Week of:
Yes, no, maybe?
Two weeks ago, the FDA provided an accelerated approval to a drug for Duchenne muscular dystrophy (DMD) called Exondys 51 (eteplirsen). With each passing week, we see a new issue with the clinical trials system that divides the public and stakeholders. While the CEO of Sarepta said it took “courage” for the FDA to provide this accelerated approval, the approval was not without controversy, with Rob Califf penning a very frank open letter on all the considerations that needed to be undertaken before approval. An important piece of the approval was the support of patient advocacy groups, who have not only funded research on DMD, but have also helped by making the voices of patients heard both through formal and informal channels.
Our group, of course, dove right into the issue. Keeping with the onion metaphor, every time we peeled back a layer thinking we had uncovered the answer, we kept finding more information, generating more questions (please see our infographic). For example, the media outlets are simply reporting on the pivotal study involving 12 patients, yet our analysis uncovered at least 48 boys who participated in at least two more trials in addition to the pivotal study. We have also found that most critics are treating this accelerated approval like a regular approval based on just 12 patients. However, this is in no way like a regular approval. The FDA expects studies to still be conducted; two trials are still ongoing, one with an accrual target of 160 patients. Based on the results of those trials, the FDA may choose to withdraw approval.
Our team’s insight from reviewing Exondys 51’s approval returns us to a topic we’ve discussed before: human beings are complex, and healthcare needs to acknowledge this complexity when dealing with disease. Currently, the clinical trials process is a linear process, and often fails to capture this complexity. Drug efficacy is often measured as binary: either the drug works or it doesn’t. In the case of DMD, taking the drug either enabled the boys to produce significantly more dystrophin — the protein that is missing in the disease — than normal or not. In reality, the real question is not simply whether or not the boys produced more dystrophin, but rather how much dystrophin they produced, whether the amount produced was significantly more than normal levels in boys with DMD, and whether that amount translated to a clinical impact on muscular dystrophy symptoms.
In fact measuring efficacy is far more complicated and part of the reason why this drug caused internal strife among key FDA personnel. In the pivotal trial, the boys who took Exondys 51 experienced an increase of only 0.28% of normal levels, but that seemingly small increase was, on average, a tripling of normal dystrophin levels for this group. However, based on results from studies of other forms of MD, FDA reviewers felt the children needed to experience an increase of over 10 percent above normal to experience improved DMD symptoms. Obviously, the question is whether DMD should be measured like other forms of MD. There are now at least 30 identified forms of MD that share only one common characteristic, muscle weakness that increases over time; but, the order in which symptoms occur and which parts of the body are affected are different for each form of MD. For diseases like DMD, a binary endpoint is not helpful, and characterizing the percent increase of dystrophin that signals success is incredibly complicated. Science has not caught up with this disease, and there is no other option for treatment. In this case (and perhaps others like it), the mechanism for measuring success (trial endpoints) must be measured differently.
Not only us, but a number of other researchers have said that the mechanism for measuring treatment success needs to be more nuanced. Andrew Lo, from the MIT Sloan School of Management, has proposed a mechanism for changing the approval process so that characteristics of the disease such as severity and patient population impacted are taken into account. Others have talked about conditional approval, and trying to get treatments to patients quicker instead of subjecting them to traditional methods. Hence, our team strongly feels that rating efficacy in a binary way does not represent the complexity of disease/ treatment interface. We need ways around it.
And in fact, this accelerated approval from the FDA did take into account the complexity! It represents the fact that this is a rare disease, so high recruitment is not possible. It took into account the comments from patients. It specifically ventured into the grey area, and did it in a manner it felt was safe. Yet, it is has been hard for many stakeholders and members of society to acknowledge that. And this, more than anything, is a cultural issue that plagues healthcare. As Atul Guwande would say, we need to think about our lives more holistically, and think about what outcomes we want in our treatments. The drug approval process, an important part of healthcare, should take this into account as well.
