A legacy pharmacology program—validated in clinical trials but locked behind outdated licensing—may be giving Marseille patients safer, more personalized treatment doses while the rest of France waits.
In the early 2000s, a team of pharmacologists, mathematicians, and clinicians in Marseille achieved something unprecedented: a clinical trial in metastatic breast cancer guided not by intuition or fixed-dose protocols, but by a mathematical model that updated dosing cycle by cycle. The model-adjusted treatment, based on patient-specific pharmacokinetics, aims to reduce toxicity while preserving or improving efficacy.
“It was the first time in the world that there has been a clinical trial based on the recommendations of a mathematical model for the dosing,” recalls Sébastien Benzekry, head of the Inria–Inserm COMPO (Computational Pharmacology & Clinical Oncology) team. “We adapted depending on how the patients responded at each cycle, how to adapt for the next cycle.”
The results were striking.The median survival of participating patients doubled compared with historical controls. Although the trial’s impact was blunted when standards of care changed before implementation could expand, the underlying tool—a dosing-adaptation software built on decades of population pharmacokinetics—endured. It continues to be used daily by pharmacists in Marseille.
And only in Marseille.
Across the rest of France, clinicians who hear about the tool ask why they cannot access a system that has proven effective and simple to use. The answer, as Benzekry puts it, is “a completely stupid” accident of licensing history.
Held Hostage by Licensing: How One City Kept Access
The software began as an academic project in Marseille but was later transferred to a private company for commercialization. When that company stopped developing the product, it inadvertently froze its distribution. Although the firm dropped the software, it retained the intellectual property rights, preventing new licenses from being issued.
“Now, the company doesn’t exploit it and has dropped the product but still holds the license,” says Benzekry. “No one else can get new licenses to use the software.”
Marseille’s hospital retains access only because it purchased four licenses before the company ceased operations. Pharmacists there continue to use the tool in routine care, expertly adjusting doses for therapies whose pharmacokinetics vary widely between patients. Benzekry believes this gives Marseille an edge in specific drugs where adaptive dosing is known to improve tolerability.
“Surely, patients could benefit from the use of the dosing software,” he explains. “Instead, they receive the same dose as everyone else.”
The contrast is stark. In many other French hospitals, clinicians must either rely on generic fixed-dose regimens or use complex, specialist-only PK software unsuitable for daily use by busy pharmacists. Meanwhile, whenever Marseille pharmacists present their work at congresses, colleagues from across France express frustration.
“They’d love to have this tool because it’s very user-friendly,’” Benzekry says. “But no one else can access it.”
This unique situation, where a clinically validated tool remains confined to one city, highlights how regulatory and ownership structures can suppress otherwise promising innovations in personalized medicine.
A Path Forward: Open-Source, Certification, or a New Startup
Despite its usefulness, the software is aging. For Marseille to keep using it safely and for the rest of France to access it, an update is needed. But modern EU medical device regulations mean that even updating an old tool requires fresh CE marking, a process that can take years.
“Nowadays, the regulation has changed, and if we want to have this software developed and move forward, we need the certification to get the CE marking,” Benzekry explains. “This gets out of the scope of a research team.”
COMPO is now exploring three potential solutions.
1. Rewriting the software from scratch
The team is already attempting to rebuild the program, but deploying it beyond research settings still requires certification.
2. Launching a startup to commercialize a modernized version
This route would allow the team to navigate CE-marking requirements with appropriate resources, something academic labs lack.
3. Releasing an open-source, browser-based version
Perhaps the most ambitious plan is to bypass installation hurdles altogether.
“What we would like to do is to bring an open-source, freely available version of the software that would be available via a web interface,” Benzekry says. “So that they would not need to install anything local.”
A web-based system would not just democratize access across France; it could also solve the hospital IT barriers that prevent institutions from installing external software and even newer commercial tools from the industry.
There is competition: a Swiss company has developed a modern dosing tool, but Marseille cannot install it because the hospital’s IT department blocks third-party installations. An open-source browser-based system could bypass those barriers completely.
“This type of system does exist, but only for a specific drug for specific therapy areas; the software in Marseille allows users to enter their own population and data for a drug, which is important in oncology because we consistently have new drugs.”
For now, Marseille remains the sole beneficiary of a technology with clear clinical value and a long scientific pedigree. But with renewed attention, committed engineering support, and potentially a new commercial vehicle, COMPO hopes to finally bring this tool out of its licensing limbo.
Because as Benzekry notes, “Patients can benefit from the use of the software.”
And if the tool were accessible beyond Marseille, many more could.
