I remember a conversation during the I-131 shortage in 2022, when we were trying to manage the limited availability of a product that was needed for patient treatment.
Different markets were asking for their share of the available supply. Some argued that their market should have priority over others.
In one of those discussions, someone argued that their market should receive priority.
And then he asked me, as the site lead:
“Do you know it is for cancer patients?”
“Yes, of course,” I replied.
Because we knew. We were very much aware that behind those orders were patients awaiting treatment. And that made the situation even more difficult.
But the patients in one market were no less important than those in another.
And knowing how important the treatment was did not create additional supply.
So how do you decide which market should receive more when there simply isn't enough product available?
We eventually decided to approach it mathematically. First, we established how much I-131 we could realistically obtain. We then divided that available quantity proportionally, based on what the different markets had ordered previously.
Was that perfectly fair? I don't know. Because what is fair in such a situation is incredibly difficult to define.
But it gave us a consistent basis for making decisions, rather than allowing the strongest arguments to determine who received what.
And while that helped us manage the situation, it did not solve the actual problem. No matter how hard we worked to mitigate the impact, we still had insufficient supply.
How did we get there?
Looking back at 2022, it wasn't one single event that created the situation.
It was a combination of developments, each with its own background, requirements and consequences.
We were still dealing with the effects of COVID-19. Technical projects had encountered delays and difficulties. At the same time, the transition from highly enriched uranium (HEU) to low-enriched uranium (LEU) continued to present challenges for parts of the medical isotope supply chain.
That transition was important, particularly from a nuclear security perspective. But changing production processes is not simply a matter of replacing one material with another. It can affect yields, processing, qualification, capacity and ultimately the availability of the resulting products.
From my own experience at the time, I remember discussions about transition timelines, requests for extensions and the practical consequences when further flexibility was no longer available.
These were not necessarily developments affecting every producer in the same way. But they were part of a wider situation in which multiple dependencies and constraints had to be managed simultaneously.
At the same time, the priority was understandably on restoring and securing Mo-99 production, the basis for Tc-99m used in diagnostic nuclear medicine.
And then there was I-131.
For some production routes, I-131 was recovered as a by-product of Mo-99 production. But not every Mo-99 producer had the capability to process that I-131 into material meeting pharmaceutical requirements.
So even when production capacity existed somewhere in the system, that didn't automatically mean pharmaceutical-grade I-131 was available to the market.
There were only a few significant producers internationally, and the supply chain had limited flexibility when production capacity became unavailable.
In August 2022, Nuclear Medicine Europe warned of potential I-131 supply disruptions, identifying three major international suppliers and highlighting the impact of planned reactor maintenance.
Other disruptions affecting medical isotope production, including planned and unplanned reactor outages, further demonstrated how vulnerable these interconnected supply chains could be.
Not every disruption affected I-131 availability in the same way, and not every development was directly connected. But together, these events exposed the limited resilience of a system in which different products, production routes and organisations depended on overlapping infrastructure and capabilities.
What I remember most is how quickly a technically complex supply-chain problem became an operational problem that affected everyone downstream.
And ultimately, the patients.
When individual decisions affect the whole system
One of the things I have learned throughout my career is that decisions can make perfect sense within their own area of responsibility, while creating difficulties somewhere else.
Improving nuclear security is important. Maintaining regulatory requirements is important. Qualifying a new production process is important. Managing investments and operational costs is important.
None of those objectives should be dismissed.
But they don't exist independently.
A technical change may affect production capacity. Reduced capacity may affect delivery commitments. Delays in qualification may prevent alternative sources from being used. And when several disruptions happen simultaneously, a supply chain that seemed sufficient under normal circumstances may suddenly have very little flexibility left.
The problem is not necessarily that individual decisions were wrong.
It is that their combined consequences may not have been sufficiently understood, communicated or managed.
And this is where I think SQDCP becomes particularly relevant.
SQDCP is more than an operational dashboard
We often use SQDCP — Safety, Quality, Delivery, Cost and People — to understand and manage operational performance.
But I believe it is equally useful when looking at the design and resilience of an entire supply chain.
Safety is not negotiable. Nuclear security, radiation protection and patient safety all need to be considered. But when we introduce changes to improve safety or security, we also need to understand the consequences for the rest of the system.
Quality means that producing an isotope is not enough. The material must meet the applicable specifications, and the production route, processing and controls must be suitable for its intended pharmaceutical use. A supplier who can physically produce an isotope is not automatically a qualified alternative supplier.
Delivery is about more than meeting an agreed delivery date. For radioisotopes, time matters. Material decays, treatment schedules depend on availability, and lost production capacity cannot always be recovered later. A missed delivery may mean a postponed patient treatment.
Cost is where we sometimes face difficult trade-offs. Additional capacity, alternative suppliers, qualified backup processes and infrastructure investment all cost money. But what happens when we optimise costs so far that there is no reserve capacity left when something goes wrong?
And then there is People.
For me, this is the dimension connecting all the others.
Because no system works without people understanding each other's needs, limitations and responsibilities.
The people developing production technologies, those responsible for qualification and regulatory acceptance, the operators running the facilities, the suppliers managing capacity, the manufacturers depending on that supply, and the hospitals treating patients.
They may all look at the same supply chain from very different perspectives.
And unless those perspectives are brought together, important dependencies can remain invisible until something goes wrong.
Are we building the next shortage?
Today, we are seeing tremendous developments in therapeutic radioisotopes.
Ac-225, Pb-212, Lu-177 and other isotopes are opening opportunities for new treatments, particularly in oncology.
Companies are investing in new production routes, facilities and technologies. Clinical programmes are developing, new products are being investigated, and expectations for future demand are growing.
That is an exciting development.
But it also makes me wonder whether we are sufficiently applying the lessons from the past.
When we talk about future Ac-225 capacity, for example, what exactly do we mean?
The ability to produce a certain amount of activity?
The ability to recover and purify the isotope consistently?
The availability of material meeting pharmaceutical requirements?
Or the actual number of patient treatments that can reliably be supplied?
Those are not necessarily the same thing.
And how independent are the different supply routes?
We may have multiple suppliers, but if they depend on the same starting materials, specialised infrastructure, processing capabilities or limited expertise, we may still have common vulnerabilities.
The same applies to production methods. Different routes can introduce different technical challenges, impurity profiles, waste streams, qualification requirements and economic considerations.
All of this needs to be understood before the supply chain is expected to support substantial commercial demand.
Because approving a new treatment and making it available to patients are two very different achievements.
Building capacity is not the same as securing supply
I think we need to look beyond the question of whether enough isotope can be produced to meet today's demand.
We also need to consider what happens when demand increases, when clinical trials progress towards commercialisation, when a facility has an unexpected shutdown, or when a new production method takes longer to qualify than anticipated.
And perhaps equally important: what happens when several of those things occur at the same time?
Do we have alternative sources that can actually be used?
Do we understand the shared dependencies between suppliers?
Are capacity forecasts based on realistic clinical demand, or mainly on current commercial commitments?
Are we investing in sufficient resilience, or assuming that everything will operate according to plan?
And are we communicating those expectations throughout the supply chain?
I don't believe every disruption can be prevented. That would not be realistic.
But we can understand dependencies better. We can plan for different scenarios. We can qualify alternative routes earlier, improve communication between stakeholders and make more informed decisions about where reserve capacity is necessary.
That also means thinking about the economic model.
Who pays for capacity that may not be needed every day but becomes essential when another producer is unavailable?
Who takes the investment risk when future clinical demand is promising but still uncertain?
And how do we make sure that decisions focused on short-term efficiency do not undermine long-term reliability?
These are not questions individual suppliers can answer on their own.
They require cooperation across the entire system.
The importance of understanding each other
In much of my professional career, I have worked at the interfaces between different functions, organisations and requirements.
And one thing I have seen repeatedly is that people are often willing to find solutions once they understand why something matters.
But that understanding doesn't always come automatically.
A reactor operator may not fully understand the consequences of a particular production delay for a pharmaceutical manufacturer. A pharmaceutical manufacturer may not understand the technical limitations of the isotope producer. And a commercial organisation may not always recognise how little flexibility is available once production has already been planned.
That doesn't mean anyone is unwilling to cooperate.
Sometimes the connections simply haven't been made.
Which is why communication cannot be something we only start organising when there is a shortage.
We need to build those relationships while we are building the supply chains.
Not only between customers and their direct suppliers, but also further upstream and downstream.
Understanding what others need, why they need it, and what the consequences are when something changes.
And being willing to explain our own limitations as well.
What will we do differently this time?
When I look back at 2022, I don't primarily remember the calculations we made to divide the available I-131.
I remember the enormous effort that went into trying to maintain supply in a situation where many of the options had already become limited.
And I remember how difficult it was to balance the needs of different markets while knowing that the products were intended for patients who were waiting for treatment.
Today, with new therapeutic isotopes moving towards broader clinical application, we have an opportunity to consider these questions much earlier.
Not only how to increase production.
But how to build supply chains that remain reliable when circumstances change.
How to connect Safety, Quality, Delivery, Cost and People across organisational boundaries, rather than managing each of them separately.
And how to create enough mutual understanding that the consequences of decisions are recognised before they become problems somewhere else.
We have experienced what happens when the supply of essential radioisotopes becomes insufficient.
So, while we are developing the next generation of radiopharmaceuticals, shouldn't we also be asking whether we are building the supply chains needed to support them — not just today, but in the future?
And perhaps an even more important question:
What have we learned from the shortages we already experienced, and where are we actually applying those lessons?
References and further reading
- Nuclear Medicine Europe (2022). Communication on I-131 Supply, 24 August 2022.
- OECD Nuclear Energy Agency (2023). The Security of Supply of Medical Radioisotopes: Demand and Capacity Projections for Mo-99/Tc-99m for the 2023–2027 Period.
- OECD Nuclear Energy Agency (2025). Current Trends in the Supply and Utilisation of Medical Radioisotopes.
- OECD Nuclear Energy Agency (2026). Medical Radioisotopes Supply Workshop Weighs Advances and Trade-offs.
