New Brain Signal Technique Could Lead to Personalized Medicine in MS

Researchers of the Human Brain Project have developed a new methodology to calculate the delay of signal propagations in brains of patients suffering from multiple sclerosis (MS), which could lead to more personalized medicine.

Pierpaolo Sorrentino, M.D., a neurology specialist and senior post-doc at the Human Brain Project, Institute for Systems Neuroscience in Marseille, France, and colleagues published the results recently in the Journal of Neuroscience.

Pierpaolo Sorrentino

Managed Healthcare Executive® conducted an email interview with Sorrentino about how the new technique works and how it could lead to better MS treatments in the future.

Please explain how the new methodology works.

The method we devised borrows techniques from statistical mechanics, a branch of physics, and uses these to merge structural and functional brain data in a principled and informative way. In short, we measure, across any two brain regions, the length of the white-matter tracts connecting them. This measurement is obtained using magnetic resonance.

Then, in the same subjects, we also record brain activity noninvasively, using electro-magnetoencephalography. These techniques provide information about brain activity, with very good temporal resolution.

We then focused on how messages travel across the brain. In short, it is possible to observe in the data that, once a region becomes active, such activations spread across the brain in a very complex, yet highly regulated and patterned fashion. These patterns appear spontaneously in the data, and they can be measured noninvasively. Hence, it is possible to track how one active region sequentially recruits more regions and, in particular, it is possible to measure how long it takes any two regions to activate sequentially.

What are the results of your research? Is this research unique?

While these results are preliminary and should be confirmed in larger cohorts, we show that, in subjects affected by MS, functional delays are longer than in healthy controls, and more so across edges that have been directly attacked by the disease.

Our research is the first attempt to estimate functional delays noninvasively on the whole brain.

The current manuscript falls in the context of personalized medicine. In short, this branch of research is concerned with developing tools to predict the effect of a specific treatment on a specific patient. To be able to predict such effects, one must build appropriate models of the brain that accommodate how everyone’s brain is different and unique.

Advanced large-scale modeling so far considers the individual large-scale wiring of the brain, which is a piece of information typically derived from magnetic resonance. This means including in the models how brain regions are connected to each other in an individual (i.e. the structure of the white matter bundles connecting gray matter areas).

How do you expect the research to lead to better future treatment of MS?

Now we wish to build even more precise models, which might not be limited to including the wiring scheme of everyone’s brain, but also how fast messages between brain regions can travel across white matter bundles. This is specifically relevant in MS, where we know that the velocity at which messages travel across the brain can be reduced due to lesions in myelin.

We hope that this information could make subject-specific, large-scale brain modelling available to MS.

In particular, we hope to help shed light on the phenomenon whereby the amount of structural damage does not predict well individual clinical disability, sometimes referred to as the “clinical-radiological paradox”.

We hope our approach can shed light on why specific lesions have a much higher clinical impact on a given subject. To do that, mechanistic large-scale have the potential to test potential mechanisms. Allowing to include information about the individual functional damage in the model might allow using such models in MS.

We have already started to include this information in personalized brain models, and to explore if including subject-specific delays can help predict individual clinical outcomes. Large-scale modeling has been used to simulate the effect of surgical resections in individual epileptic patients.

By adding the delays, we are now adjusting large-scale models to simulate subject-specific, large-scale activity in individual MS patients. We need large, multicenter, multimodal cohorts to check how generalizable our results are.