Many cancer drugs that win U.S. Food and Drug Administration (FDA) approval based on clinical trials with surrogate endpoints may not necessarily help patients live longer, a new study suggests.
Researchers conducted a retrospective review of cancer drugs approved from January 1992 through July 2019 on the basis of surrogate endpoints - either response rate or progression-free survival. They identified a total of 194 unique drug authorizations for 132 different medications, including 89 accelerated approvals and 105 therapies approved on a standard timeline.
Surrogate endpoints were used for first-time approvals of 64 drugs. In 39 cases (61%) researchers found no documented correlation between the surrogate endpoints and overall survival. For another 10 cases (16%) there was a poor correlation; one case had a medium correlation and just three had a high correlation.
"Surrogate endpoints are more convenient to use, and sometimes are more likely to yield a positive outcome," said lead study author Dr. Emerson Chen of the Knight Cancer Institute at Oregon Health & Science University in Portland.
"But because they do not measure quantity of life (overall survival) or quality of life directly, there is always a margin of error," Dr. Chen said by email.
During the study period, surrogate endpoints were used for roughly one in three first-time approvals for specific types of cancer, the authors report in JAMA Internal Medicine.
There were few approvals using surrogate endpoints in the 1990s, but this became increasingly common over time, the analysis found. From 2016 to 2019, a total of 34 accelerated approvals of cancer drugs involved surrogate endpoints, as did 36 drugs approved using the standard review process.
"More and more, the FDA has accepted the use of convenient endpoints to allow new drugs to be approved," Dr. Chen said. "It is unfair to patients when we approve therapies that have uncertain benefit to patients because they may think it will help them when it is unlikely to do so."
The FDA use of surrogate endpoints isn't justified based on publicly available data from post-marketing studies and label updates, the study team concludes.
That's because surrogate endpoints don't show whether drugs help people live longer or improve patients' quality of life, the researchers argue. Moreover, post-marketing studies are often not conducted, incomplete, and slow to be performed. And some of these studies have negative results suggesting that the drugs are of little, if any benefit, the authors point out.
One limitation of the analysis is that the authors lacked data on some label updates prior to 2006 that are not publicly accessible.
There is growing evidence that FDA approves oncology drugs with unclear or no clinical effect, and the study offers fresh evidence of this, said Dr. Alberto Falk Delgado of CityAkuten Stockholm and Uppsala University in Sweden.
At best, many oncology treatments approved with surrogate endpoints have a minimal effect on patient outcomes, and at worst, they are "very pricey placebo treatments," Dr. Delgado, who wasn't involved in the study, said by email.
There might be a role for surrogate endpoints If there is clear and strong correlation between surrogate and clinical endpoint, Dr. Delgado added. But correlation does not imply causation, he cautioned.
"Surrogates should be used cautiously," Dr. Delgado said.
SOURCE: https://bit.ly/3bNcPaE JAMA Internal Medicine, online April 27, 2020.
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου