BioPharm Executive: And So It Continues: Let's Get Rid of Clinical Trials (Part 2)

July 26, 2017
By Karl Thiel for BioSpace.com

Are overly burdensome regulations keeping great new medicines away from the public?

No, probably not. Yet that's a widespread perception, entrenched in parts of both the political right and left. I suggested last month that this simplistic view is pointing reform in the wrong direction. Freeing drug companies of red tape (or forcing them to give up the goods, depending on your view) won't improve peoples' lives. But that doesn't mean there aren't ways in which we could improve our approach to clinical trials and drug approvals by supplementing clinical studies with more real-world data.

Even the best, broadest clinical trial looks at only a small swath of the audience that will ultimately receive a drug. By shining a flashlight into a dark room, we hope to find out what's inside. That often works tolerably well, but many times we only find out later what was lurking in the corners.

The obvious example of this problem is a drug like the painkiller Darvon that appeared safe and effective in clinical trials but turned out to be more dangerous and/or less useful that we first thought. But there are reverse cases, less visible yet potentially more harmful to public health. This includes programs that got killed off at some point in development because a clinical trial returned a false negative due to a skewed control group or simply because the drug was exposed to a group of patients who aren't indicative of a broader or at least different audience that might have benefitted.

It's well documented, for instance, that the placebo effect has been getting stronger over time, at least in the U.S., and the surprisingly strong performance of control groups has torpedoed many drugs. That's particularly true in CNS areas. As a Psychology Today article from last year noted, "Antidepressants are likely to be more effective than our current estimates suggest" (my emphasis). Any number of companies can speak to how it has gotten harder over the years and harder to show a pain drug works better than placebo. Or Alkermes , which is hoping to get approval of ALKS-5461 for major depression after it files an NDA this year. Patients on this drug saw a clear improvement in their symptoms, but it's not nearly so clear that they did better than placebo patients, who also felt much better. Despite a desperate need for better drugs in this indication, a strong placebo effect could complicate Alkermes' chances for approval.

How could we avoid both kinds of losses—both the approvals and the rejections that shouldn't have happened? "Real-world" evidence, gathered both before and after approval, could vastly improve our knowledge of how drugs are likely to really impact health outcomes.

This isn't just an idea. Last year, FDA released its draft guidance on using "real-world evidence" for medical devices. But drug companies are already moving forward on the same front, mining patient medical and insurance records for information that supplements formal clinical data, or enhancing data collection with everything from smartphones to wearable devices.

For instance, Sanofi mined the U.S. Predictive Health Intelligence Environment (PHIE) database to support the switchover of patients from basal insulin to its drug Toujeo. It looked at how blood sugar was controlled in a real-world setting. That was a post-approval analysis, but The Salford Lung Study is an example of the world's first "pragmatic, randomised phase III real-world effectiveness trial" in COPD. This study is looking at Breo Ellipta in an investigational first-line setting. Noting that "double-blind RCTs differ from real life in having highly selective eligibility criteria, and enrolling participants who are not representative of patients in clinical practice and have much higher adherence," this study will allow real-time collection of data from electronic medical records and study how the drug performs in a practical setting.

That approach could benefit companies and patients. One theory on why placebo effects are getting stronger has to do with how modern trials are conducted—more care, a higher level of "touch" and counseling, longer studies with more follow-up. That's great for the patients involved, and helps explain why everyone involved does better. But that's probably not so reflective of how real-world care will play out.

Taking advantage of real-world evidence, though, will require a healthy dose of PR. A recent study from the London School of Economics illustrates the problem. It suggests that FDA is using its Fast Track authority to put "at best questionable" drugs in the marketplace, citing an analysis of 37 drugs given accelerated approval between 2000 and 2013. The analysis is based on the idea that the evidence used for approval is un-rigorous, rather than that the drugs were later proven not to work. But the real goal should be to pass drugs through the traditional clinical process more quickly, then ask more from post-marketing studies. Some real regulatory reform along these lines could help everyone.

-Karl Thiel

Read the BioPharm Executive online newsletter July 26, 2017.
Sign-up for the free monthly subscription to the BioPharm Executive