MRI already plays a central role when we prescribe anti-amyloid therapies. Every patient basically undergoes baseline imaging before starting treatment, mainly to establish diagnosis and to evaluate his eligibility for treatment, but then also for monitoring safety issues, specifically ARIA. I think that maybe one of the main limitations today is that MRI information is still underutilized as a tool to predict treatment response and to assess risk...
MRI already plays a central role when we prescribe anti-amyloid therapies. Every patient basically undergoes baseline imaging before starting treatment, mainly to establish diagnosis and to evaluate his eligibility for treatment, but then also for monitoring safety issues, specifically ARIA. I think that maybe one of the main limitations today is that MRI information is still underutilized as a tool to predict treatment response and to assess risk. In most clinical settings, MRI interpretation is still mostly qualitative. Radiologists visually assess the presence of microhemorrhages and superficial siderosis and white matter disease or any other abnormalities. And this is, of course, clinically, this approach is essential, but it may miss more subtle structural patterns that could maybe help us better understand disease stage and to predict outcomes. And I’d say that another limitation is that many of the predictors that we currently rely on come from clinical trial populations, which are usually more homogeneous than real-world patients. Because in everyday clinical practice, patients are often more variable in age, in comorbidities, in vascular disease and the underlying neurodegeneration that they actually have. And because of this, the imaging predictors identified in trials may not fully capture the complexity of patients we treat in routine care. And a third important issue is that imaging is often evaluated in isolation, while Alzheimer’s disease is a complex biological process. So ideally, imaging predictors should be interpreted together with other biological information, with genotype, with plasma or CSF biomarkers, and of course, clinical characteristics. So for these reasons, I think that there is a growing interest in quantitative and AI-derived imaging approaches, which can extract much more information from routine MRI scans. And these are tools that allow us to measure brain volumes, regional atrophy, and of course, microhemorrhage burden in a more standardized and reproducible way. So ultimately, the goal is to move from simply asking, is this patient eligible for treatment? To asking, how likely is this patient to benefit? And what is their specific risk profile? So I think that the main limitation today is not the MRI itself, but rather how we use it. And with more quantitative approaches and integration with other biomarkers, imaging could become an important part of personalized treatment decisions in Alzheimer’s disease.
This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.
