An Unmet Need in Epilepsy? Diagnostic/Prognostic Biomarkers

Michael R. Sperling, MD: I would say that one of the problems we have in epilepsy is that if you come to me with a seizure and I give you a pill to take, and then you say to me, “Well, will this work?” I tell the patient, “You’ll tell me if it works.” That’s the equivalent of somebody coming with heart disease and I say take this pill and if you don’t have a heart attack, I guess it’s worked. It’s very unsatisfying. We don’t even have something like blood pressure or blood glucose to monitor along the way. There’s really no good surrogate. Are there any prognostic biomarkers or diagnostic biomarkers that you tend to use or think might be promising at some point in the future that might help guide our advice to our patients and guide your choice of therapy?

Trevor J. Resnick, MD: I think we’ve discussed a couple of those already. I think some of them also are biomarkers that come with diagnostic tests. Just as an example, we were talking about genetic mutations. One genetic mutation, SCN2A, is specifically sensitive to treatment with phenytoin, which is a drug we never think of anymore, especially in pediatrics. So a lot of these additional tests are making us look back and think about specifically focused treatments that are not currently part of our usual thinking.

Michael R. Sperling, MD: And the old EEG [electroencephalogram] is actually not a bad biomarker, and the MRI [magnetic resonance imaging], right?

Kathryn A. Davis, MD, MS, FAES: Right. I can comment on that. The MRI has advanced significantly over the last several decades. And for a patient particularly who is not responding to medications, and if an initial MRI is unrevealing, getting an epilepsy protocol 3 Tesla [T] MRI can be instrumental with fine cuts through the hippocampi to improving your diagnostic accuracy. What’s happened in very recent time is the advent of higher field MRI. 7T MRI is now approved by the FDA for clinical use, and centers are starting to use that in patients who have nondiagnostic 3 Tesla MRIs, and that I think will lead to improvement in finding focal lesions. And we know if we can find them, that the treatment options surgically are much improved. And then there are other types of MRI and imaging tests that are being studied. We’re applying machine learning to these different modalities, and I’m hopeful that in the coming years we’ll have many more biomarkers at our disposal.

Michael R. Sperling, MD: Eric, did you want to add any comments to that?

Jesus E. Pina-Garza, MD: Just briefly to add that in some of the patients, the genetic testing or the markers that we have may transform a situation where you have an unknown etiology that seems to be focal, and we’re trying to find that microscopic heterotopia after surgery. And we may be able to diagnose that epileptic brain as a genetic etiology, which makes it less attractive for a surgery regarding outcome. So it is an area we need to explore more, and at least it allows you to select the best possible candidates for surgery.