This tragedy — recounted in an article in the March 4 issue of The Lancet by two Stanford University School of Medicine neurologists — serves as a telling case study of what can go wrong in clinical trials. In their article, Annette Langer-Gould, MD, and Lawrence Steinman, MD, warn of the pitfalls of testing a drug with unknown side effects in patients who would do fine without the drug.
The drug in question is natalizumab, which has the brand name of Tysabri. In November 2004, the U.S. Food and Drug Administration fast-tracked its approval for use in multiple sclerosis patients following promising results seen early in two clinical trials. But within months of the approval, some patients taking the drug had developed a rare infection — progressive multifocal leukoencephalopathy, or PML — and Smith and one other patient had died.
Langer-Gould, a clinical instructor in neurology, treated a patient who was part of the clinical trial and developed PML after taking Tysabri; the patient survived. But that experience, coupled with an examination of Smith’s case, prompted Langer-Gould to approach Steinman about writing an article that would examine the appropriateness of testing a drug on people with no evidence of the disease and who are not disabled at the time of the trial.
“We are arguing that people with no disability should probably not enter into a clinical trial or be recruited into clinical trials, because where is the potential benefit to them if nothing is wrong?” said Steinman, professor of neurology and neurological sciences and of pediatrics.
“This situation represents a systemic problem,” said Langer-Gould. “It is not just one company being a rogue, doing something out in left field.”
Langer-Gould and Steinman argue that if a drug has a known risk of death, it should only be used on patients who are likely to suffer severe disability from the targeted disease — and for whom there are no other options. In other words, those who have tried all the other available therapies. That is almost the reverse of what happened in the Tysabri trial, which excluded the most severely affected patients.
“A big mistake was made in these trials that, in my opinion, is easily preventable,” said Langer-Gould. “All they need to do is tighten up entry criteria into multiple sclerosis clinical trials and we could avoid similar types of problems in other trials.”
Multiple sclerosis results when the immune system attacks the protective myelin sheath surrounding nerve cells, causing them to misfire and leading to loss of motor control and possibly paralysis. Tysabri appeared to block this effect and, after the first year of the two-year clinical trials, did not appear to cause more infections.
Steinman was involved early on in the development of the drug, publishing on its effects in 1992. Even then, he had suspicions that the drug’s mechanism of action — blocking the entry of immune cells into the nervous system — might also make patients more vulnerable to infections. Indeed, PML is an infection that usually affects people whose immune systems are compromised.
“It was a shocking development that a drug that had so much promise and so many potential benefits ended up causing two deaths and one very serious injury,” said Steinman. “It is kind of a cruel Greek drama, something that may be more beneficial than anything yet developed for multiple sclerosis, but yet may be far more dangerous than those other approved drugs.”
The FDA withdrew Tysabri only three months after its approval. The FDA is now considering re-approving the drug. On March 7 and 8, an FDA advisory panel is meeting about the possibility of bringing back Tysabri as a single therapy (in the trials, it was combined with another drug).
“I predict it will come back with really hellacious warnings,” said Steinman. “I think the right course would be to have it undergo more testing, but I don’t think that is practical or fair to patients; they ought to have the opportunity to decide with their physicians if they are willing to take the one in a thousand risk of dying.”
But Steinman and Langer-Gould expressed reservations about the drug returning to the market. They noted that its effects, while impressive, are in general not much better than what is seen with other available drugs: The risk of relapse dropped from an average of two relapses every three years using other approved multiple sclerosis drugs to one every three years with Tysabri.
“Do you want to expose someone to the risk of death for eliminating one relapse every three years?” said Steinman. “I say no.”
“I’m not sure if it is wise to re-approve it,” added Langer-Gould. “The question is, will the FDA rise to the occasion and admit their mistake and try to prevent future mistakes or are they going to ignore it?”