Atrophied Lesion Volume as a Biomarker of Disability in Progressive Multiple Sclerosis

Robert Zivadinov, MD, PhD

Recently, a group of investigators conducted a post hoc analysis of the ORATORIO trial (NCT01194570) of ocrelizumab (Ocrevus; Genentech) in the treatment of individuals with primary progressive multiple sclerosis (PPMS), specifically assessing the accumulation of a novel biomarker, atrophied T2-lesion volume (aT2-LV).

Led by Robert Zivadinov, MD, PhD, professor of neurology and director, Translational Imaging Center, Clinical Translational Research Center, Buffalo Neuroimaging Analysis Center, and director, Buffalo Neuroimaging Analysis Center, the dataset included 732 patients with PPMS who were randomly assigned to ocrelizumab (n = 488) or placebo (n = 244). The data confirmed that accumulation of aT2-LV in patients receiving placebo (366.1 mm3 in 120 weeks) was consistent with prior literature, helping to validate the accuracy of this biomarker. Additionally, those treated with ocrelizumab had significantly slower accumulation of aT2-LV (319.4 mm3; P = .013). Conducting the analysis by accounting for scanner model, software, protocol changes, and additional covariates further confirmed these results (P = .029).

To find out more about the analysis, the potential of having a biomarker for PPMS such as aT2-LV, and what the clinician community should take away from these data, NeurologyLive inquired with Zivadanov for more details.

What were the observations you and your colleagues made in this analysis?

Robert Zivadinov, MD, PhD: The primary objective of the study was to see if this biomarker is reflective of progression to more disability and the progressive stages of MS, to see if ocrelizumab’s mechanism of action can slow down the accumulation of atrophied lesion volume. We found a significant difference over 24 months of the trial, with ocrelizumab being able to decrease the number of lesions that disappeared or shrunk into the CSF. It was 319 mm3 with ocrelizumab versus 366 mm3, which was significantly different. That was the primary end point of the study and given that the analysis was done in a hierarchical way, once this primary objective was fulfilled, we begin to look to the secondary objectives.

The secondary objective was to look at the difference or similarity in people who develop disability and those who did not develop disability. In placebo patients and ocrelizumab patients that did not develop disability during the study, we did not find differences in atrophied lesion volume. That seems logical because if we want to have a perfect biomarker of disease progression, then you should not find that this biomarker really correlates with disease progression in those who did not progress, right? That leads us to look at the people who progressed in the study. Progression was termed in different ways and defined in ORATORIO as follows: Those who had an EDSS [Expanded Disability Status Scale] score of less than 5.5 for disability had to advance 1 point on EDSS, and those with an EDSS score of more than 5.5 had to advance 0.5 in EDSS. Also, there was a definition with the 9-Hole Peg Test [9HPT] to measure dexterity of the upper hands and limbs, and Time 25-Foot Walk [T25FW] to measure the walking ability of the patient. If there was 20% deterioration over the 2 years on these outcomes, the patient was defined as progressive. There was also a total composite between the EDSS, 9-HPT, and T25FW to determine overall composite progression. When we began to work on that and compare the placebo with ocrelizumab, actually, the results were pretty stunning. Very surprising to me, in all instances, was that whether the progression was confirmed at 12 weeks or was confirmed at 24 weeks, ocrelizumab was highly superior in preventing atrophied lesion volume in the patients who progressed in EDSS composite, in 9-hole peg test composite, in walking composite, and in the overall composite. That was significant for all at 12 and 24 weeks.

The main message of this finding is atrophied lesion volume is really a good marker over such a short period of time—2 years in this study—to distinguish people who progress and people who do not progress, and to find such a treatment effect. The second important thing was that it clearly provides us with more understanding that drugs against B cell component of the disease, like ocrelizumab, may really change the neurodegenerative outcomes. In particular, the strongest result in the regression analysis was on the walking composite, with an odds ratio of about 3.2 between placebo and ocrelizumab. That tells you that why atrophied lesion volume and walking are so related. We really did not previously look at how walking correlates really with atrophied lesion volume, but the axons that are passing near the ventricles where centers for walking visibility and motor visibility are located—mostly between the cortex and the spinal cord. These are most affected by the shrinkage of the lesions because those lesions are shrinking the most around the ventricles where these fibers and axons pass. We believe and interpret these data, to mean that the shrinkage of the lesions around the ventricles is probably the most responsible for walking, and in this case, ocrelizumab was able to prevent that.

Are there any implications about the potential underlying biology of progressive MS related to what this biomarker has shown thus far?

This is a great question. One thing that, clearly, we need to understand better is what is leading to the lesions disappearing. The concept of the tissue being lost in MS—healthy or lesioned tissue—is pretty much 20 plus years old, around when I started my first research in MS and published my first brain atrophy studies. But to distinguish between the non-lesioned tissue and the contribution of the non-lesioned tissue disappearance or atrophy, versus the lesioned tissue or atrophy, and what's leading to the 2 processes, is extremely important. We believe that aging and other health-related processes can also lead to the loss of brain tissue. But clearly, the loss of the lesioned tissue might be driven, must be driven, with more specific immunological or immunity-related mechanisms.

We believe that there are 2 possibilities for that. One, that the lesion may undergo, or part of the lesions that are disappearing, may undergo much more severe damage. So microglial activation, or specifically neurodegeneration, etc, in those lesions may really lead to faster loss. But, when the lesion forms—whether it's enhancing lesion or a T2 lesion—it usually becomes, over time, a black hole. Then, once it's a black hole, it may degenerate and disappear. In this study, the disappearance over 2 years was probably much faster than the lesions can go into this trajectory of the various pathological substrates until they become CSF. At the moment, we don't that understand fully. We did one study that will hopefully be published soon in which will look at diffusion tensor imaging, or nonconventional MRI, to understand which part of those lesions are more prone to disappear into the CSF. And we measured so-called free-water fraction, or how much of the water these lesions had, and so those that had more water at the beginning at the baseline are more prone to become CSF at the end. But I would say that the mechanisms are not fully elucidated. We are planning a number of grants in that direction.

Also, there is potential for these biomarkers to become relevant in the clinical routine because when you move the scan and you say, “Okay, what I see on that scan is this,” you always look for a new or enlarging lesion. You almost never go, “Okay, it's a good thing if there are less lesions over time,” right? I think this will have to be a reeducation for the neurologists that when they see less lesions on scans, that may not be necessarily a good thing. It could really be a bad thing because some of these lesions have completely been destroyed and gone. As a matter of fact, that's how everything started a couple of years ago. I have seen thousands of MRI scans, and one thing that you always see in patients with MS is ventricles are enlarging and with time, the lesion volume is rising, but then at some point, probably after 20 years of the disease, you see a plateauing of the lesions. Then, if the patient is followed long enough, you begin to see that decrease. That's how we came to the discovery of this biomarker. We asked what is driving that lesion decrease and enlargement of the ventricles. Based on the MS hypothesis, you could say that there is something toxic in the CSF—chemotactic factors and other factors whether in the cortex or in the ventricles—and that this actually may be the most destructive for the nearby lesions. There have been studies that show that most myelin changes are happening closest to the CSF. So possibly, those lesions that are destroyed may also have contact with the CSF chemotactic and toxic factors that are creating this lesion disappearance and shrinkage.

What should the neurology community be looking for regarding this biomarker, as its clinical application is further explored?

The key outcome of this study is that new and enlarging lesions are not those who drive disability, really it is the lesions that are disappearing. That's a really dramatic change in the concept for what we believed over the last 30 years of measuring new and enlarging lesions in MS. I think that's very significant. If that marker is 3 to 5 times more accelerated in progressive patients, then we also need to begin looking on everyday scans. When we scan these patients in clinical routine, how many of these lesions are really going away? That may be a very negative sign of disease progression and may prompt physicians to prescribe more aggressive therapies in this case, like ocrelizumab. The third point is that the drugs that may act against neurodegeneration, as ocrelizumab has shown in this study, may be important to slow down disability progression through alternative pathways other than just those that are anti-inflammatory. It's very important to say that because at baseline there were no differences in T2 lesion volume between placebo and ocrelizumab patients, so what we observed during the study was truly a result of the treatment effect.

Is there anything else you’d like to add?

I think that there are a couple of possible important issues to discuss. One, clearly, is that other centers around the world should begin to use this biomarker. Certainly, the more multicenter studies, the better—although ORATORIO had more than 70 centers in the world and over 750 patients participating. That's the first multicenter proof-of-concept, but I think more studies are needed. Also, we need more automatic measures of atrophied lesion volume, clearly, to increase the reproducibility of such results. This is an initial step, and hopefully, this could be done in clinical routine and become a go-to biomarker in progressive MS.

Transcript edited for clarity.

REFERENCE
Zivadinov R, Pei J, Clayton D, et al. Evolution of lesions that shrink or disappear into cerebrospinal fluid (atrophied t2 lesion volume) in primary-progressive multiple sclerosis: Results from the phase III ORATORIO study. Presented at 2021 American Academy of Neurology Annual Meeting; April 17-22, 2021. Abstract P15 151