Rare Disease · In-Market · Updated August 2026
Every Dravet-specific agent is an add-on to a generic backbone. Nothing here is prescribed as monotherapy, so a switching-share model describes a market that does not exist.
Dravet syndrome is a genetically defined, paediatric-onset epilepsy, and the identification funnel is the strategy. Patients reach a Dravet-specific therapy only after genetic testing confirms the diagnosis, so testing rates and referral patterns govern the size of the treatable population far more than promotion does.
Every approved Dravet agent layers onto a generic backbone and competes as an add-on. Sizing therefore runs off add-on eligibility rather than off share of a switchable pool: how many diagnosed patients remain inadequately controlled on the backbone, and how many add-ons a clinician will stack before stopping. A model that assumes patients move from one branded agent to another misreads the clinical reality, in which they accumulate rather than switch.
The pool is small and paediatric, which constrains every forecasting assumption. Numbers at this scale are too small for market-research sampling to resolve, so the funnel has to be built bottom-up and defended on its logic rather than on survey data. It also means a single change in testing or referral practice moves the forecast more than a competitive event does.
Beyond the current symptomatic agents, an antisense oligonucleotide in development targets the underlying genetic defect rather than seizure frequency. If that succeeds, the add-on model the market is built on changes shape, and a forecast that assumes the current structure persists will need rebuilding.
AXLRx Dravet reports build the genetic-identification funnel end to end and size on add-on eligibility. Each report is a sample output of the AXLRx commercial consulting engine, and the same analysis can be commissioned for your asset and market.
For the wider context, see orphan drug approvals in numbers: FDA and EMA orphan approvals counted by year, condition and sponsor.
Dravet syndrome is a rare paediatric-onset developmental and epileptic encephalopathy. US incidence runs about 1 in 15,700 births, consistent with a Dravet Syndrome Foundation registry estimate of roughly 6,000 to 8,000 US patients. It carries one of the highest syndrome-specific SUDEP rates in epilepsy, 9.32 per 1,000 person-years, with a median age at death of 7 years. True prevalence exceeds the molecularly confirmed count everywhere: the UK has an estimated 2,000 to 2,500 patients but only 400 to 500 SCN1A-confirmed, and the GCC 600 to 800 with fewer than 200 confirmed.
Diagnosis combines a characteristic seizure history with genetic confirmation. Onset falls in the first year of life, with febrile seizures and at least two seizures before 12 months, and a likely-pathogenic de novo SCN1A mutation is found in roughly 75% of clinical cases. SCN1A genetic testing defines the addressable funnel: the UK NHS Genomic Medicine Service offers free testing reaching about 75% molecular confirmation, whereas the GCC confirms fewer than 200 patients across only 5 to 6 genetic centres. The genetic-confirmation step, not raw prevalence, gates the market.
Three Dravet-specific agents sit on a generic valproate and clobazam backbone. Stiripentol (Diacomit, Biocodex), a GABA-A modulator, was approved in August 2018 as an adjunct and posted a 71% responder rate in STICLO. Cannabidiol (Epidiolex, Jazz Pharmaceuticals), a plant-derived oral CBD, was approved in June 2018 and showed a 38.9% convulsive-seizure reduction in GWPCARE1. Fenfluramine (Fintepla, UCB), a low-dose serotonin-releasing agent, was approved in June 2020 with a REMS cardiac-monitoring requirement and posted a 62.3% greater seizure reduction versus placebo in Study 1.
The commercial pool is the molecularly confirmed subset, not epidemiological prevalence, so AXLRx sizes from the genetic-diagnosis funnel outward. We start with incidence-derived prevalence, then apply market-specific SCN1A confirmation rates: about 75% where the NHS Genomic Medicine Service funds free testing, under 30% across the GCC's 5 to 6 genetic centres. That gap is the growth lever. Earlier genetic testing in early-life epilepsy converts undiagnosed febrile-seizure children into an identified, addressable pool, expanding the numerator faster than raw prevalence moves. A patient-flow model that segments prevalence down to the SCN1A-confirmed refractory cohort, rather than treating total prevalence as addressable, is the honest basis for any Dravet forecast.
Every Dravet-specific agent layers onto a generic valproate and clobazam backbone and competes as an add-on, never as monotherapy, so switching-share models mislead. AXLRx sizes by add-on eligibility: within the SCN1A-diagnosed funnel, the addressable pool is the refractory subset inadequately controlled on the existing regimen, an estimated 1,400 to 2,000 of 6,000 to 8,000 US patients on cannabidiol plus fenfluramine. We then model combination sequencing, not displacement, asking where a new agent slots against the entrenched cannabidiol-plus-fenfluramine background and what incremental seizure-reduction bar, roughly 40%, it must clear. Pediatric weight-based dosing sets per-patient value, so the forecast becomes eligibility times sequencing share times weight-adjusted price.
Dravet is a small paediatric-onset pool, and that shapes every forecasting assumption. With roughly 6,000 to 8,000 US, 2,000 to 2,500 UK, and 600 to 800 GCC patients, the numbers are too small for market-share smoothing, so AXLRx forecasts patient-by-patient off named specialist centres rather than territory averages. Refractory care concentrates: six NHS centres in the UK, a five-institution Saudi consortium in the GCC, so a KOL-anchored, site-level build beats top-down penetration curves. Paediatric onset means the pool refreshes through incident births rather than a large prevalent backlog, and weight-based dosing, about $96,000 a year for fenfluramine at the high end, makes per-patient value and access, not volume, the commercial battleground.
Today's approved agents are symptomatic, layering seizure control onto the backbone, but the pipeline threatens that model. Zorevunersen (STK-001, Stoke Therapeutics), an antisense oligonucleotide, targets the underlying SCN1A haploinsufficiency rather than symptoms, and a disease-modifying mechanism would reset the value conversation from incremental seizure reduction to trajectory change. AXLRx models this as a scenario overlay on the symptomatic forecast: a genetic-diagnosis funnel that identifies patients earlier becomes even more valuable when the intervention modifies disease, since earlier confirmation widens the treatable window. Soticlestat (Takeda/Ovid), studied in ELEKTRA with no cardiac-monitoring requirement, is the nearer-term symptomatic benchmark. A credible forecast prices both the add-on base case and the ASO disruption case, not one alone.
Every answer above rests on one thesis: in Dravet, the genetic-identification funnel is the strategy, not broad promotion. AXLRx builds that funnel end to end, disease landscape, competitive intelligence, and payer briefs across the US, UK, and GCC; market-sizing and patient-flow models that segment prevalence down to the SCN1A-confirmed refractory cohort; KOL mapping of the concentrated paediatric-neurology centres; and launch-readiness and HTA-strategy work benchmarked against the approaching soticlestat and zorevunersen milestones. Each figure is live-sourced and independently verified before delivery.