PNES After Years of Presumed Epilepsy: How Fast to Taper the AEDs
A single patient newly diagnosed with psychogenic nonepileptic seizures after nine years of presumed epilepsy. The diagnosis is settled — how fast to taper the drugs she never needed is not.
S.R., a 31-year-old woman, was diagnosed with epilepsy at twenty-two after a series of events described as convulsions during an especially difficult year that included her parents' divorce and a sexual assault she has since discussed in therapy, and has been on levetiracetam and lamotrigine together for nine years, with “breakthrough seizures” roughly monthly despite escalating doses of both drugs over that time. She left a nursing degree program in her second year, in part because faculty grew uncomfortable supervising a student with an unpredictable seizure disorder in clinical rotations, and has worked retail jobs since — a career path she has told her new neurologist she still grieves. A five-day inpatient video-EEG admission, arranged after her third neurologist in nine years finally questioned whether her events matched a genuinely epileptic pattern, captured four of her typical events — all without any corresponding EEG change, a clean, unambiguous finding consistent with psychogenic nonepileptic seizures rather than epilepsy.
She has started psychotherapy with a clinician experienced in PNES specifically, on her new neurologist's referral, and has had one further event since the admission, four days ago — also captured on her home seizure log, though obviously without video-EEG confirmation this time. The diagnosis itself isn't in question anymore; both her new neurologist and her therapist agree PNES, not epilepsy, has always been the correct diagnosis. What remains genuinely unresolved is what happens to the two AEDs she's been on for nine years: the reported prevalence of comorbid epilepsy among PNES patients runs from about 10% to past 30% depending on setting and how strictly epilepsy is defined, with the strictest-criteria studies such as Benbadis et al. (2001) landing near the low end, and abrupt discontinuation of either levetiracetam or lamotrigine carries its own real withdrawal-seizure risk, entirely separate from whether her events are epileptic.
With the diagnosis settled, the taper still open
I'd taper both AEDs fairly promptly, over a few months rather than a year. She's been on two drugs for nine years for a condition she never had, and continuing an unnecessary regimen carries its own real costs — side-effect burden, cognitive fog, the ongoing message that she's a person managing a seizure disorder when she isn't. A relatively brisk taper reinforces the diagnosis for her the way a slow, ambivalent one wouldn't.
I understand the reinforcement argument, and I don't think it's wrong on its own terms. But comorbid epilepsy in PNES patients is common enough to matter here, and I'd be honest that the estimates are genuinely unsettled rather than one clean number — reported rates run from about 10% up past 30% depending on the setting and how strictly epilepsy is defined, and studies using the strictest criteria, like Benbadis et al. (2001), land at the low end near 10%. Even the low end isn't a rounding error, and her single event four days after discharge, without video-EEG confirmation this time, is genuinely ambiguous. It could be a PNES event, or it could be a real seizure we no longer have the tools in the room to distinguish from one.
Treating a brisk taper as primarily therapeutic — reinforcing the diagnosis — sidelines that withdrawal seizures are a real, physiologically distinct risk from anything about how she experiences or internalizes her diagnosis; her brain doesn't know the taper is meant to be a message.
There's a way to honor both concerns without picking a fixed pace in the abstract. Taper gradually — months, with the exact pace adjusted based on how the recent ambiguous event and any future events actually resolve — rather than committing today to either a brisk or a slow schedule. That keeps withdrawal-seizure risk genuinely low while still moving in the direction the diagnosis supports.
And the ambiguous event four days ago deserves a real answer, not a shrug: outpatient ambulatory EEG monitoring around her next expected event, if one occurs during the taper, would tell us directly whether recent events are following the vEEG-confirmed PNES pattern or something the video-EEG unit simply hasn't captured yet. With one caveat that has to be stated up front, because it shapes how any result gets read: a seizure captured during an active taper may be a withdrawal seizure the taper itself provoked rather than evidence of underlying epilepsy she has had all along. Those look identical on the recording and mean opposite things for her diagnosis. It argues for the slower end of the taper range and for monitoring during a stable interval rather than mid-reduction wherever that's possible.
Agreed: a gradual, multi-month taper of both AEDs, with pace explicitly reassessed at each new event rather than fixed today, and outpatient ambulatory EEG arranged to capture and clarify her next event if one occurs.
Not agreed: whether the four-day-post-discharge event should be treated as more likely PNES (matching the confirmed pattern) or genuinely uncertain until directly captured — the Neurologist leaned toward the former given the strength of the vEEG finding overall; the Psychiatrist held it as genuinely open. Both positions were recorded rather than one being adopted as the team's answer.