Endpoints News - When efficacy is borderline: FDA needs to get more consistent on close-call drug approvals, agency-funded research finds
In the exceedingly rare instances in which clinical efficacy is the only barrier to a new drug’s approval, new FDA-funded research from FDA and Stanford found that the agency does not have a consistent standard for defining “substantial evidence” when flexible criteria are used for an approval.
The research comes as the FDA is at a crossroads with its expedited-review pathways. The accelerated approval pathway is under fire as the agency recently signed off on a controversial new Alzheimer’s drug, with little precedent to explain its decision. Meanwhile, top officials like Rick Pazdur have called for a major push to simplify and clarify all of the various expedited pathways, which have grown to be must-haves for sponsors of nearly every newly approved drug.
Part of the reason the Aduhelm decision was so controversial is that it occurred without any precedent, and the authors of the article noted how a database to identify related decisions and a tradition to build on them is necessary for the FDA to explain itself moving forward.
“The FDA needs a database of its own decisions that enables the rapid retrieval of cases with similar evidential characteristics and context and an expectation that FDA officials cite previous decisions in their decisional memos. This would enable the FDA to improve institutional memory and learn from its own decisions, thereby increasing the transparency and consistency of its determinations,” the researchers wrote.
That crucial institutional memory is an Achilles heel for the FDA, hampered by the revolving door with industry and the retirement of top officials, like CBER’s Marion Gruber.
“Multiple FDA staff leaders and collaborators confirmed that the FDA has weak structures to support institutional memory, particularly one that crosses FDA therapeutic areas, offices, or centers; much depends on staff memories and is lost with turnover. This balkanization and fragility of institutional knowledge diminishes institutional efficiency and consistency,” the research published Tuesday in the Annals of Internal Medicine said.
For the study, members of both CBER and CDER’s offices of biostatistics and investigators at Stanford University pored through 912 applications for new drugs, finding 117 that went through multiple review cycles, although only 22 of these applications faced additional FDA review primarily because of issues related to clinical efficacy.
“Concerns about the end point, the clinical meaningfulness of the observed effect, and inconsistent results were common bases for initial rejection,” they wrote. In seven of the 22 cases, the approval did not require new evidence but rather new interpretations of the original evidence, and importantly, “No FDA decisions cited reasoning used in previous decisions.” The clinical meaningfulness of the treatment effect was questioned for eight applications, and in six of these, the issue was resolved by submitting new pivotal studies.
Overall, 15 of the 22 applicants provided additional data, data corrections, or new subgroup analyses, while seven applicants supplied none. For eight applications, the products were approved by the division director despite unresolved issues.
The researchers also highlighted some concerning issues where a simple change in review division led to an approval (which was also the case with Biogen’s Aduhelm):
"We saw multiple instances of evidential requirements changing across review cycles, and some decisions changed not with new evidence but with a change in the division considering the drug. In 2 cases, the applicant’s apprising the FDA of its own past decisions shifted the balance in favor of approval. Although each decision had an explainable internal logic, it was not clear whether the same reasoning was applied in other cases; each decision invoked a bespoke rationale."
More expedited designations, less use of 2 pivotal trials
The FDA’s increasing reliance on expedited pathways, particularly in cases where there’s a thin line between an approval and a CRL, and its reliance on less data are also coming into focus.
The researchers found the use of expedited designations has increased dramatically in a decade, from 51% of submissions (163 of 319) submitted between 1997 and 2007, to 81% (48 of 59) in 2018. The number of approvals supported by the traditional standard — two pivotal studies with statistically significant results — decreased from 81% in 1995 to 1997 to 53% in 2015 to 2017.
“When we take a look at all of the pathways — we have fast track, breakthrough, priority review, accelerated approval, RMAT — should we try to simplify some of these pathways?” FDA’s top oncology official Rick Pazdur, speaking at a Friends of Cancer Research conference on Monday, said. “Should there be an early and a late breakthrough therapy designation?”
Pazdur, who earlier this summer called the accelerated approval pathway “under attack,” noted how the needs of sponsors vary dramatically, as some submit for their breakthrough designations early in development and some submit at the time of NDA or BLA submission.
“Rather than taking a look at all of these different programs, perhaps we need some simplification and to concentrate on what are the core messages of these programs,” Pazdur said while mentioning other pathways that are not codified in regulations, including OCE’s real-time review and assessment aid.
While again noting the importance of the accelerated approval pathway, Pazdur, who has sharply critiqued some dangling accelerated approvals in the past, also pointed to the need for sponsors to engage with FDA early on about using the pathway and not to just use it as a last resort.
“Frequently, you find sponsors that kind of stumble on a finding and they say, ‘Aha, I want to do accelerated approval.’ But really you need a discussion on the confirmatory study, when is it going to be submitted, how many patients are going to be entered on that study before accelerated approval is granted,” he added.