What can we learn from the Food and Drug Administration’s (FDA’s) controversial approval of the first drug for Duchenne Muscular Dystrophy (DMD)?

Everybody at the FDA agrees: “There were major flaws in the design and conduct of the clinical trials using eteplirsen (brand-name Exondys 51).” Nevertheless, against the recommendations of the FDA’s expert panel, Janet Woodcock, director of the FDA’s Center for Drug Evaluation and Research, approved the drug under the FDA’s accelerated approval pathway for a subset of patients, approximately 13 percent of the DMD population with a gene mutation amenable to the drug’s action. Upon appeal by agency scientists who disagreed with Dr. Woodcock, FDA Commissioner Robert Califf upheld her decision, saying that both sides, in reaching their different conclusions, had exercised reasonable scientific judgment. One week later, he called for retraction of one of the papers that Sarepta Therapeutics, the maker of Exondys 51, had relied upon for data on the drug’s effectiveness.

What’s going on?

How can a regulator approve a drug, while simultaneously questioning the quality of the evidence about its efficacy?

Has the FDA succumbed to the pleas of desperate parents and in so doing, lowered its scientific standards? Or has it demonstrated compassion and a path forward that is justifiable?

My answer is that the controversial decision is defensible, but only if there is adequate follow through.

Compassionate response to the most vulnerable among us is a virtue, often described as emanating from the “rule of rescue,” which calls on us to meet the needs of strangers in dire circumstances. Nothing is more dire than children living with fatal illnesses, like DMD. These children are terminally ill, there are no other viable alternatives, the drug seems to be safe, and while it is being used, it can be studied. Dr. Woodcock seemed to be making an explicitly compassionate argument based on the virtue of rescue, when she noted at a 2016 FDA advisory committee meeting that there is a great deal of effort that goes into avoiding the mistake of approving a drug that is not effective, but “there is often little consideration of another error, which is failing to approve a drug that actually works.”

What’s next?

But now there must be follow through. First, responsible drug development under accelerated approval calls for corporate restraint from hyperbolic claims. Indeed, in approving Exondys 51 for accelerated approval, Califf noted that statutory and regulatory provisions governing accelerated approval “require sponsors to submit promotional materials to the FDA for pre-dissemination review.” Hopefully, as the Commissioner urged, the FDA will be vigilant in ensuring that packaging and marketing materials do not overstate what is known about the drug’s efficacy.

Second, if the now-required post-marketing trial does not demonstrate clinical benefit, FDA approval must be withdrawn. There are precedents for withdrawing approved indications that are later discovered to be mistaken, like Avastin, which was approved in 2008, and had approval withdrawn in 2011, when subsequent trial data showed no benefit in overall survival or quality of life and significant life threatening side effects. Nevertheless, once approved, it is hard to roll back all the consequences of a premature approval. Medicare continues to pay for Avastin, and the National Comprehensive Care Network’s 2014 guidelines for patients still list the drug as a treatment for Stage IV breast cancer.

Moreover, if evidence proves lacking, insurers will refuse to pay for it (as Anthem already has), and they will deserve back up from health policymakers and others to withstand the public controversy that will arise from coverage refusals.

If approval is ultimately withdrawn for Exondys 51, its use over the years until we have better evidence will have cost millions and burdened families, who—believing in it—may have had to make dire decisions about how to cover co-pays. Clinicians have a role to play here. Most, if not all of these children are under the care of physicians, who should make a point of explaining to parents how uncertain the science still is on this drug, helping to modulate false hopes and guide appropriate parental decision making.

Third, going forward, the FDA should do much more to insist that sponsors follow FDA advice earlier in the research process. In this case, the FDA alerted Sarepta that it needed a larger sample and controls, but that advice did not translate into action. More emphasis on the importance of complying with FDA advice about study design should be placed earlier in the process, so that sponsors follow FDA recommendations for study design before bringing applications for accelerated approval.

Patient Engagement: A Double-Edged Sword

This is some of what can be done to ensure the bar on scientific standards has not been lowered. But there is more that as a society we have to figure out. Patient engagement in science policy in general, and specifically in drug policy, is a principle we can, and should, embrace. Were it not for the AIDS activists of the 1980s and early 1990s, HIV/AIDS might not have transformed from a death sentence to a manageable chronic illness. Listening to patient experiences, even when the scientific evidence is scant, is a good policy that ensures the experts and the powerful stay alert to the human needs science aims to address.

However, patient engagement is a double-edged sword, especially when coupled with the financial interests of other stakeholders. Consider, for example, the hundreds of millions of dollars and the human pain and suffering that was caused by over-exuberant patients who were desperate and health systems that stood to make huge revenues, by too quickly embracing unproven bone marrow transplantation for the treatment of breast cancer. There is no substitute for the scientific method.

As this election cycle demonstrates, we are already in what has been called a “post-fact” world in our political life. Let us hope that science and science policy do not find themselves in the same predicament — where evidence is pushed aside by the overly fervent.

That said, it would be naïve to think that evidence alone can be determinative. It behooves us to find ways to engage patients, but we also should engage other more disinterested members of the public, who can be charged with considering a policy’s impact on a range of issues vital to the common good, not just on rescue. There are many questions that citizens should weigh in on — questions, like: Will approval of this drug impede, or even eliminate, our ability to gain important knowledge about it? Will approval limit the evidentiary standard the FDA should use? If a drug is being considered for approval on the basis of limited evidence, how will both the potential benefits and the full range of possible negative side effects be explained to patients?

Institutional integrity—and the very ability of our society to carry out the biomedical research enterprise in ethically appropriate ways—requires that scientists and policymakers withstand pressure from industry, patients, Congress, and the press. In drug development these pressures are intense and omnipresent, as desperate patients and their families seek rescue, research sponsors seek cures, and investors seek returns. Adherence to evidence must remain a sine qua non. In addition, we must find ways that experts, patients, and the public can examine evidence, weigh competing claims, and balance or prioritize values, so that together we can make the hard choices that confront us.