Among 55 cancer drugs recently approved on the basis of a surrogate endpoint, less than one-fifth have been shown to improve survival in follow-up clinical trials.
Fully two-thirds of new cancer drugs in recent years gained regulatory approval based on a so-called surrogate end point, such as tumor shrinkage, rather than a clinical end point directly measuring how patients feel, how well they function or how long they survive.
This shortcut strategy makes sense if the surrogate reliably predicts improvements in survival or quality of life. But that connection remains unknown for a surprisingly large share of surrogate-approved cancer drugs in the U.S., according to a study appearing this week in Mayo Clinic Proceedings.
The authors analyzed cancer drug approvals by the Food and Drug Administration between January 1, 2009, and December 31, 2014. They found that 55 of 83 approvals were based on a surrogate end point. In 25 of the 55 surrogate approvals, the authors were unable to find any published research addressing whether the surrogate correlated with survival.
“How can you say it’s reasonably likely to predict true clinical efficacy if nobody has ever studied it?” says senior author Vinay Prasad, M.D., M.P.H., an OHSU Knight Cancer Institute hematologist-oncologist and assistant professor of medicine. “How can you say a surrogate is established when nobody can find a paper?”
There are sound reasons why cancer clinical trials have come to rely more on surrogate end points. They can be measured more easily, more cheaply and more quickly – months or years earlier than survival. That makes it possible to accelerate the completion of trials and the availability of new cancer treatments.
But there are serious downsides. Unacceptable safety issues may be more likely to go undiscovered in clinical trials using surrogate end points. Such trials tend to provide less reliable information about the safety of investigational treatments because they enroll fewer patients and monitor them for a shorter time than trials designed to measure overall survival.
And when put to the test, surrogate measures of cancer therapies have often shown little or no correlation with survival or other clinically meaningful result. Progression-free survival, for example, measures the share of patients at a given point in time with tumor growth that has not exceeded some arbitrary threshold, typically a 20 percent increase.
But these arbitrary thresholds for change in tumor size may be too small to mark a difference in quality of life or survival. Surrogate end points also are more open to interpretation, measurement error and bias than the straightforward discernment of survival versus death.
There’s no shortage of examples showing that experimental therapies can cause delayed and dangerous side effects that don’t show up in the surrogate measure.
The approval of erythropoiesis stimulating agents for treating chemotherapy-induced anemia is a case in point. The FDA based the approval on data showing improvement in hemoglobin levels and a reduced need for blood transfusions. But later trials linked the agents to a significantly increased risk of stroke and possibly worse survival. The FDA eventually required black box warning labels on the products, and medical societies rewrote practice guidelines to limit their use.
The drug bevacizumab (Avastin) gained an accelerated approval for metastatic breast cancer in 2008 based on dramatic gains in progression-free survival. But in subsequent randomized trials the drug did not help patients live longer even as it caused a range of severe side effects, some of them life-threatening. The FDA withdrew its approval for breast cancer in 2011.
Even if a surrogate end point proves valid for one cancer treatment, that does not mean it will work for predicting the effectiveness of other cancer treatments, nor that it will be valid for the same treatment used in different clinical circumstances. Progression-free survival, for instance, has been shown to be a good surrogate for overall survival in advanced colorectal cancer treated with fluoropyrimidine drugs, but not with newer drugs. And progression-free survival has failed as a surrogate measure of treatments for advanced breast cancer and non–small cell lung cancer.
“A surrogate is a short-cut that doesn’t always tell you about the things you really care about – how long are you going to live, how well are you going to live,” says Prasad.
Prasad and co-author Chul Kim at the National Cancer Institute say that the FDA is too lax in enforcing its own standards for surrogate end points. FDA guidance says the agency grants traditional approvals if a drug demonstrates “a longer or better life or a favorable effect on an established surrogate for a longer or better life.” But Prasad and Kim found that 11 of 30 traditional approvals (37 percent) had no formal analyses of the surrogate-survival correlation.
The FDA grants accelerated approvals based on a surrogate end point that is “reasonably likely to predict” survival or quality of life. Prasad and Kim said this standard has not been enforced. They were unable to find any studies of the validity of the surrogate for more than half of accelerated approvals, 14 out of 25.
In a study last year, Prasad and Kim concluded that more than half of the cancer drugs approved by the FDA between 2008 and 2012 may not be helping patients to live longer. The researchers focused on the 54 cancer drugs that gained FDA approval in those years, and systematically searched for post-approval reports on the clinical end point of survival. They were unable to find evidence of a survival benefit for 57 percent of the approved cancer drugs.
In the new study, the authors performed a similar analysis on the 55 surrogate approvals from January 2009 and December 2014. Only 25 had a subsequent analysis of survival. In 15 of the 25 cases the analysis concluded that the drug did not improve survival.
◊ ◊ ◊
Strength of validation for surrogate end points used in the US Food and Drug Administration’s approval of oncology drugs by Chul Kim and Vinay Prasad, Mayo Clinic Proceedings (2016)
The Strength of Association Between Surrogate End Points and Survival in Oncology: A Systematic Review of Trial-Level Meta-analyses by Vinay Prasad, Chul Kim, Mauricio Burotto, and Andrae Vandross, JAMA Internal Medicine (2015)
Pharmacovigilance in practice: erythropoiesis-stimulating agents by Michael Hedenus, Heinz Ludwig, David H. Henry and Eduard Gasal, Cancer Medicine (2014)
Cancer Drugs Approved on the Basis of a Surrogate End Point and Subsequent Overall Survival: An Analysis of 5 Years of US Food and Drug Administration Approvals by Chul Kim and Vinay Prasad, JAMA Internal Medicine (2015)