New Cancer Drugs With No Benefit 'Should Be Withdrawn'

Pam Harrison

December 01, 2016

Most new cancer drugs that were approved by the US Food and Drug Administration (FDA) on the basis of surrogate endpoints retain their approval status and remain available to prescribing physicians even when postmarketing studies show that the new agents do not extend overall survival (OS) or improve quality of life (QoL) compared to placebo or observation groups, a new analysis suggests.

The new analysis, by Diana Zuckerman, PhD, and Tracy Rupp, PharmD, MPH, both from the National Center for Health Research, Washington, DC, was published online November 29 in JAMA Internal Medicine.

The investigators identified six drugs that do not improve OS or QoL and yet are hugely expensive.

"The FDA should promptly withdraw approval for cancer drugs that are proven to have no clinical benefit," declare Scott Bauer, MD, and Rita Redberg, MD, both from the University of California, San Francisco, in an accompanying editorial.

"Removing these drugs, each of which costs between $20,000 and $170,000 per year, from the market will improve the quality and value of cancer care," they add.

Approached for comment, Vinay Prasad, MD, MPH, senior scholar, Center for Health Care Ethics, Oregon Health and Sciences University, Portland, said he couldn't agree more, telling Medscape Medical News in an email, "In our effort to speed drugs to market, we have stopped asking if they truly make patients better off."

Recently, Dr Prasad and colleague Chul Kim, MD, MPH, who is a fellow in medical oncology at the National Cancer Institute, examined the effect that 36 new cancer drugs had on OS.

These drugs had all been approved by the FDA between 2008 and 2012 on the basis of surrogate endpoints.

Dr Prasad and Dr Kim reported that postmarketing studies showed no OS survival benefit for 18 of the 36 drugs they analyzed.

It was these same 18 drugs that Dr Rupp and Dr Zuckerman investigated in the latest analysis. To assess the drugs' true value in the management of cancer, they studied effects on QoL as well as the cost of these drugs.

The authors identified 31 articles describing findings of clinical trials that included a comparison group in which a validated instrument was used for assessing QoL.

Importantly, QoL outcomes had not been studied for more than half of the 18 drugs included in the new analysis, Dr Bauer and Dr Redberg point out.

Indeed, Dr Rupp and Dr Zuckerman identified only a handful of studies in which QoL was assessed through comparisons with placebo or observation.

They found only one drug that improved QoL. That drug was was crizotinib (Xalkori, PF Prism CV) for the treatment of non–small cell lung cancer.

Among the other drugs for which the researchers found QoL data, two worsened QoL (peginterferon alpha for melanoma and cabozantinib for medullary thyroid cancer), and four others had no measurable effect on QoL (everolimus in both renal cell carcinoma and breast cancer, pazopanib for renal cell carcinoma, and bevacizumab for glioblastoma).

"The average cost of the 6 drugs with no different or worse QoL was $87,922 per year," Dr Rupp and Dr Zuckerman observe.

The estimated annual costs of the 18 drugs in their analysis ranged from $20,237 for rituximab to $169,836 for cabozantinib; 13 drugs had annual costs that exceeded $100,000.

The most expensive drug, cabozantinib, did not improve OS and worsened QoL in comparison with placebo, the authors point out.

Dr Rupp and Dr Zuckerman comment that despite having no benefit on OS or QoL compared with placebo and observation, these drugs present significant risks for serious adverse effects, yet all but one of these drugs retain FDA approval.

Not only that, they "remain on the market at prices comparable to those of the most expensive cancer drugs," the authors point out.

They note that the FDA rescinded the accelerated approval for bevacizumab (Avastin, Genentech) for metastatic breast cancer in 2011.

But they also point out that despite this move, Medicare and other major insurers still pay for the drug when it is used for the treatment metastatic breast cancer.

Strict Standards

"For many years, the FDA had strict standards, and drugs had to be proven to have actual clinical benefit," Dr Zuckerman told Medscape Medical News.

However, for more than a decade, the FDA has been under pressure from pharmaceutical companies and even the law to approve drugs more quickly, she said.

"One of the easiest ways to do that is to depend on preliminary data, particularly surrogate endpoints, because you can measure tumor shrinkage in a few months, but it takes a lot longer to measure how long a person lives," Dr Zuckerman explained.

To compensate for basing drug approvals on preliminary data, the FDA felt that they could rely on postmarketing studies to establish how good a drug really was.

In reality, postmarketing studies represent a "catch-22," Dr Zuckerman argued. If patients in a postmarketing study know they have a pretty good chance of receiving placebo, "they are going to drop out of the study if they don't seem to be getting any better," she observed.

Unfortunately, the design of these postmarketing studies is such that patients who drop out often are allowed to cross over from placebo to active treatment. This ability to cross over makes it difficult to assess the true effect of the treatment under investigation.

"We have a system that is already quite broken, and it looks like it's aiming to get more broken," Dr Zuckerman said. (A new law, the 21st Century Cures Act, is to be voted on very shortly by Congress. If passed, it is expected to lower FDA standards for drug approval even further.)

"As it now stands, there is a lot of emphasis on industry as the customer, whereas we think the American public, patients, and physicians should be the most important customer for the FDA, because taxpayers are paying most of the FDA's budget," she added.

"So I think it would be very helpful if oncologists spoke up and said, 'This is not helpful to us. We want to be able to rely on the FDA to tell us what is known and not known about a drug.' And if these surrogate endpoints are not predicting survival, patients and doctors need to know that, because we don't want our patients suffering from nausea and vomiting and exhaustion from drugs that are not going to help them live any longer," Dr Zuckerman told Medscape Medical News.

Sobering Results

Dr Prasad pointed out that his study asked the question, For new cancer drugs that were approved on the basis of a surrogate endpoint, how many later showed improvements in survival?

"Results were sobering, as we found that just 5 out of the 36 we analyzed did improve survival," Dr Prasad said.

In 2009, a report from the Government Accountability Office criticized the FDA for failing to enforce postmarketing study commitments for drugs approved on the basis of surrogate endpoints.

"Our analysis fits with work by the Government Accountability Office showing the FDA is very bad at enforcing postmarketing commitments," Dr Prasad observed.

Dr Prasad did feel that some of the new cancer drugs have earned a place in cancer medicine.

For example, both nilotinib (Tasigna, Novartis) and dasatinib (Sprycel, Apotex) are clearly "transformative drugs" when used for patients with chronic myelogenous leukemia who are intolerant of or resistant to the first-line agent imatinib (Gleevec, Novartis).

However, if used as initial therapy instead of imatinib, "these drugs offer no survival or QoL benefit," Dr Prasad emphasized.

"And the reason why cabozantinib for medulary thyroid cancer is on this list, as well as other drugs that offer no clear survival or quality of life benefit at tremendous cost, is simply regulatory failure," he said.

The authors and editorialists have disclosed no relevant financial relationships.

JAMA Intern Med. Published online November 29, 2016. Full text, Editorial

Follow Medscape Oncology on Twitter for more cancer news: @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.