Ganaxolone for CDKL5 Deficiency Disorder: How Early to Add the Newest Agent
A single pediatric patient with confirmed CDKL5 deficiency disorder, already on two antiepileptic drugs. The newest approved agent for her exact diagnosis is available — the disagreement is about how soon to use it.
E.V., a 4-year-old girl, was genetically confirmed with CDKL5 deficiency disorder at eleven months of age after presenting with seizures in the first weeks of life — among the earliest-onset epilepsies clinicians see, and one that comes with substantial developmental delay independent of seizure control itself. Her parents have become fluent in a disease most pediatricians have never encountered, and track her seizure clusters, feeding, and sleep in a shared spreadsheet they update daily. She currently has ten to fifteen brief tonic seizures most days despite valproate and clobazam, both titrated to their practical ceiling given her weight and tolerability.
Ganaxolone, a neurosteroid that works as a positive allosteric modulator at both synaptic and extrasynaptic GABA-A receptors, is the first and only agent specifically approved for seizures associated with CDKL5 deficiency disorder — her diagnosis exactly, not an adjacent or related condition. Its approval trial, Marigold (Knupp et al., 2022), randomized 101 children and young adults and found a median 30.7% reduction in 28-day major motor seizure frequency against 6.9% on placebo — a genuine result in a population where most agents show only modest effect. What the trial couldn't yet establish, because the drug is simply too new, is how it performs over years of use, what its long-term tolerability looks like in a child this young, or how it behaves layered onto her existing two-drug regimen rather than tested largely as monotherapy addition against placebo.
Her parents brought a printout of the approval trial's results to today's visit and asked directly why she isn't on it yet, given that it's approved for exactly her diagnosis and her current seizure burden is, by anyone's assessment, still substantial.
With the trial printout on the table
Add ganaxolone now. This is the one agent developed and approved specifically for her exact diagnosis. The Marigold trial (Knupp et al., 2022) randomized 101 children and young adults with CDKL5 deficiency disorder and found a median 30.7% reduction in 28-day major motor seizure frequency on ganaxolone against 6.9% on placebo — the first placebo-controlled evidence of efficacy in CDD-associated seizures. She's at ten to fifteen seizures a day on two drugs already at their practical ceiling. Waiting for a diagnosis-matched, evidence-backed option to accumulate more years of real-world data while she keeps having this many seizures is a real cost too, not a neutral default.
You're right that the diagnosis match and the trial result are both genuinely strong, and I'm not arguing to withhold it indefinitely. What gives me pause is the shape of what Marigold actually measured: ganaxolone added onto an existing regimen across a seventeen-week double-blind window, not years of layered three-drug use in a four-year-old. Somnolence was its most common adverse event. We don't yet know how that sedative burden compounds with clobazam specifically over years, or what her long-term developmental trajectory looks like on three GABAergic-acting agents together.
Framing this purely as 'evidence-backed option versus continued seizures' skips the part of the trial evidence that's actually thin — not whether ganaxolone reduces seizures, which Marigold showed it does, but what three-drug, years-long exposure in a child this young actually looks like, which a seventeen-week trial wasn't built to show.
Both of those are right, and there's a way to add the drug without pretending the long-term question is already answered. Start ganaxolone now, given her actual current seizure burden and the strength of the diagnosis-specific trial data — but build in scheduled, explicit reassessment of the clobazam dose specifically, since two GABAergic-acting sedating agents together is exactly where a dose-reduction opportunity is most likely to appear if ganaxolone works as well in her as it did in the trial population.
That treats her current seizure burden as the real, present cost it is, without simply stacking a third drug on top of two others and calling the regimen finished.
Agreed: ganaxolone started at a weight-based starting dose with planned titration, and a specific follow-up visit scheduled at eight weeks to reassess her clobazam dose in light of her actual response and any new sedation.
Not agreed: how aggressively to reduce clobazam if ganaxolone does bring meaningful seizure reduction — the Clinical Pharmacologist wanted a conservative, slow taper given the developmental stakes of getting a benzodiazepine reduction wrong in a child this young; the Pediatric Neurologist preferred a more decisive reduction once real improvement is confirmed, to limit total sedating-drug burden sooner rather than later. Left for the eight-week visit to decide with real data in hand.