I think sometimes we talk about biomarkers in a single context and it actually helps to separate them out into different types of biomarkers. And I think the most important first step for a biomarker is a target engagement marker. So if you know how your treatment works, try to see if it’s engaging the target you believe it’s engaging. And that’s sort of a first pass to say, “Is my treatment actually doing what I think it is on a molecular level?”...
I think sometimes we talk about biomarkers in a single context and it actually helps to separate them out into different types of biomarkers. And I think the most important first step for a biomarker is a target engagement marker. So if you know how your treatment works, try to see if it’s engaging the target you believe it’s engaging. And that’s sort of a first pass to say, “Is my treatment actually doing what I think it is on a molecular level?”. And then the second step would be to find a biomarker that is predictive of clinical outcome and see if that biomarker is also changed and usually that would be like a downstream marker, it could be something like tau where you think that the changes in this biomarker will predict changes in clinical outcomes and you’re unlikely to see changes on that biomarker unless you’re also impacting the disease progression. And what I’ve seen is that we’re in a situation where amyloid has been used for accelerated approval for aducanumab and lecanemab and because of that it’s sort of taken on a surrogate-like status, but it’s really more of a target engagement marker in a lot of ways. But at the same time, those changes have been at least somewhat predictive of the changes we’ve seen clinically. But we run into these issues of timing. How quickly or how early does the change in the biomarker happen? How long does that take to impact the actual clinical aspects? And figuring out those correlations is an important part of this equation too.
And I think where we are right now is people are happy to have something we can use for an accelerated approval and something that can be considered there, but they also know that it’s not a true surrogate in the sense that change on the biomarker directly corresponds to change in the clinical outcome. And because it’s not a perfect surrogate, I think we’re now looking for somewhat downstream markers that can more easily align with the clinical benefits and potentially be a surrogate and maybe be a surrogate across more mechanisms of action. Amyloid certainly works well for amyloid-targeted agents but there’s a lot of agents that are targeting other mechanisms and with those other mechanisms we may not impact amyloid very much at all or maybe not even at all. And in that case, we need something that’s more downstream that will be, again, connected to the clinical outcome, something such as tau.
So what I see happening is a lot of companies are designing their early studies with target engagement markers and with surrogate markers or markers that are more downstream, they want to see whether their treatment has an effect in early disease, like Phase II study, and then when they follow it into Phase III, they will carry forward some of those same biomarkers to try to connect what they’ve seen in the earlier stage with something in the later stage. And the hope is that this will save us time and that we’ll have fewer Phase IIs that look like they might have shown something and then fail in Phase III.
And so I think biomarkers are really important part of that equation and we’re getting more and more information about which biomarkers are the most relevant as either targeting engagement or surrogate markers in different stages of disease and it may be different which ones are better in the different stages of disease. So we advise a lot of companies along those lines and study design and data collection.
