Escalating Therapy Specifically to Cut SUDEP Risk in an Otherwise Stable Patient
A single patient with occasional nocturnal breakthrough seizures who otherwise feels his epilepsy is well managed. The disagreement is whether SUDEP risk alone justifies pushing his regimen harder.
C.B., a 39-year-old man, has had drug-resistant focal epilepsy for eleven years and has been on lamotrigine and levetiracetam together for the past three, a combination that controls his daytime focal seizures completely and has reduced his nocturnal generalized tonic-clonic seizures to roughly one every six to eight weeks, almost always caught only by his partner noticing rumpled bedding or a bitten tongue the next morning. He works full time as a high school teacher, coaches soccer in the fall, and describes his current life, by his own account, as genuinely good — seizures that used to derail entire weeks now barely register against everything else going on.
His neurologist raised escalating therapy at today's visit for a reason unrelated to how C.B. himself experiences his seizure control: generalized tonic-clonic seizure frequency is the strongest modifiable risk factor for SUDEP identified in the epidemiologic literature. Hesdorffer et al. (2012), pooling case-control studies for the ILAE mortality subcommission, found it was GTC frequency rather than any particular drug that drove the risk; the AAN/AES guideline (Harden et al., 2017) puts three or more GTCs a year at roughly a fifteen-fold increase, and population-based case-control data find nocturnal convulsions specifically carry an odds ratio in that same range. A seizure every six to eight weeks sounds, and functionally is, well controlled — but from the specific angle of SUDEP risk, any ongoing nocturnal GTC frequency above zero carries real, quantifiable risk that doesn't track neatly with how well-managed a patient otherwise feels.
C.B. had never heard the term SUDEP before today, and said so directly — a fact that itself became part of the conversation, since escalating a regimen he currently experiences as working, in response to a risk he didn't know existed until this visit, is a genuinely different conversation than adjusting a regimen a patient already feels is failing him.
Naming a risk he'd never heard of
I think we should escalate — add a third agent or optimize dosing further, specifically targeting the nocturnal GTCs rather than his overall seizure burden. The epidemiologic data are consistent on this. Hesdorffer et al. (2012), pooling case-control studies for the ILAE mortality subcommission, found that generalized tonic-clonic seizure frequency — not which drug a patient is on — is what drives SUDEP risk. The AAN/AES practice guideline (Harden et al., 2017) puts three or more GTCs a year at roughly a fifteen-fold increased risk, and Swedish population-based case-control data find nocturnal convulsions specifically carry an odds ratio in that same range. Any frequency above zero carries real risk regardless of how manageable it feels day to day.
I don't dispute the epidemiology, and I think he deserves to know it — he just did, for the first time, today. But he's also functioning well on a regimen that took years to reach, and a third AED carries its own real cost: additional side-effect burden, drug interactions, and for a lot of patients, a regimen that starts to feel like it's chasing a number rather than treating a person who feels genuinely okay.
Framing this purely as 'any frequency above zero carries risk' is true but incomplete — escalating therapy also carries risk, in tolerability and adherence, and a regimen he abandons because it's become too heavy protects him from nothing.
I don't think this resolves into a single right answer today, and I'd be honest with him about that rather than presenting escalation as an obvious next step. Give him the actual numbers — Harden and colleagues put the absolute risk for patients with frequent GTCs at up to 18 deaths per 1,000 patient-years, which is real and also not the same thing as a coin flip — alongside what's known about the tolerability trade of adding a third agent, and let this be a genuinely shared decision rather than one made for him in the room.
If he chooses to escalate, a low-dose clobazam add-on at night specifically, timed to his sleep rather than dosed like a daytime AED, is a reasonable next step that doesn't require abandoning a regimen that's otherwise working. If he chooses not to, a seizure-detection device for nighttime monitoring is a real, lower-burden alternative that addresses part of the risk without changing his medications at all.
Agreed: C.B. was given the actual SUDEP-risk data explicitly, told plainly this was new information as of today's visit, and offered both a pharmacologic (nocturnal clobazam) and non-pharmacologic (seizure-detection device) option rather than a single recommended path. He asked for a week to think it over and discuss it with his partner before deciding.
Not agreed, and explicitly left unresolved rather than defaulted either direction: whether a physician's obligation, once aware of a strongest-known modifiable risk factor, is to recommend escalation actively or to present the full picture and let a well-informed patient decide for himself. Both positions were named directly rather than one being treated as simply correct.