Experts Say Surrogate Endpoints Need More Regulatory Oversight
By Deborah Borfitz
October 4, 2021 | “The benefits of using surrogate endpoints in drug trials are often outweighed by the harms of not knowing in a timely manner whether a new drug is effective in improving patient-relevant outcomes,” according to Huseyin Naci, associate professor of health policy at the London School of Economics and Political Science. Drugs approved based on surrogate endpoints—indirect measures for predicting outcomes such as laboratory values and radiographic images or, in oncology, response rates and disease-free survival—are sometimes later found to be ineffective or harmful.
As Naci and his colleagues argue in a recent analysis published in The BMJ (DOI: 10.1136/bmj.n2191), growing regulatory reliance on surrogate endpoints is problematic for stakeholders across the board and their use should be restricted to chronic diseases where collection of the relevant outcomes might otherwise require unreasonably lengthy studies. In oncology, their use should be restricted to phase 2 studies, says Naci, as suggested by a University of California, San Francisco oncologist in a feature article in the same issue.
Likewise, he concurs with a list of recommendations by physicians interviewed for the article, including convening a panel of independent industry experts and patient representatives to establish when expedited pathways are warranted and requiring confirmatory trials be fully enrolled before a drug is approved—and having companies provide products at low or no cost in the interim. Naci co-authored an article published last year in The Lancet (DOI: 10.1016/S0140-6736(19)33177-0) making the case that regulators should limit the use of expedited programs “for a clearly demarcated and pre-specified set of circumstances,” he notes.
Earlier research in JAMA (DOI: 10.1001/jama.2017.9415) showed many required post-marketing studies remained incomplete or delayed even many years after approval. Naci therefore supports proposals that post-marketing studies be well underway when a “conditional” approval is granted by regulators.
Another good move would be to balance surrogate endpoints with greater use of patient-reported outcomes, including health-related quality of life, says Naci. Quality-of-life outcomes data are “rarely collected” when new drugs are approved.
Reasonable use of surrogate endpoints would include relapse rate for multiple sclerosis studies and an absolute reduction in the level of low-density lipoprotein cholesterol for cardiovascular disease trials, he says, as well as disease-free survival for early-stage gastric cancer.
Concern over growing use of surrogate endpoints recently came to a head with approval by the U.S. Food and Drug Administration (FDA) of Biogen’s aducanumab for treating Alzheimer’s disease. The decision was supported by clinical trials where prediction of clinical benefit was based on reduction in brain amyloid plaques.
But the new analysis builds on a larger body of research finding a “lack of reliable relationship between surrogate endpoints and patient-relevant outcomes,” says Naci. “It is important to note that we are not only advocating for more limited use of surrogate endpoints. We are also highlighting the importance of using appropriate methods to evaluate surrogate endpoints when their use is warranted [in chronic disease settings when collecting data on clinical outcomes can take many years].”
Unfortunately, he continues, “established meta-analysis methods are rarely used by regulators and health technology assessment bodies to examine the strength of the relationship between surrogate endpoints and clinical outcomes.”
At last count, a shocking 60% of clinical trials (almost 80% in oncology) were measuring the efficacy of new drugs using surrogate endpoints alone, says Naci. This is directly related to increased use of expedited regulatory pathways in both the U.S. and Europe, based on pressure from drug developers.
While the drugs may be only “conditionally” approved, confirmatory studies tend to also use surrogate endpoints, he points out. To make matters worse, an FDA-approved drug gets automatically included in certain clinical guidelines that in turn govern the payment policies of the Centers for Medicare & Medicaid Services.
“What this emphasizes is the importance of regulators’ decisions as gatekeepers to the market,” says Naci. “Regulators can influence the ultimate quantity and quality of evidence available on new drugs. Once drugs are approved, companies have limited, if any, incentives to continue evaluating [those products] and generating additional evidence.”
When regulators accept surrogates as a basis for approval decisions, other stakeholders—payers, health technology assessment bodies, and guideline panels—are “put under immense pressure to appraise the drugs using near-identical data.”
The German Institute for Quality and Efficiency in Healthcare is one of few regulatory bodies to have prescriptive criteria for accepting surrogate endpoints, says Naci, so what constitutes a “valid” surrogate is not well defined. Published frameworks for evaluating surrogate endpoints rarely get used by either regulators or health technology assessment bodies.
One positive sign in the U.K. is that the National Institute for Health and Care Excellence (NICE) recently proposed changes to strengthen evidence requirements for the use of surrogate endpoints, he adds. “NICE’s recent changes to its methodological guide is a very important step in the right direction. NICE is an exemplary health technology assessment body, and its standards are closely monitored by other organizations worldwide.”
In the U.S., there are also promising signs on the payer side, says Naci. In a recent report to Congress, the Medicaid and CHIP Payment and Access Commission advised increasing statutory Medicaid rebates for new drugs approved under its accelerated pathway until their clinical benefits have been verified.