FDA-Approved Cancer Drugs That Are 'Deeply Problematic'

Nick Mulcahy

June 01, 2016

More than one-third (37%) of traditional full approvals of cancer drugs in recent years have been based on surrogate end points that have never been formally analyzed to determine if they actually translate into a survival benefit for patients, a new study indicates.

In other words, these drugs (and their surrogate end points) might be duds.

Perhaps worse yet, because these drugs received full approval, their approval is unlikely to be rescinded even if future trials show no survival benefit, which is an option with accelerated approvals.

Full approvals made on the basis of unproven surrogates are "deeply problematic," said study coauthor Vinay Prasad, MD, MPH, a medical oncologist at the Knight Cancer Institute, Oregon Health and Sciences University, in Portland.

The study was published online May 20 in the Mayo Clinic Proceedings.

To illustrate the problem, Dr Prasad gave the example of two breast cancer drugs.

Everolimus (Afinitor, Novartis) received traditional full approval for the treatment of advanced breast cancer in 2012 on the basis of a "low correlation" surrogate (progression-free survival).

This meant that it was "biologically plausible" (but not strongly proven) that the progression-free survival improvement might translate into an overall survival improvement, so the drug was approved.

Unfortunately, the surrogate did not pan out.

For everolimus, "the pivotal randomized trial eventually showed no survival benefit," Dr Prasad reported. Nevertheless, the drug will not be pulled off the market for efficacy because it received full approval, he explained.

The case of everolimus is much, much more problematic than that of bevacizumab.

"The case of everolimus is much, much more problematic than that of bevacizumab," Dr Prasad told Medscape Medical News.

In the higher-profile case of bevacizumab (Avastin, Roche), the system actually worked, he pointed out.

Bevacizumab received accelerated approval in 2008 on the basis of progression-free survival. But when three postmarketing studies showed no improvement in survival, the accelerated approval for metastatic breast cancer was rescinded by the US Food and Drug Administration (FDA). "That is how the system should work," he said.

Study Details

For their study, Dr Prasad and his coauthor, Chul Kim, MD, MPH, from the National Cancer Institute, investigated 55 oncologic drug approvals based on a surrogate from 2009 to 2014. Of these, 25 were accelerated approvals and 30 were traditional full approvals.

The pair found that only 30 of the 55 approvals (55%) had a published analysis on the strength of the association between the surrogate end point and overall survival.

They also found that only 19 of the approvals had undergone "high quality" level 1 analysis of the surrogate–survival association.

The bar for traditional approval must be surrogates whose correlation [with survival] has at least been examined.

Dr Prasad said that he "supports approvals based on surrogates." But he is critical of the FDA processes that allow a drug such as everolimus to receive full approval with sketchy, unexamined end points. "The bar for traditional approval must be surrogates whose correlation [with survival] has at least been examined," he said.

Another expert is less worried about the FDA processes.

"In my view, the FDA has struck the right balance between the need to save lives and safety," writes S. Vincent Rajkumar, MD, from the division of hematology at the Mayo Clinic in Rochester, Minnesota, in accompanying editorial.

He says he believes that the new study is "careful" and "thoughtful," and suggests that the problems outlined can be addressed by a "stricter postmarketing enforcement policy" and "aggressive policing to make sure any required confirmatory phase 3 trials are performed in a timely manner."

Dr Rajkumar is worried about impeding the progress of drug development and accessibility, especially for uncommon cancers, which have fewer data for the kinds of analyses that Drs Prasad and Kim would like to see.

The United States often leads the way globally with cancer drug approvals, he notes.

Dr Rajkumar, who is an expert in multiple myeloma, points out that the recent slew of new drugs for this condition (bortezomib, carfilzomib, pomalidomide, and daratumumab) were all approved by the FDA on the basis of single-group uncontrolled studies that used response rate as a surrogate end point.

"In each of these instances, we assume that in many countries where there are more restrictive rules for drug approval, numerous patients must have died before they had access to the possible benefits of these drugs," Dr Rajkumar writes.

But he also believes that approvals should not be too lax.

"The current FDA standards for oncology drug approval are rigorous but fair and should not be further relaxed, especially with new precision medicine drugs for which some have advocated a more liberal approval policy," he says.

Dr Rajkumar also gives an example of a cancer drug that has been an effective treatment but lacked the gold standard of overall survival data at approval.

Thalidomide was approved in the United States to treat newly diagnosed multiple myeloma on the basis of a surrogate end point (overall response rate) that is still not validated as a reliable surrogate for overall survival in previously untreated patients, he explains.

With thalidomide, "the high probability of saving lives trumped concerns about safety or inadequate proof of efficacy, and time has confirmed that this was the right decision," says Dr Rajkumar.

Ironically, thalidomide is the drug that led, in 1962, to the Kefauver–Harris Amendment to the United States Federal Food, Drug, and Cosmetic Act. The landmark amendment mandated that pharmaceutical companies seeking drug approval must provide proof of clinical benefit (efficacy) from well-controlled studies, observes Dr Rajkumar.

But Drs Prasad and Kim emphasize that surrogate end points used in approvals are supposed to have a standard: they must be "reasonably likely to predict" clinical efficacy. And that standard is not met when there is no previously published formal analysis of the strength of a surrogate–survival correlation, they argue.

Dr Prasad and Dr Rajkumar have disclosed no relevant financial relationships.

Mayo Clin Proc. 2016;91:689-691, 713-726. Editorial, Abstract

Follow Medscape senior journalist Nick Mulcahy on Twitter: @MulcahyNick

For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.