Is FDA Getting It Right on Cancer Drug Approvals?

Not every cancer drug approved by the FDA receives approval and coverage in England and Canada, two studies in JAMA Internal Medicine showed.

In England, a retrospective cohort study looked at 68 drugs given accelerated approval by the FDA from December 1992 to May 2017 that also received market approval in the European Union. The researchers, Elias Mossialos, MD, PhD, of the London School of Economics and Political Science, and colleagues, found that only 45 drug indications (66.2%) were recommended for public coverage through the National Health Services (NHS), and that 39 of the 45 (86.7%) were recommended only after price negotiation or collection of further data to verify the clinical benefit.

The Canadian study, by Bishal Gyawali, MD, PhD, of Queen’s University Cancer Research Institute in Kingston, Ontario, and colleagues, looked at 104 drugs submitted to the pan-Canadian Oncology Drug Review (pCODR) from 2011 to February 2020 and found that only 75% were recommended for coverage, and that the majority of those (92.3%) were given conditional approval — i.e., pending proof of cost effectiveness (97.2%).

The data in both studies are in contrast to policies in the U.S., where most drug approvals are based on surrogate markers, explained Vinay Prasad, MD, MPH, of the University of California San Francisco, and Myung S, Kim, MD, of Oregon Health and Science University in Portland, writing in an accompanying commentary.

In the U.S., cancer drugs are a “protected” class of drugs, where coverage of all drugs approved in that therapeutic area require coverage without price negotiation, Prasad and Kim explained. “The well-intentioned effort to provide drugs to patients with cancer faster has led the U.S. to approve and cover many expensive drugs with substantial uncertainty about their clinical benefit. These choices paradoxically lead other high-income nations to delay or abandon uptake of these medications entirely because of reservations about efficacy and value.”

Further, a lack of routinely documented survival or quality-of-life benefits in trials in the U.S. “makes it more difficult for other countries to obtain the evidence they need to justify coverage,” the commentators wrote.

In the study by Mossialos and colleagues, 60% of approved drug indications relied on the same surrogate measures used by the FDA for accelerated approval. Of the seven non-recommended drug indications, five were denied authorization or coverage by the U.K.’s National Institute for Health and Care Excellence (NICE) because of a combination of clinical benefit or cost-effectiveness concerns; the remaining two were denied based on low cost-effectiveness alone.

The seven drugs had a mean cost per quality-adjusted life-year (QALY) of $221,000 — almost four times greater than the mean QALY of drugs in the recommended group, the researchers reported.

Of the 39 drugs recommended by NICE that required either price negotiation of additional evidence of clinical benefit, 48.7% required more than one agreement designed to improve cost-effectiveness, restriction of the drug indication to a specific subpopulation, or collection of additional evidence, Mossialos and colleagues said.

They suggested that the U.S. public insurance programs could learn lessons from the NHS: “First, after clinical outcome data from postmarket confirmatory clinical trials become available, these data can be used to create formularies that are similar to those of the NHS, reevaluating coverage decisions for drugs based on their safety and efficacy benefits. Second, for drugs which have yet to complete confirmatory studies, public insurance programs can make access conditional on price discounts or additional collection of data measuring clinical outcomes, such as overall survival and quality of life,” the team noted.

During the period of the Canadian study, pCODR received only 104 submissions, in contrast to 163 FDA approvals during the same time, indicating that the pCODR framework might filter out drugs with minimal clinical utility, Gyawali and co-authors suggested.

The 75% of drugs that received a positive recommendation were more likely than those that received a negative recommendation to have a phase III randomized clinical trial design (92.3% vs 53.8%; P<0.001) and had substantially more benefit according to the European Society for Medical Oncology-Magnitude of Clinical Benefit Scale (61.5% vs 19.2%; P<.001).

“Even in a regulatory system like the pCODR with a central evidence appraisal process, including assessment of evidence quality, magnitude of clinical benefit, and cost-effectiveness, cancer drugs without meaningful patient benefit are entering the market,” the investigators wrote.

Progression-free survival (53.8%) and overall survival (32.1%) were the most common primary endpoints associated with successful drug submissions, and half of these that received a positive recommendation had shown an overall survival benefit, with a median benefit of only 3.7 months.

There was 100% concordance between pCODR recommendations and provincial funding decisions, the researchers said.

Prasad and Kim observed that the U.S. system, where the FDA is a regulator and formulary maker, “provides little downward market pressure and works to keep drug prices high.”

“High prices make the U.S. the most valuable market for new drugs, which in turn leads to global development of drugs with marginal or unproven benefit and little focus on cost-effectiveness,” the commentators continued. “The U.S. should understand and pay attention to the dynamics of drug approval in other countries. We should consider the possibility that our drug policy has negative repercussions for patients with cancer worldwide.”