This is a 2-part interview. For part 1, click here.
A recently published retrospective study examined the time-domain and frequency-domain components of patients admitted for video-electroencephalogram (EEG) monitoring between 2003 and 2014, and subsequently died of SUDEP (sudden unexpected death in epilepsy). Considered the largest SUDEP biomarker study to date, which comprised of 31 SUDEP cases and 56 controls, investigators identified an association between heart rate variability (HRV) and SUDEP.
The data also showed that reduced short-term low-frequency power (LFP), a validated biomarker for sudden death, was also associated with SUDEP. Normalized LFP during wakefulness was found to be lower in SUDEP cases (median, 42.5 [interquartile range (IQR), 32.6-52.6) than epilepsy controls (median, 55.5 [IQR, 40.7-68.9]; P = .015; critical value = .025). Furthermore, a negative correlation between LFP and latency to SUDEP was observed, where each 1% incremental reduction in normalized LFP conferred to a 2.7% decrease in the latency to SUDEP (95% CI, 0.95-0.995; P = .017; critical value = .025).
In the second half of our conversation with NeurologyLive®, Devinsky, director of the Comprehensive Epilepsy Center at NYU Langone, provided insight on how these findings impact the SUDEP research space going forward. More specifically, he touched on whether HRV will have a role in future seizure-detecting devices, the ways clinicians are actively trying to learn about SUDEP, and whether the overall awareness for the event needs improvement.
NeurologyLive®: What’s the likelihood that these findings help the advancement of devices or modalities that may help detect or signal SUDEP?
Orrin Devinsky, MD: Other than what we’ve done with heart rate variability, the other major marker is when someone who’s married or has a roommate has noticeable convulsions in their sleep or during the daytime. They’re aware that they’re going to be at high-risk. There are watches that can be worn, motion detectors under the mattress, or other movement-detecting infrared software during sleep, that detect seizures and are already in widespread use. We don’t have clinical trial data to prove that they will reduce the risk of SUDEP, but we believe they will. It makes perfect sense; we know that a big risk factor for SUDEP is having a nighttime seizure and not being supervised. There have been multiple studies that have looked at individuals living in a group home, some of which do extremely careful monitoring and have seen the rate of SUDEP go down with the intensity.
One of those studies in England, they have the young adults go home on some weekends where they aren’t monitored, and they typically sleep in their own bedroom while their parent sleeps down the hall. God forbid a seizure happens at 4AM when the parents are sleeping and they’re not aware of it. Those individuals have a much higher rate of SUDEP. We have good anecdotal data. If you go through the series of steps like we’ve done, something like 80% to 90% of SUDEP cases are not witnessed. What that tells us is that there’s a higher percentage of seizures overall. If I’m a patient with epilepsy and I have 10 seizures and only 1 is witnessed, and I wake up with a bit tongue or blood on my pillow, I know I’m going to be at high-risk. Whereas if I was married and I roll facedown during my seizure, my wife can roll me on my side and ask me if I’m doing okay. That might be live saving.
Will clinicians have to monitor heart rate variability of their patients, or is this more of an observation?
Right now, it’s not something I would recommend for my colleagues in general neurology practice, but it’s something to be aware of, and that it’s coming down the pike. Hopefully, within 2 years, our group and other groups will repeat this study in a new cohort. If its validated, then it raises up the ladder and then we can try to figure out how we can incorporate this into clinical practice.
How is the clinician/research community actively trying to improve SUDEP risk?
One project that I’m excited about that hasn’t quite gotten off the ground yet is another paper we published in Neurology using a Bayesian model to do an individualized SUDEP risk. We know, for example if you have frequent tonic-clonic seizures or nocturnal seizures, your risk is much higher than people who don’t have that. But what we can’t do is say, “well you’re a 43-year-old man, you have generalized epilepsy, you’re on these 2 medications, you miss your medications once a month,” and then throw that into a model and come back and say “you [patient] currently have a 1 in 800 risk per year of SUDEP. Now if you could avoid missing your meds 2 times a month, that will go to a 1 in 1500 chance or whatever the number would be.” And use that as both a counseling tool and an education tool to help let people understand how they can reduce their risk, because I don’t think people understand what their risk is right now. They don’t understand exactly how improved adherence can help. If you’re a college student, getting intoxicated with alcohol, forgetting to take your meds, and sleep depriving yourself, you don’t know how risky that could be.
Is there anything else you’d like to mention about your results?
We’re making progress. We need to pivot as a community from studying SUDEP, to preventing it. In the last decade, we’ve seen an explosion of research and publications on SUDEP, and Neurology has been a home for many of them, but now we need to start to move forward on changing. The lowest hanging fruit is educating our patients. Every neurologist should discuss SUDEP with parents, adult patients, and adolescents, because they need to understand that if they take their medications, they don’t drink alcohol, and get good sleep, they’re going to lower the risk of seizures, which can prevent head trauma, prevent dislocated shoulders and vertebral fractures, downing, and above all, prevent SUDEP. That’s what we can do right now, and I think we’ll be able to do much more.
What was your reaction to the recent findings form Nascimento et al? Are you just as stunned to see the SUDEP education levels so low?
No, I wasn’t [stunned]. We’ve surveyed neurologists, epileptologists, patients, and it’s terrible. Many more know about it because we as a community have risen up, but the patient community has been the leader here, not the neurology community. Unfortunately, it remains something where the basic information is not known. When I wrote a review article for the New England Journal of Medicine on the incidence of SUDEP, the editor at the time emailed me and said, “this must be wrong.” I sent him a few articles and said, “no, I don’t think this is wrong, I just don’t think anybody has any idea.” That was the editor of the New England Journal of Medicine, who in fairness was not a neurologist. We should know, he shouldn’t know.
Transcript edited for clarity.
Sivathamboo S, Friedman D, Laze J, et al. Association of short-term heart rate variability and sudden unexpected death in epilepsy. Neurology. Published online October 14, 2021. doi:10.1212/WNL.0000000000012946