I was speaking at the Diagnostic Marketing Association meeting in Philadelphia yesterday. Several speakers got up and made the claim that companion diagnostics and/or biomarkers make it possible to shorten drug development timelines. This isn't the first time I've heard this proposition, but it seems to be gaining traction as a talking point.
There's a nugget of truth to the idea that CoDx will shorten development timelines... it certainly could happen that way in specific cases. Mostly, however, that's just wishful thinking.
The key assumption behind shorter timelines is that biomarkers make it possible to conduct "selected" or "enriched" clinical trials. If you know who your drug works best for and you only test those people... voila! Shorter trials!
There are three key problems with using a selected/enriched design for your pivotal trial:
- FDA is not fond of this study design and may resist accepting it for purposes other than patient safety
- Prior to the pivotal, you probably don't have any idea what the actual relevance of the biomarker(s) may be
- You could be trading moderately shorter timelines for significantly reduced patient populations
Other than that, it's a great idea.
Remember that the intended use of a companion diagnostic is bound to a particular therapy. The pivotal trial not only tests the therapy, but the proposition that the diagnostic is a meaningful adjunct to treatment decisions.
If you don't study the population that isn't selected, how could you know that the diagnostic works as intended? If you can't show that the diagnostic works as intended, how do you know that you've tested the correct patient population for the therapy? Those are difficult questions to answer in the best-case scenario and I promise that "we decided to take a shortcut" isn't going to sound like much of an answer to FDA.
Companion diagnostics are a very exciting new area with many advantages and strategic possibilities. Let's not over-hype them, though. They're important and strategic, but they aren't magical.