The Shortage Can End Before the System Scales: An Ac-225 Systems Analysis 2026-2035
If the alpha pipeline works, bulk Ac-225 is not the first part of the system to fall behind. This analysis models the whole chain - evidence, approval, reimbursement, centres, finished dose, bulk isotope and workforce - as a delayed feedback system, and asks which layer binds in which year under four scenarios.
What this report answers
- If Ac-225 supply stops being the binding constraint, what becomes the constraint instead?
- What has to be true before bulk isotope is genuinely less restrictive, and when could that happen?
- How much of the announced Ac-225 capacity is usable in a product file today?
- Does an existing Lu-177 treatment centre translate into alpha capacity?
- Which capabilities take longest to build, and where does overbuild risk sit?
- How would an operator know the bottleneck had moved before it showed up in results?
12.6%
the nuclear-medicine technologist vacancy rate the report cites from a 2025 professional-body staffing survey, set against an average of roughly four budgeted FTEs per responding department.
4.25 years
the observed interval from groundbreaking to stated routine weekly output at one Ac-225 producer, used in the report as the anchor for its production lead-time ranges.
What the analysis establishes
What the end of the shortage would actually require
The analysis sets a threshold test rather than a supply forecast. It defines the conditions that must hold at the same time before bulk Ac-225 becomes materially less restrictive, notes that most of them are operating conditions rather than announcements, and gives three dated windows under different upstream outcomes. Each window is stated as an inference with a confidence level, and none carries a probability.
Announced capacity against usable supply
A capacity-maturity ladder runs from press release to administered treatment, and the analysis places the current field of producers and projects across it using their own stated positions. It marks where the gate sits, shows how far most public reporting stops short of it, and explains why production route changes the downstream assay, waste and release work.
The clinical clock, and why the programme count misleads
Four public pivotal programmes carry the commercially consequential weight, with registry completion dates and company guidance mapped onto a catalyst timeline. The analysis then regroups them by target biology, selection imaging and referral pathway, and shows how that regrouping changes what a given number of clinical successes actually means for demand.
What a Lu-177 network carries over, and what it does not
The Lu-177 franchises built licences, imaging pathways, radiopharmacy workflows and payer experience. The analysis treats that estate as an option on alpha capacity and sets out, capability by capability, the additional requirement an alpha programme adds on top of the mature beta baseline - including which of those items capital cannot accelerate.
Lead times, and the capacity that money cannot buy
Eleven expansion capabilities are compared on stated planning ranges, from licence amendments through finished-dose sites to forming new specialists, sorted slowest last. Public workforce evidence from US and European professional bodies and labour statistics sits underneath the slowest lines, and the analysis separates capital-elastic capacity from experience-dependent capacity.
Four futures, and the signal that the bottleneck has moved
Four scenarios are tested against five questions each, then resolved into a year-by-year map of which layer binds under each one. The analysis argues that migration is confirmed operationally rather than by announcement, names the reporting granularity required to see it, and sets out fifteen dated decision nodes and a quarterly dashboard for tracking it.
How it was built
Built entirely from dated public sources - trial registries, regulator notices and proposed rules, government isotope programme reporting, IAEA guidance, peer-reviewed literature, professional-body surveys and company disclosure - all current at the 31 July 2026 evidence cutoff, with 32 sources listed. The report separates four kinds of statement: established fact, company disclosure reported as what the company says rather than as demonstrated output, inference drawn across sources, and planning assumption used to compare lead times against each other. Six interpretation rules are applied throughout, the wider of two defensible ranges is used, and patient numbers, market size, aggregate curie totals, scenario probabilities and company revenue are deliberately not modelled because the public evidence does not support them.
Written for radiopharma operators, isotope producers, CDMO and finished-dose leadership, investors and health-system planners making supply, manufacturing and treatment-capacity decisions across the US and Europe.
Contents
- If the pipeline works, the isotope is not what fails first
- Five conditions, all at the same time
- Eleven nodes, three loops, and one unit of capacity
- Later than the narrative, and more correlated than the count suggests
- Eight rungs between a press release and a treated patient
- Bulk isotope does not treat a patient
- A Lu-177 center is an option on alpha capacity
- Capital moves in quarters. Experience moves in years.
- What must be true, and what breaks first
- The migration signal is a reason code, not a press release
- Three questions resolve into four futures
- What the system does that no single function can see
- Decisions that hold up in all four futures
- How to argue with this report
Organisations, agencies and professional bodies referenced
Novartis, Bristol Myers Squibb, AstraZeneca, Bayer, TerraPower Isotopes, PanTera, SCK CEN, NorthStar Medical Radioisotopes, Cardinal Health, Nusano, Ionetix, Eckert & Ziegler, Niowave, ARTBIO, US Department of Energy Isotope Program, IAEA, FDA, US Nuclear Regulatory Commission, EMA, SNMMI, EANM, Centers for Medicare & Medicaid Services, US Bureau of Labor Statistics, American Society of Radiologic Technologists
The full 31-page briefing is free. Tell us where to send it and it downloads immediately.