Why a Failed Confirmatory Trial May Not Mean a Failed Drug

The US Food and Drug Administration (FDA) sometimes allows cancer drugs with accelerated approvals to remain on the market after confirmatory trials fail.

The FDA’s Richard Pazdur, MD, and Gautam Mehta, MD, explained why in a recent article in Journal of Clinical Oncology.

In short, Pazdur and Mehta highlighted that a failed confirmatory trial does not necessarily equate to a failed drug. Instead, a failed confirmatory trial — a trial that does not confirm a drug’s clinical benefit — might simply mean that the trial had design issues. The trial may not, for instance, have selected the optimal patient population or drug dosage to verify a clinical benefit.

Failing to show a clinical benefit can mean several things: The trial did not achieve a statistically significant prespecified primary endpoint, the trial demonstrated worse overall survival in the experimental group, or the trial found the toxicities of the drug outweighed its benefits.

So far, 15% (30 of 198) of accelerated approvals in oncology have been withdrawn because the confirmatory trials failed or were incomplete.

But when making withdrawal decisions, the agency carefully considers the nuances of the confirmatory trials. Even if a confirmatory trial fails, the FDA may let an accelerated approval stand if there are no alternative treatments for the indication and the confirmatory trial did not signal a major toxicity or survival issue, explained Pazdur, director of the FDA’s Oncology Center of Excellence, and Mehta, the agency’s associate director for oncology clinical policy.

After all, to win accelerated approval, cancer drugs must demonstrate a direct antitumor effect, typically tumor shrinkage, which signals that some subgroup of patients might still benefit, even if a confirmatory trial fails to hit its key endpoints.

Oncologist Manni Mohyuddin, MD, a blood cancer specialist at the University of Utah, Salt Lake City, told Medscape Medical News that these arguments are reasonable.

“You can have a drug that is active, meaning that it shrinks cancer cells, but it can still fail its confirmatory trial,” Mohyuddin said. And “these drugs may be useful in certain situations.”

But this issue of failed confirmatory trials is more likely to arise with marginal drugs rather than transformative ones, Mohyuddin noted. With “true game changers” — such as checkpoint inhibitors for melanoma and non–small cell lung cancer (NSCLC) — the benefits are likely to emerge in confirmatory trials, regardless of methodological issues in these trials, Mohyuddin said.

Most accelerated approvals today — 60% — are for oncology drugs, and currently, there are 64 oncology drug indications on the US market with accelerated approvals only.

The FDA estimates the accelerated process gets drugs to patients with cancer about 3 years faster than if initial approvals required proof of clinical benefit, but many drugs linger on the market for years with an accelerated approval and no confirmation that the drug provides a clinical benefit.

To address the issue, Congress granted the FDA the power to require companies to have confirmatory trials underway as a condition for the accelerated approval. The withdrawal process has also been expedited for drugs that don’t show a benefit.

Nonetheless, Pazdur and Mehta said there will likely continue to be drugs that remain on the market despite failed confirmatory trials, sometimes to give companies a second chance to prove benefit with a better confirmatory trial design.

Pazdur and Mehta gave several examples of how poorly designed trials can derail otherwise useful cancer drugs.

For instance, the 2003 accelerated approval of the EGFR-targeting tyrosine kinase inhibitor gefitinib was withdrawn in 2011 after several failed confirmatory trials in patients with locally advanced or metastatic NSCLC. It wasn’t until studies were limited to patients with EGFR-mutated disease that a benefit emerged, at which point gefitinib was granted full approval for appropriately selected patients in 2015.

In 2010, the accelerated approval for gemtuzumab ozogamicin for patients with acute myeloid leukemia was withdrawn after the confirmatory trial demonstrated excess deaths. It turned out the dose was too high. New trials at a lower dose given on a different schedule and in a more select population showed a favorable benefit-risk profile, leading to a traditional approval in 2017.

The confirmatory trial for lurbinectedin in refractory metastatic small cell lung cancer didn’t demonstrate an overall survival benefit but also didn’t show harm. Because few other options for this patient population existed at the time, the FDA let the 2020 accelerated approval stand as additional trials were conducted.

Mohyuddin said that as a clinician, he’s glad for the FDA’s nuanced approach because it gives him more options for his patients. At the same time, however, he worries from a policy perspective that the FDA’s leniency might allow more drugs on the market with marginal or no real benefit.

“I see both sides of the argument,” he said.

Overall, the goal of the article in Journal of Clinical Oncology was simply to get the information out to the medical community, Mehta told Medscape Medical News in an email.

“The challenges of regulatory decision-making when a confirmatory trial after accelerated approval ‘fails’ have not been described,” Mehta said. “We hope that this report communicates the critical factors that are considered…and some of the complexities of regulatory decision-making related to accelerated approval[s].”

M. Alexander Otto is a physician assistant with a master’s degree in medical science and a journalism degree from Newhouse. He is an award-winning medical journalist who worked for several major news outlets before joining Medscape Medical News. Email: [email protected]