New-Onset Seizure on High-Dose Carbidopa/Levodopa: The March 2026 Vitamin B6 Warning in Practice
A new-onset seizure in a patient on high-dose carbidopa/levodopa, matching a pattern the FDA flagged in a March 2026 safety warning.
Harold T., 74, worked the loading dock and then supervised it for the last decade of a thirty-year career with the postal service before his Parkinson’s disease, now fifteen years in, made standing through a full shift impossible. He lives with his daughter, who found him on the kitchen floor two days ago, one side of his face and arm jerking rhythmically for what she estimates was under a minute before he lost consciousness entirely and the jerking spread to both sides — a focal seizure with secondary generalization, by the emergency department’s read of her description and his exam. He has never had a seizure before, has no history of head injury, alcohol use, or stroke, and outpatient imaging from two years ago, done for a fall-risk workup, showed no structural lesion. What he does have is fifteen years of progressively intensified levodopa therapy: his current regimen, extended-release carbidopa/levodopa dosed to control fluctuations oral options alone can no longer fully smooth, totals a levodopa-equivalent dose above 1,000mg daily — well past what most patients started on the drug decades ago ever needed.
That detail matters more than it would have eighteen months ago. On March 20, 2026, the FDA issued a safety communication requiring a new label warning across every carbidopa/levodopa product — oral, enteral, and subcutaneous infusion formulations alike — after identifying fourteen cases of seizures tied to vitamin B6 deficiency in patients on the drug — thirteen reported to FDA’s adverse event system, one from the published literature — every one of them at a daily levodopa dose above 1,000mg, the same threshold Harold’s own regimen crosses. The seizures in that case series were typically focal at onset before generalizing, the same pattern his daughter described, and all nine of the patients who received vitamin B6 had their seizures resolve, the majority of them only after multiple antiseizure medications had already failed. Some of the reported cases progressed to status epilepticus before the deficiency was identified, which is why the agency framed this as needing rapid recognition rather than watchful observation. The mechanism runs through the drug’s own core pharmacology rather than around it: levodopa’s peripheral conversion to dopamine consumes vitamin B6 as a cofactor, and carbidopa itself independently binds and inactivates the vitamin’s active form, so the same combination that makes high-dose levodopa tolerable in the first place is also what depletes the nutrient whose absence, at the extreme, can provoke exactly the kind of seizure he had. His actual vitamin B6 level has not yet been checked.
Admission, day one
The FDA's March 2026 communication reviewed fourteen cases of seizures tied to vitamin B6 deficiency in patients on carbidopa/levodopa — every one at a daily dose above 1,000mg, seizures typically focal-onset before generalizing, and every patient who received B6 resolving once it was replaced, mostly after standard antiseizure drugs alone hadn’t worked. Two of the fourteen patients died, both with documented deficiency and poorly controlled seizures. Harold matches that pattern on every count I can check right now — the dose, the seizure semiology, the absence of any other identified cause. I’d start supplementation today rather than wait.
The pattern match is real, and I’m not arguing his B6 shouldn’t be replaced — just that we shouldn’t dose empirically before confirming it. High-dose pyridoxine has its own chronic toxicity, a sensory peripheral neuropathy, and there’s an older, real pharmacologic history here worth remembering: before carbidopa co-administration was standard, pyridoxine supplementation was known to blunt levodopa’s own central effect by accelerating peripheral decarboxylation.
Where I’d push back is on treating the level as optional. Once we start repleting, his B6 result stops being interpretable, and we lose the one piece of evidence that tells us whether this was the mechanism at all. Draw it before the first dose — but I’m not asking anyone to hold the dose waiting on it. My caution is about the size and duration of what we give, not about whether he gets it today.
Then we don’t have a real disagreement about sequencing — we have one about dose. Draw the level, then give the first dose of pyridoxine on the way out of the room. What I won’t do is hold repletion for a turnaround time, even a fast one: the agency’s own review found some of these cases progressing to status epilepticus, and found that these seizures characteristically don’t answer to standard antiseizure drugs. A patient sitting at neurological baseline between focal seizures is exactly the patient who looks stable until he isn’t. The neuropathy risk you’re describing comes from sustained high-dose pyridoxine over months, not from starting replacement dosing tonight.
Agreed: B6 level drawn STAT and repletion started the same visit without waiting for the result, levetiracetam continued as seizure precaution while the B6 question is settled rather than in place of it, and his PD regimen left unchanged pending the full picture.
Not agreed: how long repletion should run and at what dose once the level returns — conservative and closely monitored, given both the sensory-neuropathy risk of sustained high-dose pyridoxine and the older pyridoxine-antagonism history, or sustained more aggressively given how closely his case matches the FDA's own reported pattern. Neither physician treated starting it tonight as the contested part; what stays open is the regimen after that, left for the outpatient neurology team once the result and his repeat level are both in hand.