This article was published on April 1, 2023, at NEJM.org.
List of authors.
- Joseph S. Ross, M.D., M.H.S
- Karina M. Berg, M.D.,
- and Reshma Ramachandran, M.D., M.P.P., M.H.S.
On January 6, 2023, the Food and Drug Administration (FDA) granted accelerated approval to lecanemab (Leqembi), a monoclonal antibody for treating Alzheimer’s disease that targets βamyloid. New drug approvals for diseases causing substantial morbidity and mortality are typically lauded by patients, clinicians, and industry alike. The lecanemab approval, however, has been clouded by controversy over recent FDA authorizations under the accelerated approval program, illustrating the challenges faced by an agency tasked with both enabling timely access to innovative products and ensuring product safety and efficacy.
The accelerated approval program enables earlier regulatory approval of drugs that treat serious conditions and fill unmet medical needs. Approval is based on results from pivotal trials using surrogate end points that are deemed “reasonably likely” to predict clinical benefit and requires postmarketing studies verifying such benefit. Originally used for HIV drugs, the program has most commonly been used for oncology therapies,1 and questions have arisen regarding the predictiveness of surrogate markers, delays in completion of postmarketing requirements, the evidence used to verify benefit, and the FDA’s ability to withdraw products that fail in confirmatory trials. 1-3 For instance, for 71% of oncology drugs receiving accelerated approval for which associated postmarketing requirements were fulfilled, benefit was “verified” using surrogate end points. 2 Nevertheless, the FDA has increasingly expanded use of the accelerated approval program beyond HIV and oncology therapies.
This shift became apparent in September 2016, when the agency granted accelerated approval to eteplirsen (Exondys 51) for Duchenne’s muscular dystrophy, a typically fatal genetic disease.
The approval was based on a single trial involving 12 patients and using a controversial surrogate end point — the increase in dystrophin found on muscle biopsy (no benefit was observed using the functional 6-minute walk test) — and it was granted despite
recommendations against approval from both the FDA’s scientific staff and its external advisory committee. The postmarketing trial to verify benefit did not begin recruitment until July 2020, anticipating enrollment of 154 patients who will all be randomly assigned to receive different doses of the drug (without a placebo control); it is not expected to be completed until November 2024.
Public awareness of concerns about accelerated approval peaked in June 2021, when the FDA granted such approval to aducanumab (Aduhelm), another monoclonal antibody targeting βamyloid in Alzheimer’s disease. In that case, approval was based on two large randomized, controlled trials that were terminated for futility because aducanumab did not slow cognitive decline, but both trials revealed reductions in brain β-amyloid plaque, another controversial surrogate marker. Again, approval was granted against the recommendation of FDA staff and nearly unanimous agreement of an external advisory committee that aducanumab was not effective and might not be safe. The completion deadline for the postmarketing trial was set for August 2029.
Aducanumab’s approval established a precedent for lecanemab’s, which was based on reductions in brain β-amyloid plaque observed in a single 856-patient trial. Again, approval was granted despite objections from FDA biostatisticians that the study’s prespecified primary clinical end point was not met, and despite uncertainty over whether reducing β-amyloid predicts cognitive benefit. The decision was made without input from an external advisory committee and also without formal consideration of the sponsor’s 18-month clinical trial involving 1795 patients, which was published a month before the approval. 4 Although the trial revealed slightly delayed cognitive decline, independent experts characterized this change as having unclear clinical relevance and raised concerns about trial loss to follow-up, functional unblinding, and serious safety problems, including patient deaths.5
in internal communications discussing the FDA’s eteplirsen approval, Janet Woodcock, then director of the FDA’s Center for Drug Evaluation and Research (CDER), said that the FDA should exercise “the greatest flexibility possible” under its statutory authority in considering eteplirsen’s efficacy. She cautioned that “if Sarepta [the sponsor] did not receive accelerated approval for eteplirsen, it would have insufficient funding to continue.” Billy Dunn, then director of the FDA’s Office of Neuroscience, made similar comments at a recent advisory-panel meeting, explaining that “[the FDA’s] underlying legal authority is clear in not only allowing, but also endorsing and encouraging, the application of regulatory flexibility in the setting of serious and life-threatening diseases.” And when the FDA approved aducanumab, Patrizia Cavazzoni, current CDER director, noted that, “As we have learned from the fight against cancer, the accelerated approval pathway can bring therapies to patients faster while spurring more research and innovation.”
These comments exemplify what appears to be a newly empowered stance by the agency in its use of accelerated approval. The sentiment will resonate with people who believe the FDA should prioritize timely access to innovative medical products, but it raises serious questions for clinicians, patients, and caregivers. Previously, decisions to prescribe or use medications would generally be straightforward: FDA approval clearly signaled that drugs were safe and effective for use and that their benefits outweighed their risks. But with more aggressive adoption of the accelerated approval pathway, particularly in areas other than oncology, clinicians, patients, and caregivers are being asked to make individual decisions about whether “reasonably likely” to predict benefit outweighs personal risk, despite substantial residual uncertainty. Their decisions are complicated by the considerable financial costs of these new medicines, the effort needed to navigate health insurance plans, the time and money spent traveling to receive treatment, and the products’ potential harms.
In the recently passed Food and Drug Omnibus Reform Act, Congress included FDA-supported modifications to the accelerated approval program, giving the agency the power to request thatconfirmatory trials are initiated before a company receives accelerated approval, streamlining the process for withdrawing drugs when postmarketing studies are negative or delayed, and requiring companies to provide confirmatory trial status reports to the FDA every 6 months. These reforms are long overdue and are likely to strengthen the program.
There are also other steps that the FDA might consider to better support clinicians, patients, and caregivers as the agency broadens use of accelerated approval. First, the agency could consider whether the pathway is appropriate for common diseases associated with substantial morbidity and mortality, even if other program conditions are met. Unlike rare neurologic disorders such as Duchenne’s muscular dystrophy, Alzheimer’s disease affects millions of Americans and has a measurable health effect within years, not decades, making it feasible to recruit participants for large, rigorously designed trials examining clinical benefit before marketing authorization.
Second, the FDA could consider establishing opportunities for public input whenever new surrogate markers are being considered “reasonably likely” to predict clinical benefit and establishing standardized reevaluation periods, such as every 5 years, for all surrogate markers to ensure that evolving scientific evidence continues to support their use. Third, external advisory committees could be asked to weigh in on all potential accelerated approvals, which would foster transparency and build public trust. Fourth, the FDA could prepare informational materials to support patients and clinicians that leverage validated decision-support tools to communicate anticipated benefit and potential harm, given the residual uncertainty at the time of approval. And finally, the period allowed for fulfilling postmarketing requirements could be shortened.
Clinicians will also need to take steps to help patients and caregivers make individual decisions about using drugs granted accelerated approval. The ongoing uncertainty about these drugs’ safety and efficacy puts clinicians in an unusually challenging position, requiring substantially more time spent with patients to clarify “known” versus “anticipated” benefit. It is particularly difficult to provide patients with information on potential harms, since trials supporting accelerated approvals are smaller and shorter than those used to verify clinical benefit, and thus less likely to identify safety concerns. Moreover, real-world use differs from trials in terms of adverse-event monitoring; the aducanumab and lecanemab trials, for instance, included monthly brain imaging for early identification of amyloid-related harms, such as brain edema and microhemorrhage, before patients experienced clinical consequences.
An empowered FDA may prioritize using its regulatory authority to enable timely access to innovative products, but to ensure continued trust in the agency, this priority should be balanced against the challenges clinicians, patients, and caregivers face when there is substantial residual uncertainty about product safety and efficacy.
Disclosure forms provided by the authors are available at NEJM.org.
This article was published on April 1, 2023, at NEJM.org.
From the Section of General Internal Medicine, Department of Internal Medicine, and the Yale Collaboration for Regulatory Rigor, Integrity, and Transparency, Yale School of Medicine (J.S.R., R.R.), the Department of Health Policy and Management, Yale School of Public Health (J.S.R.), and the Center for Outcomes Research and Evaluation, Yale New Haven Health System (J.S.R.), New Haven, and the Center on Aging, University of Connecticut School of Medicine, Farmington (K.M.B.) — all in Connecticut.