drug reactions.
In an all-too-common scenario, the recommended drug dosages are too high for older patients, some of whom may end up hospitalized with treatment-related complications. Over time, oncologists learn to adjust drug regimens so that older, frailer patients can tolerate them. “We cause harm and then we adapt,” said Supriya Mohile, MD, who directs the geriatric oncology research program at the University of Rochester Wilmot Cancer Institute in New York.
Compounding the problem, outcomes that are top of mind for older patients and their caregivers, such as functional improvements and quality of life, aren’t always measured, Mohile said. “When you’re talking about 85- and 90-year-olds, longevity may not be as appropriate as the primary end point as it is for a younger person.”
The dearth of older adults in cancer trials has long been recognized. In the late 1980s, a study in the Archives of Internal Medicine (now JAMA Internal Medicine) found that people older than 70 years comprised about a third of patients with cancer in New Mexico but made up just 7% of participants in trials funded by the state’s Southwest Oncology Group.
Several larger analyses, including Murthy’s article and other studies published in 1994, 1999, 2003, and 2004, have gone on to demonstrate similar age mismatches. The disparities increase among the oldest demographics, according to Harpreet Singh, MD, a medical oncologist with the US Food and Drug Administration (FDA). In an analysis she led, people aged 80 years or older composed 16% of patients with cancer but just 4% of participants in registration trials for therapies and indications the FDA approved from 2005 to 2015. Meanwhile, the oldest old may be the most vulnerable to treatment harms.
A small proportion of cancer trials—approximately 2% of phase 3 studies, according to Ludmir’s unpublished research—specifically focus on best practices for older patients. Recently, 1 such trial found that the lowest tested dose of a chemotherapy regimen was just as effective and less toxic than the highest dose in older, frail patients with advanced gastroesophageal cancer. Another will compare the safety and tolerability of recently approved drugs among patients aged 70 years or older with metastatic breast cancer.
Ludmir said that while these studies are important, the other 98% should not exclusively enroll younger patients: “The bulk of our trials, which do not have overt age restriction eligibility criteria, are used to justify standard practice for patients of all ages, and these trials need to enroll appropriate proportions of older cancer patients to ensure that the trial findings are generalizable across the age spectrum.”
Best-Case Scenario Patients
Recognizing the growing need for evidence in older patients, in 2013, the Institute of Medicine (now the National Academy of Medicine) said that cancer research funders should require investigators to have a plan for testing their treatments in a population “that mirrors the age distribution and health risk profile of patients with the disease.” Since then, the FDA, the National Institutes of Health (NIH), the National Cancer Institute (NCI), and the American Society of Clinical Oncology (ASCO) have stepped up initiatives to increase enrollment of older patients.
A major effort to modernize overly restrictive cancer clinical trial eligibility requirements could benefit older adults. Eligibility criteria exist to define the study population and protect trial participants from being harmed. “You want the least variability in your patient population, so you can test in the cleanest way the effect of your intervention,” said Klepin, who is immediate-past chair of the ASCO Cancer Research Committee.
But the strict standards—which Klepin said in many cases are passed down from one study protocol to the next with little scientific basis—are a factor in why only 3% to 5% of adults with cancer participate in clinical trials.
Exclusion criteria other than age limits frequently disqualify seniors from cancer studies. (Only about 9% of trials in Ludmir’s study had upper age limits.) It’s often more subtle age-related factors—such as organ dysfunction or prior malignancies—that eliminate older patients, even when they have little bearing on the treatment that’s being studied.
The result is that “we are enrolling best-case scenario patients disproportionally,” Ludmir said. A 2017 FDA analysis of eligibility criteria in trials for cancer drug applications confirmed this. The parameters limited the studies to lower-risk patients, which the agency said may not reflect the real-world population.
In Ludmir’s analysis, industry-sponsored trials had larger age disparities than studies with other funding sources like the NCI. Whether drug companies intentionally stack the deck with younger, fitter patients is an open question.
“I think it’s far too early to suggest with any definitive data that there’s something overt or insidious about the way industry is supporting trials,” Ludmir said. “Rather, we should explore in more detail why we’re seeing these differences.” To that end, his team is conducting follow-up studies to explore whether industry-supported cancer trials use more restrictive eligibility criteria than do studies with other funding sources.
Toothless Recommendations?
The FDA is now partnering with ASCO and the nonprofit Friends of Cancer Research to relax the eligibility rules used in trials. Among other suggestions, their recommendations include using liberal creatinine clearance criteria when the kidneys aren’t important for a drug’s excretion and including patients with brain metastases and prior or concurrent cancers when possible. The groups are also weighing changes to performance status, a widely used functional assessment associated with the age disparities in Ludmir’s analysis.
If broadly implemented, the new criteria could substantially move the needle. A recent proof-of-principle study projected that these modernized parameters would increase the percentage of trial-eligible patients with non–small cell lung cancer in ASCO’s CancerLinQ database from about 52% to 99%.
Last November, the NCI adopted broadened eligibility criteria based on the groups’ recommendations. When appropriate, the relaxed criteria are recommended for all NCI-sponsored clinical trials and are required for studies overseen by the institute’s Cancer Therapy Evaluation Program (CTEP), the largest publicly funded US clinical trials organization. And early this year, the NIH began requiring that all investigators applying for new or renewed trial funding must have a plan for including participants across the lifespan unless age restrictions can be scientifically or ethically justified. The NIH will also keep tabs on implementation in trials it conducts or sponsors.
And yet these efforts aren’t likely to be enough. The CTEP manages only about a third of NCI-funded interventional trials. More critically, drug companies today sponsor most cancer trials—about 83% of those in Ludmir’s study. (In contrast, just 4% of cancer trials in 1975 to 1984 were industry sponsored, according to 1 study.) Any meaningful changes would therefore require industry collaboration and buy-in.
The FDA works closely with trial funders during the drug application process and has an opportunity to influence study designs. “We make recommendations to sponsors so that their clinical trials are diverse and represent the American population for race, sex, age, and sexual orientation,” said Singh, who is working within the FDA to expand trial access for the geriatric oncology community.
In June, the agency published draft guidance for the broader pharmaceutical industry reflecting the modernized eligibility criteria recommendations. More inclusive trial designs were also suggested, such as the use of so-called expansion cohorts, which could involve lower-dose interventions for older patients. And the agency encouraged the drug industry to make trial participation less burdensome, for example, by offering travel expenses or by replacing some site visits with electronic communication.
But the FDA lacks the teeth to enforce its guidance. Legally granting the agency increased authority to incentivize and require research that includes older adults would help, but no such bills have been introduced in Congress, according to ASCO Chief Medical Officer Richard L. Schilsky, MD. Without such “carrot-and-stick approaches,” Mohile said, “this problem will continue to perpetuate.”
There are other barriers, too. Most older adults get their cancer treatment at community clinics, whereas trials tend to take place at major urban academic medical centers. Experts said a collaborative effort involving stakeholders such as the FDA, the NCI, industry, and community oncologists, to decentralize the clinical trial process is needed to reach more older adults. “We need to take trials to where the patients are,” Mohile said.
She also noted that an age bias underpins the common assumption that seniors with cancer don’t want to join trials. “There [are] data to suggest that when older adults are provided the opportunity, they agree to participate at the same prevalence as younger patients,” she said.
Ludmir said oncologists can do their part by challenging their unconscious biases about which patients are willing and able to take part in studies. This type of thinking could prevent older people from learning about the trials for which they’re actually eligible. “We have a prototype in our mind of somebody who’s fit, younger, and able to do the treatments,” he said. “It’s a bias I think a lot of us have and sometimes are loath to admit.”
