Rare Diseases and Powering Randomized Clinical Trials

How do you get data when there are only 140 patients in the entire country?

What was the FDA thinking when they rejected the initial application for elamipretide? I think this graphic, posted a few days ago on You Can Know Things, provides some insights.

In 1979, a Canadian task force summarized what eventually became known as the “hierarchy of evidence.” While most depictions of the hierarchy display a simple pyramid, the graphic above includes a scale of “Flaw level” on the left, qualifying statements within the pyramid, and three general evaluation statements on the right. I think it’s safe to say that these additions generally reflect the larger scientific community’s opinions on different types of evidence.

But what if your patient population is so critically ill that true randomization and placebo testing becomes unethical? Or what if your patient population is so small that you will never be able to get a sample size large enough to move out of the “Meh, need more data” category?

I posed this question to the creator of the graphic, Dr. Kristen Panthagani, and she graciously responded:

“Your question applies broadly: what do we do when we don't have the higher quality studies, either because the disease is rare or the research isn't being prioritized? In that case, I think it's totally reasonable to go with the highest quality data that is available, understanding the limitations. If all that is available is case reports because the disease is so rare -- that's still valuable information, but should also be interpreted with the limitations in mind. Many times clinicians must make decisions with limited information, and a case report can be a valuable tool to help guide them. But if we have better studies, then those take precedence. That's the overall intent of the pyramid: to help people rank the quality of information, not to dismiss every study that doesn't fall into the top of the pyramid.”

In his recent book Proof: The Art and Science of Certainty, Adam Kucharski tells the story of the development of ECMO (the same life support system featured in The Pitt). In the 1980s, despite “over seven hundred critically ill newborns [having] received the treatment, with 82 percent surviving, far higher than the 20 percent predicted chance of survival,” skepticism about the use of ECMO with infants remained because of the lack of randomized treatment. This led to a randomized trial design with a built-in stopping point if four babies in the control group died. Four infants in the control group did indeed die, and all study participants thereafter received ECMO.

We need the FDA to understand that the hierarchy of evidence should not be an altar at which we sacrifice the lives of patients in pursuit of the ideal clinical trial. Instead, it should be a tool that empowers us to make thoughtful decisions that improve patient outcomes. In early July, Stealth BioTherapeutics submitted a request for the FDA to reverse their previous decision. It’s time to approve elamipretide.

Help advocate for Gilbert and other patients like him: gogilbygo.com