Multiple sclerosis clinical trials typically use the reduction in annualized relapse rates as their primary outcome. A paradox has emerged in these studies, as approved high-efficacy medicines typically prevent more than 90 percent of new lesions but only around 45 percent of relapses. In this project, we reanalyzed a recent set of phase 3 clinical trials in multiple sclerosis, using a new relapse definition that includes an MRI metric. In this way, we have tried to remove noise from this important clinical trial outcome and redefine relapse in a way that more truly reflects the efficacy of MS treatments.
Hi, I'm Steven Krieger, a professor of Neurology at the icon School of Medicine at Mount Sinai in New York. Uh and an MS specialist here at our Corine Goldsmith Dickinson Center for multiple sclerosis. Uh And I'm gonna give you a quick uh highlight of a poster that uh I presented along with my colleague, Enrique Alvarez from the University of Colorado Denver um at the American Academy of Neurology Meeting uh held in spring 2024 in Denver, Colorado. And this is a project we called MS relapse redefined or redefining relapse learning how to distinguish true relapses from pseudo exacerbations. And we used for this the ultimate one in two trials of tux out. So to begin, why do we need to think about pseudo exacerbations and real exacerbations? And the annualized relapse rate outcome has been the main outcome in relapsing MS studies for many years. And there's a little bit of a paradox, a little bit of a conundrum. Um Most of our highly efficacious therapies studied in modern relapsing MS trials like the B cell depleter, including tux IAB reduce the annualized relapse rate by around 50% but they reduced the number of new lesions and enhancing lesions by well over 90% sometimes closer to 97%. And in truth and new enhancing lesion and a relapse should mean the same thing. So why would these medicines reduce 50% of relapses? But the incredible vast majority of new lesions. And we think that that paradox comes from the presence of pseudo exacerbations or fluctuations in our clinical trials. So we wanted to find ways of taking out those pseudo exacerbations. So if we look here the results of the original ultimate one and two trials, this was tera flutamide versus Luima oxime reduced the annualized relapse rate by 54%. Great. What we did in an earlier version of this project that we presented six weeks earlier at the Act TRMS conference is we applied a new relapse criteria that defined a relapse as an event that had a new T two lesion found on MRI scan any time subsequently during the study, this is based on the idea that every lesion is forever, every two T two lesion is forever. And so if there really was disease activity, if a true relapse had occurred, we confirmed that by the presence of a new T two lesion scene subsequently, when we defined relapses that way, Tu IAB reduced the number of relapses versus tide by 87.5%. Much closer to what we know is its sort of true efficacy. From the perspective of preventing new lesions. So with that in mind, what we looked for in the current study was to say we understand that using an MRI supported relapse definition, a new lesion definition Oxin produces 88% of relapses. Could we define relapses better clinically to try to weed out pseudo exacerbations? So we created a stepwise more stringent clinical definition of relapse. Here's just a quick summary of it. We look to see if we define relapses, not by a half point increase in E DS S but a one point change, not a one point functional system change, but a two point change. And then whether the new E DS S or functional system score at the time of relapse was a new maximum for that patient, a new sort of high water mark for disability, whether that would better define relapses to eliminate these fluctuations and pseudo exacerbations, which we know happened. We also look to see whether we could make a more stringent infection free criteria for relapse. So in every clinical trial, if a suspected relapse is happening, they have to rule out infection. We widened out the time window to say a new relapse. A true relapse couldn't occur if there was an infection within a two month window. But sometimes ad infection declares itself after the relapse has already occurred. So we widened out that window and then finally, we added back in our MRI criteria as an additive measure. So here's what we found again in the original pivotal trial of Luima versus Teruto. With the original definition, the protocol definition tuum reduced relapses by 54%. When we used our clinical redefinition. This new definition of one point change, higher functional system score change maximal ever for that patient and ruling out infections within a couple of months. The reduction in annualized relapse rate increased to 60%. The modest change, it didn't totally get rid of the fluctuations in pseudo exacerbations which we feel kind of pollute the relapse data when we put back in our MRI supported relapse definition as previously presented. That's where we saw the 87.5% reduction in MRI confirmed relapses versus uh tera flutamide. So if you think about it this way, how many confirmed relapses were there on tux IAB versus pseudo exacerbation? And how many confirmed relapses were there on Teruto? What you see is the great majority of the relapses that occurred on oxime didn't meet this new MRI definition. They were pseudo exacerbation. A uh different percentage, a much smaller percentage of pseudo exacerbations happen in tera flutamide. So you see using the MRI definition, there were only 22 confirmed relapses in the entire trial in the tux IAB treated patients that has a little bit of face validity to it. In so far as we don't think of people on B cell medicines as having a lot of relapses here if you're interested is the breakdown of how much each criteria helped to remove pseudo exacerbation. So here you see the incremental contribution of those different clinical definitions. And the big point at the end is the T two lesion MRI supported relapse definition. That's really where most of these pseudo exacerbations or fluctuations get cut out. You see that for Go Latu IAB and now you see it here for tera flutamide with more relapses remaining confirmed, fewer of them being weeded out as pseudo exacerbations. And that's why we have a much higher annualized relapse rate reduction for Laxima using these new definitions. So in summary, our objective here, we are trying to develop both an MRI and a clinical methodology of eliminating pseudo exacerbations from the data on annualized relapse rate in pivotal trials. We saw that uh using the MRI definition, there's a reduction of 87.5%. But using clinical more stringent definitions, we really only got to a 60 66% reduction in relapses. What this meant to us was that while an MRI supported relapse definition and really improves the stringency of the determination of true relapses and the the determination of true clinical efficacy. Um Our efforts to optimize the relapse definition on a clinical basis wasn't good enough. Um which is interesting, we really need an objective marker for true relapse for inflammatory disease activity. It's very, very hard to do it on a clinical basis alone. Some limitations here, we don't have spinal cord imaging. So we can't look at spinal cord relapses, which of course are probably uh the most indicative of things that cause new findings on the E DS S. But the improved signal to noise ratio of using an MRI supported relapse definition might have some real implications for how we do relapsing MS trials. If we can get a closer sense of the true efficacy of the drug, we might be able to do that with smaller trials, we might be able to do it with faster trials that have implications for power uh analysis and sample size. And we could allow for a more expedient uh conduction of relapsing MS trials in the future using a redefined relapse definition. So this is an ongoing project and uh Doctor Alvarez and I expect to continue to take it forward and I appreciate you watching. Thank you.
