A cancer drug carries its approval like a promise, but a growing body of research suggests that promise was made to someone else. Scientists from Zurich, Yale, and Harvard have found that one in three cancer therapies approved in the United States—and one in five in Europe—are routinely administered to patients whose profiles were never included in the trials that justified those approvals. This is not a minor technical discrepancy; it is a quiet erosion of the evidentiary foundation upon which modern medicine rests, leaving patients and physicians navigating treatment decisions in the dark.
Cancer drugs approved for patient profiles untested in clinical trials
Cancer patients receive medications without evidence of safety or efficacy in their specific disease profiles, risking serious adverse effects and treatment failure.
You have a right to know that your drug was never tested on someone like you.
Mark
Why does this gap exist at all? Couldn't regulators simply require that drugs only be approved for the exact populations tested?
Mimi
In theory, yes. But cancer is urgent, and there's genuine pressure to get treatments to patients quickly. The problem is that the urgency has outpaced the rigor.
Mark
So accelerated approval is the culprit?
Mimi
It's part of it, but not the whole story. Even drugs approved through standard pathways end up being used in untested populations. The real issue is that once a drug is approved for one indication, doctors can prescribe it for others—and regulators have limited ability to stop that.
Mark
And patients don't know they're in uncharted territory?
Mimi
Exactly. The label might say the drug was tested in patients with metastatic disease, but a doctor might prescribe it to someone with early-stage cancer. Both doctor and patient may not realize the evidence doesn't cover that situation.
Mark
What's the harm if the drug works anyway?
Mimi
That's the gamble. It might work. But it might also fail silently, or cause severe side effects no one anticipated because no one had tested it in that population. You're essentially running an experiment on a patient without calling it that.
Mark
So what would fix this?
Mimi
Transparency, first. Make the gaps visible on the label. Then, make trial populations more representative of who actually gets sick. And hold regulators accountable for the decisions they make.
Il Polso
- Thousands of cancer patients are receiving drugs each year without any clinical evidence that those drugs are safe or effective for people like them—their age, their mutation, their stage of disease.
- Accelerated approval pathways, designed to rush promising treatments to the sick, are compounding the problem by allowing drugs onto the market before confirmatory studies are complete—studies that sometimes never arrive at all.
- Regulatory agencies on both sides of the Atlantic are permitting this drift at different rates, with the FDA approving misaligned uses in nearly a third of cases and the EMA in roughly a fifth, yet neither has made its internal decision-making transparent.
- Researchers are calling the combination of off-label use and fast-tracked approvals a 'double risk'—where reduced efficacy and serious adverse effects can emerge in populations that were simply never part of the original research.
- Health systems spending enormous sums on these therapies have no reliable basis for those funding decisions when the drugs have not been tested in the very patients being treated.
A cancer drug carries its approval like a promise, but a growing body of research suggests that promise was made to someone else. Scientists from Zurich, Yale, and Harvard have found that one in three cancer therapies approved in the United States—and one in five in Europe—are routinely administered to patients whose profiles were never included in the trials that justified those approvals. This is not a minor technical discrepancy; it is a quiet erosion of the evidentiary foundation upon which modern medicine rests, leaving patients and physicians navigating treatment decisions in the dark.
A cancer patient takes a newly approved drug home from the pharmacy. Neither they nor their doctor may know it was never tested on someone quite like them—their age, their mutation, their stage of disease. This gap between clinical trial populations and real-world patients is wider than most realize, and it is growing.
Researchers from the universities of Zurich, Yale, and Harvard analyzed 89 cancer drugs across 92 indications, comparing the patient profiles enrolled in trials against those listed on labels after regulatory approval. Their findings, presented at the European Society for Medical Oncology congress in Barcelona, reveal a troubling misalignment: in the United States, roughly one in three approved cancer medications are given to patients whose characteristics were never tested. In Europe, the figure is one in five. The researchers expected some drift—that is inevitable—but not this magnitude.
Leading the investigation, physician and legal scholar Kerstin Noëlle Vokinger of the University of Zurich noted that regulatory agencies differ in how strictly they police this boundary. The FDA permits such misalignment in 32.9 percent of cases; the EMA in 20.3 percent. When a drug has not been tested in a specific age group, genetic profile, or disease stage, no one can predict whether it will work or what harm it might cause—and many cancer medications carry severe side effects.
The problem deepens with accelerated approval pathways, now common in oncology, which require less upfront evidence on the assumption that confirmatory data will follow. When those studies are delayed for years or never conducted, the evidentiary gap only widens. Vokinger describes the result as a 'double risk': reduced effectiveness combined with the potential for serious adverse effects in people who were never part of the original research.
The forces driving these approvals remain opaque. Industry pressure, the genuine urgency to reach sick patients, and a lack of transparency inside regulatory agencies all likely play a role. What is clear is that patients and physicians are making decisions without the information they need. If a drug was never tested in someone with your profile, that fact should appear explicitly on the label—but these gaps often remain invisible.
The stakes extend beyond individual safety. Public health systems paying for these extraordinarily expensive treatments deserve to know whether the evidence supports use in the populations they serve. Better labeling, more representative trial designs, and stricter regulatory alignment between what was studied and what is approved could all help—but only if patients, physicians, and policymakers first understand the scale of what is happening.
A cancer drug arrives at the pharmacy with regulatory approval stamped on its label. A patient takes it home. Neither the patient nor their doctor may know that this particular medication was never actually tested on someone quite like them—someone their age, with their specific mutation, at their stage of disease. This gap between the patients who appear in clinical trials and those who eventually receive the drugs in hospitals and clinics is wider than most people realize, and it is growing.
Researchers from the universities of Zurich, Yale, and Harvard have spent months comparing what happens in the controlled world of drug trials against what unfolds in real medical practice. Their findings, presented this week at the European Society for Medical Oncology congress in Barcelona, reveal a troubling misalignment. In the United States, roughly one in three newly approved cancer medications end up being given to patients whose characteristics were never tested in the trials that led to approval. In Europe, the figure is one in five. Switzerland sits between them at roughly one in four. The difference matters because it means thousands of patients are receiving treatments with unknown safety profiles and unproven effectiveness in their particular situation.
Kerstin Noëlle Vokinger, a physician and legal scholar at the University of Zurich, led the investigation. She and her colleagues analyzed 89 cancer drugs approved across 92 different indications, comparing the patient profiles included in clinical trials against the populations listed on drug labels after regulatory approval. They looked at age, disease subtype, severity, and prior treatments. The researchers expected some drift between trial populations and real-world patients—that is inevitable and even expected. What surprised them was the magnitude. "We thought there would be differences," Vokinger explained in an interview. "But we did not expect to find that one in three drugs approved in the United States and one in five in Europe were being given to patients with profiles not included in the trials."
The regulatory agencies themselves differ in how strictly they police this boundary. The U.S. Food and Drug Administration approves cancer drugs for use in patient populations not represented in trials at a rate of 32.9 percent. The European Medicines Agency is more cautious at 20.3 percent. This permissiveness carries real consequences. When a drug has not been tested in patients with specific characteristics—a particular age group, a certain genetic mutation, advanced metastatic disease—no one can predict whether it will work or what harm it might cause. Many cancer medications carry severe side effects. Administering them to untested populations is a gamble with incomplete information.
The problem is compounded by the accelerated approval pathways now common in oncology. These routes to market require less evidence upfront, with the understanding that additional data will arrive later to confirm safety and efficacy. But when those confirmatory studies are delayed for years or never conducted at all, the gap between what was proven and what is being used only widens. A drug approved on limited evidence, then prescribed to patients outside its tested population, creates what Vokinger calls a "double risk"—both reduced effectiveness and the emergence of serious adverse effects in people who were never part of the original research.
Why does this happen? The answer is not entirely clear. Vokinger notes that the decision-making processes inside regulatory agencies remain opaque. Pressure from the pharmaceutical industry may play a role. So may the genuine desire to get promising treatments to sick patients as quickly as possible. The truth is that no one has fully investigated the mechanisms driving these approvals. What is clear is that patients and their doctors often lack the information they need to make informed choices. If a medication was not tested in someone with your age, your disease stage, or your genetic profile, you have a right to know that. The label should say so explicitly. Instead, these crucial gaps often remain invisible.
The implications extend beyond individual patient safety. Public health systems must decide whether to fund these treatments, and they are extraordinarily expensive. It is reasonable for a health system to hesitate before paying for a drug that has never been tested in the population it is being asked to treat. Better labeling transparency, more representative trial populations, and stricter regulatory alignment between what was studied and what is approved could all help. But none of that will happen without awareness of the problem itself—and without pressure from patients, physicians, and policymakers who understand what is at stake.
Citazioni salienti
We expected some differences between trial patients and real-world patients, but not that one in three US-approved and one in five European-approved cancer drugs would be given to people whose characteristics were never tested.— Kerstin Noëlle Vokinger, University of Zurich
If a drug was not tested in patients with your age, disease stage, or genetic profile, you have a right to know that. The label should say so explicitly.— Kerstin Noëlle Vokinger, University of Zurich
Vuoi la storia completa?
Leggi l'originale su El País ↗
Health & Medicine