But the study published in the Journal of Clinical Oncology found startlingly low uptake of those treatments in the first six months after diagnosis: Among nearly 400 Medicare patients diagnosed with the disease between 2007 and 2011, nearly a third never started on the essential drug regimen. Those patients were more likely to be older than 80 or have other medical conditions.
The researchers also found that among patients who started taking the drug within six months of diagnosis, those who received subsidies to defray the cost were quicker to adopt the drug than those who did not have financial help. People who received subsidies started treatment at a median of 58 days after diagnosis, compared with 108 days for people without subsidies.
"I think this is a really good example of a larger problem," said Stacie Dusetzina, an assistant professor of pharmacy and public health at the University of North Carolina at Chapel Hill and an author of the study. "I think people would try very hard to take this medication because of the importance clinicians put on it, but I think there are many drugs now — and many cancer medications in particular ... patients are facing these high upfront costs to initiate treatments."
Under Medicare prescription-drug plans, patients must pay a larger share of medication costs until they have reached $4,850 in out-of-pocket spending. After that, they pay a small copay or coinsurance — which may not be trivial for a drug like Gleevec, which today carries a list price of more than $10,000 a month. A Kaiser Family Foundation study of 2016 data found that the median out-of-pocket cost for Gleevec was $8,503 per year under Medicare prescription-drug plans.
Dusetzina and colleagues have already shown that for people with private insurance, even a relatively modest co-pay can decrease adherence to a lifesaving cancer-drug regimen, so they hypothesized that the burden of cost-sharing under Medicare drug plans could explain why people who do not receive subsidies are slower to start the drug. Although they were slower to take the drug, the people without subsidies eventually caught up and a similar proportion of people in both groups took the drugs by six months after diagnosis.
"Eventually, people find a way — this is a very important drug, so eventually people find a way to pay for it. ... I think that's probably what's happening," said S. Vincent Rajkumar, a professor of medicine at Mayo Clinic. "It probably takes several months to work out a solution, but people will find a way."
The study raises important questions about how cost is affecting patients' use of drugs, but it also has limitations. Researchers could not determine why people began taking a drug or not. The study did not provide information on whether patients ultimately did better or worse depending on when or if they started the drug. Outside researchers also raised questions about how broadly to interpret the findings, because a large number of the roughly 5,000 people diagnosed with the rare leukemia each year should be older than 65, but the study included only 400 patients over a five-year time period. A generic form of Gleevec became available this year, and eventually the use of that drug could decrease costs.
"After a physician prescribes a necessary therapy, any delay in treatment initiation is of concern," Eric Althoff, a spokesman for Novartis, the company that makes Gleevec and another drug in the class, said in a statement.
Although there is not specific evidence that the delays shown in the paper will affect people's health, oncologists said that it is not a type of cancer where physicians typically wait before initiating treatment.
"I’m not aware of any study that proves a delay will worsen outcomes," but the disease can progress to a more acute form of cancer, said Vinay Prasad, a hematologist at Oregon Health and Science University. "All things being equal, if you have a person you’re thinking about starting, you wouldn't want to delay."
Richard Larson, a hematologist and professor of medicine at the University of Chicago, had a number of questions about the paper's methods. But he noted that in a Swedish study, in which the cost of the drug should not be a factor because of differences in how the health-care system works in that country, people older than 80 diagnosed with chronic myeloid leukemia were also less likely to use the drugs. That suggests cost may not be the only factor behind older patients' not taking the drug at all.
But he said that the study adds to the growing body of evidence that the patient portion of the cost is an important factor in whether patients take a drug, even when there is medical consensus that they should.
"If it’s valid, it's just further evidence that out-of-pocket costs do impact on access to these drugs," Larson said.