Half of 2025's Orphan Approvals Rested on 5,001 Patients

The FDA’s Center for Drug Evaluation and Research approved 46 novel drugs in 2025. Twenty-three of them carried Orphan Drug Designation — exactly half, and the roundest number in the year’s approval statistics.

The half that carries the commercial argument is a different half. Between them, those 23 programmes were supported by 5,001 patients in the pivotal trials that took them to approval. On the roughly 26,000 pivotal participants behind all 46 novel approvals, the orphan half is 19.2 per cent of the evidence and 50.0 per cent of the approvals.

That is not a scandal. It is the arithmetic of rare disease working as designed, and it is the most useful single fact for anyone pricing this market — because it explains almost everything else about it. Why 23 approvals arrived from something close to 23 companies rather than from five. Why the price at approval is what it is. Why the binding constraint sits downstream of approval rather than at it. And why the most valuable act available to an orphan sponsor in 2025 was not a price rise.

What 5,001 patients across 23 programmes looks like

A mean of 217 patients a programme. A median of 114. The distribution matters more than either: the cohort is not clustered around 114, it is split into a majority of small programmes and two very large ones.

Two programmes — Grafapex at 570 patients and Jascayd at 1,322 — carry 37.8 per cent of the cohort’s entire pivotal evidence. That is 8.7 per cent of the approvals holding more than a third of the patients. The nine smallest programmes together enrolled 601 patients, which is 2.2 times fewer than Jascayd enrolled alone.

601 plus 2,508 plus 1,892 is 5,001 — the cohort total reconstructed from the bands rather than asserted, which is the only reason to trust any of the shares above it.

The nine

These are the programmes where the entire evidence base for a US approval is smaller than a single mid-sized Phase II in a common disease.

Forzinity’s twelve patients is the extreme case, and it is worth sitting with rather than skimming. Twelve patients is a US approval for a fatal paediatric mitochondrial disease, and it is also a commercial proposition whose entire post-approval safety record will be written by the first two or three hundred people who take it. Both statements are true at once, and a market model that only registers the first one will misprice the asset.

The comparison that reframes the year

Three non-orphan approvals in the same year enrolled more patients between them than the entire orphan cohort did across all 23 programmes.

9,279 against 5,001 is 1.86 times. This is the cleanest way to state what the orphan cohort is: not a smaller version of the standard approval cohort, but a structurally different one, where the evidentiary threshold that a payer, a prescriber and a litigator would each want is not what the approval rests on.

Why the small programme is the product

Follow the small programme through to the industry structure and it explains something that otherwise looks like fragmentation for its own sake.

The capital required to reach a US approval in an orphan indication is a fraction of what a common-disease programme costs — a median of 114 patients, no large safety database, a single-arm design in many cases. That lowers the barrier to a first approval far enough for a company with one asset and a contract research organisation. It is why the approval layer atomises into single-asset sponsors even as the revenue layer consolidates, and it is why the buyer of an orphan company is buying a label and bolting it onto a distribution machine it already owns.

That inference is drawn from structure, not from a disclosed transaction. There is no 2025 orphan deal with a published price that would let this be stated as fact, and the market report behind this article says so rather than dressing it up — a single disclosed acquisition materially above the cohort’s implied average would confirm it, and one materially below would falsify it.

Where the small programme meets the payer

The evidence base that gets a drug approved is not the evidence base that gets it reimbursed, and the gap is measurable.

Roughly one third of outpatient prescriptions fill within 30 days in the first post-approval year. At the median net price that means a filled patient-year is worth USD 257,000 and a written prescription is worth USD 85,658.

And the channel split is not random against this cohort. Acromegaly, generalised myasthenia gravis, idiopathic pulmonary fibrosis, multiple myeloma and obstructive hypertrophic cardiomyopathy are Medicare-weighted populations, and between them they carry five of the cohort’s larger programmes. A therapeutic-area segmentation without a channel segmentation overstates revenue in exactly the segments that sit outside oncology — the opposite of the intuitive error.

The price stopped cooperating at the same time

ICER and West Health put the 2025 median annual net launch price at USD 257,000 against a median list price of USD 293,000. That is up 35 per cent on USD 190,000 in 2022 — 10.59 per cent a year — but down 2 per cent on 2024, the first decline in the series.

Two readings are available and the report declines to pick between them as though it could. The discipline reading: payers pushed back. The arithmetic reading: route of administration predicts price better than the science does, and the 2025 cohort is unusual on route.

Twelve of 23 are oral solids. A cohort that is 52.2 per cent oral, with nine programmes under a hundred patients, should have a list median below the USD 314,000 oral average — and USD 293,000 is exactly that. So part of the 2 per cent net decline is probably product mix rather than price discipline. It is stated as a hypothesis the data support and do not prove, because ICER does not publish a 2025 route breakdown.

The second rare label became the whole strategy

While the price stopped cooperating, policy moved the other way. Section 71203 of the Working Families Tax Cuts Act (Public Law 119-21) widened the Medicare negotiation orphan exclusion: a product is now excluded if all its approved indications are rare, however many there are, rather than only if it has a single indication.

That inverts the historic calculus. Before, a second indication risked the exclusion. Now a second rare indication both expands revenue and permanently re-affirms it — a non-rare label is the mistake, not a second label. The base rate for follow-on activity is already 19.9 per cent of specialty drug-indication pairs (188 of 947), at a mean of 1.7 per product, and the policy now pays for the second one in a way it did not in 2024.

The floor is less settled than the statute reads. CMS’s IPAY 2028 final guidance, published 30 September 2025, still frames the exclusion around a product designated for “only one rare disease or condition”, and its comment record shows CMS replying that it cannot amend the statutory requirement. The statute is broader than the policy that articulates it, and any investment case leaning on this floor should size how much of its terminal value depends on which reading survives. The report treats it as open, and so does this article.

Three things that would prove this wrong

1. A second year of falling orphan share. ICER’s 2025 mix table has the orphan share of approvals at 55 per cent against 60 per cent in 2022–24, and every high-price category falling. The report’s base case has that share rising from 50.0 to 54.0 per cent by 2033. One down year is noise; two would falsify the central input, and it is the cheapest test available.
2. Fill rates improving without a coverage change. If first-year 30-day fill moves materially above one third on commercial coverage alone, the value-chain argument in the report is wrong about where the constraint sits.
3. A disclosed orphan acquisition priced materially below the cohort’s implied average. That would falsify the inference-from-structure above, and nothing else in this analysis depends on it.

How this was built, and where it is soft

Every figure here traces to a primary document. The approval count and the orphan split come from CDER’s 2025 novel approvals list and the FDA’s designation programme pages; pivotal programme sizes come from the approval documents themselves; prices from ICER and West Health; the policy change from the statute text and CMS’s guidance; the base year from Tufts Medical Center’s specialty-drug analysis.

Two soft points, stated rather than buried. The “roughly 26,000” total pivotal participants across all 46 approvals is a sum across approval documents of varying reporting completeness, so the 19.2 per cent share is directional and the 5,001 is exact. And the market-level numbers — USD 157.7 billion in 2025 to USD 224.9 billion in 2033 at 4.54 per cent — rest on a base year that is derived: USD 263 billion of 2024 US specialty drug spending, times the 55 per cent orphan share of specialty indication pairs, grown 9.0 per cent. A reader who rejects either input can rebuild it, which is why the full chain is published on the US orphan drug market report rather than only its outputs.

The adjacent questions this library already sizes separately are the ones the cohort keeps pointing at: acute myeloid leukemia and CAR-T cell therapy, where the follow-on label rather than the first approval carries the revenue; malignant infantile osteopetrosis, where a population measured in the low hundreds nationally sets the ceiling instead of competition; and clinical laboratory services and diagnostic tests, which sit upstream of every one of these diagnoses and therefore upstream of the fill.

All figures are drawn from the Pheonix Research report US Orphan Drug Market Report, Size, Share and Forecast 2026–2033.