Ifosfamide is not a new or marginal drug. It is decades old, off patent, and part of regimens that cure people — sarcomas, testicular cancer, the cancers that disproportionately affect the young and that oncology can often actually beat.
It has been in critical shortage for about three months. American supply ran overwhelmingly through one contract manufacturer, Simtra BioPharma Solutions, formerly Baxter Oncology GmbH, at a facility in Germany. In March the FDA sent that manufacturer a warning letter, following an unannounced inspection in September 2025 that documented recurrent microbial contamination.
Line those dates up. Inspection: September 2025. Warning letter: March 2026. Shortage: around June 2026.
The enforcement and the shortage are one event
This is the part that gets reported as two stories and is one.
A regulator inspecting a sterile injectable plant and finding recurrent microbial contamination has essentially two options. It can permit production to continue, in which case cancer patients receive infusions from a line with a documented contamination problem. Or it can force remediation, in which case the line stops.
If there is a second plant, that is an easy decision with a manageable consequence. Supply shifts, the bad line is fixed, patients notice nothing.
If there is one plant, the decision is not between safe and unsafe. It is between a contaminated drug and no drug, and the regulator has no third option, because the FDA can compel a manufacturer to meet a standard and cannot compel anyone to manufacture at all.
So the warning letter did not fail to prevent the shortage. Under single sourcing, a warning letter of that severity is the shortage, deferred by however long remediation takes to become visible in the pharmacy.
Why the plant is alone
Not through anyone's misconduct. Through the ordinary arithmetic of a generic sterile injectable.
The drug is off patent, so it sells at a price competed down to near cost. It requires aseptic manufacturing, which is among the most demanding and capital-intensive processes in pharmaceuticals — the contamination the FDA found is exactly the failure mode that facility class exists to prevent, and preventing it is expensive. Volumes are modest, because sarcoma and testicular cancer are not common.
Put those together and you have a product with high fixed costs, low margins and limited volume. One producer can just about make it work. Two cannot, because the second one halves the volume without halving the plant. So the market converges on a single supplier, and it does so for every drug with that profile — which is why injectables are about sixty percent of this year's shortages and why nearly half of them ran through a single foreign manufacturer.
Concentration here is not a market failure. It is the market working exactly as designed on a product whose social value and commercial value have almost nothing to do with each other.
What it looks like in a clinic
The substitution question is what separates this from a shortage of something else.
For many drugs there is an alternative that is nearly as good. For a curative regimen there frequently is not: the protocol that produces the cure rate was established with particular agents at particular doses, and departing from it means treating a patient with something whose outcome data does not exist. Oncologists in this position are choosing between delaying treatment, reducing dose, and substituting into a regimen with weaker evidence — three options that are all worse than the one they had in May.
Those decisions do not generate a statistic. Nobody records "cure not achieved because the schedule slipped six weeks." The cost of this shortage will be paid entirely in outcomes that are never attributed to it, which is the same invisibility this desk found in the hospital that runs on a supply chain with one of everything.
The permission problem, again
There is a structural irony here that this paper has written about before: the permission to make a drug expires the moment it works. A manufacturer that has run a product reliably for years has no route to being rewarded for that reliability, and a manufacturer that stops has no penalty beyond lost revenue on a low-margin line.
Nothing in the system pays for redundancy. A second plant sitting at partial utilisation as insurance against the first one failing is, on any commercial analysis, a bad investment — right up to the moment it is the only reason anyone gets treated.
Meanwhile the FDA expects supply problems to persist into the autumn and the European Medicines Agency expects them into 2027, which is the honest measure of how long it takes to remediate an aseptic facility. That interval is not a scandal. It is how long the work takes. The scandal, if there is one, is that the interval has a patient at the end of it rather than an alternative supplier.
The number that would change behaviour
Publish, for every drug on the essential list, the number of qualified manufacturing sites and the share of US supply from the largest one.
Both figures are known to the FDA. Neither is routinely published in a form a hospital pharmacy, a purchaser or a legislator can act on before the shortage. Concentration is currently discovered by everyone at the same moment, which is the moment the plant stops — and by then the only remaining question is which failure the patient gets.
The duration and severity of the ifosfamide shortage, its role in curative and salvage regimens for sarcomas and testicular cancers, the identification of Simtra BioPharma Solutions — formerly Baxter Oncology GmbH — and its German contract facility as the source, the FDA warning letter of March 2026 following an unannounced September 2025 inspection documenting recurrent microbial contamination, the figure that injectables are about 60 percent of drugs in shortage in 2026, the observation that nearly half of this year's shortages ran through a single foreign manufacturer, and the FDA and European Medicines Agency projections for when supply recovers are as reported by Medscape on 2 September 2026. The analysis is our own.
Topics healthdrug shortagesoncologymanufacturing
