Bashing accelerated approval isn’t supported by the data

The Food and Drug Administration’s accelerated approval program aims to speed “approval of drugs that treat serious conditions, and fill an unmet medical need based on a surrogate endpoint” of overall survival (progression-free survival). Overall survival is usually considered the gold standard in oncology because people with cancer generally want to take medications that can help them live longer. So judging the FDA’s accelerated approval program without assessing its full impact on overall survival presents a very slanted story.

Case in point: Under the seemingly benign title, “Clinical Benefit and Regulatory Outcomes of Cancer Drugs Receiving Accelerated Approval,” an article published recently in the Journal of the American Medical Association claims that among “cancer drugs granted accelerated approval from 2013 to 2017, 41% (19/46) did not improve overall survival or quality of life in confirmatory trials after more than 5 years of follow-up, with results not yet available for another 15%.”

To the media, that is catnip. A news item in BMJ reported on the study under the headline “Fast tracked drugs often do not improve clinical outcomes.” Bloomberg News got in on the fray with its story, “Americans Are Paying Billions to Take Drugs That Don’t Work.” Even STAT got into the act.

But if you read the JAMA analysis, it’s apparent that patients were not disadvantaged. It’s important to avoid sweeping generalizations, as did an earlier analysis, based on value reference points focusing on the prioritization of cost-effectiveness over clinical benefit. “Although accelerated approval can be useful, some cancer drugs do not end up demonstrating benefit in extending patients’ lives or improving their quality of life,” wrote the authors of the JAMA analysis. Does that really mean anything?

First, here’s what it doesn’t mean: It does not mean these drugs — all approved via the accelerated approval pathway — lack clinical benefit. It does mean they are, at minimum, equivalent to existing first-line therapies, some of which, it’s important to note, are themselves beneficiaries of expedited FDA reviews.

Almost all accelerated approval medicines work — but they’re not all necessarily superior to the current standard of care. Missing from the authors’ analysis is that conversion to full approval is most often based on demonstrating an overall survival benefit.

An analysis using more complete data on how accelerated approvals within the same universe of oncology medicines affect overall survival tells a very different story.

Appropriate focus on overall survival

An analysis published Monday in the Journal of the National Comprehensive Cancer Network examined 69 oncology therapies approved via the accelerated approval pathway through December 2022 for which overall survival data were available. The indications covered rare and fast-progressing late-line cancers, such as persistent, recurrent, or metastatic cervical cancer with a median overall survival of 16 months with six months of improvement in life expectancy, as well as for therapies for slow-progressing cancers.

To collect overall survival data, the authors turned to trial results published on ClinicalTrials.gov and in peer-reviewed publications. If necessary, they extracted overall survival data from sources, such as abstracts from oncology conferences, published health technology assessments, or, in rare cases, real-world evidence.

The study’s key findings include:

From 2006 to 2022, the analysis estimated that, across 69 accelerated approval products for which overall survival data were available, they added about 263,000 life years among approximately 911,000 cancer patients. If this timeline were to be extended to 2026, the additional life years would increase to 382,000.

In a subgroup analysis, the 264,061 people treated with a product given an orphan designation had a gain of 145,413 life years, a 16.5% gain over standard of care.

Moving forward

The new analysis doesn’t mean accelerated approval is perfect. Some drugs approved via this pathway do fail. But it’s important to remember that product failure isn’t the same as process failure.

We have some thoughts on building on its success.

For starters, it’s important to not overgeneralize cancer studies to other rare diseases and some common disorders such as hepatitis B for which biomarkers have been established. Each therapeutic area is unique and different. As Adm. Hyman Rickover said, “The devil is in the details … but so is salvation.”

Next, the health care ecosystem must recognize that it’s hard to anticipate which accelerated approval therapies will result in products that are superior to existing first-line therapies. (Payers pay heed.)

The claim that traditional randomized controlled trials (RCTs) take about the same time as surrogate studies in many tumor types is wrong. Real-world evidence must play a key role in expediting confirmatory work for accelerated approvals, but that doesn’t mean it takes longer — it just means it’s different. A subtext to the “accelerated approval pathway is bad” mindset is the incorrect assumption that any data generated outside of an RCT is similarly bad.

Highly intelligent, dedicated, and experienced researchers and regulators can look at the same datasets and reach different conclusions. That’s science. And the more advanced the science, the greater the nuance and the more difficult the decisions. Nobody said it was going to be easy.

The accelerated approval pathway is available under federal law when a drug intended for a serious disease is expected to provide a meaningful advantage over available treatments. But what does “meaningful” mean? Is it a specific p-value? To be sure, data must always be at the foundation of every FDA decision, but data aren’t always black and white. Claude Debussy has been credited as saying that “Music is the silence between the notes.” The same is true for data, because what often resides between the notes is the patient voice.

A crucial lesson learned from the FDA’s Patient-Focused Drug Development initiative is that patients are often willing to accept heightened risk for what reviewers (both inside the FDA and outside its walls) view as secondary benefits.

The accelerated approval pathway works, but it isn’t — nor is it intended to be — a flawless oracle of therapeutic success. Perfect mustn’t be allowed to be the enemy … of patients with unmet needs.

Peter J. Pitts is president of the Center for Medicine in the Public Interest, a visiting professor at the University of Paris School of Medicine, and a former associate commissioner of the Food and Drug Administration. Timothy R. Franson, M.D., is a principal at Faegre Drinker Consulting, a fellow of the Infectious Diseases Society of America, and a past president of the United States Pharmacopeial Convention.