A 65-year-old woman who was receiving a promising experimental treatment to slow the cognitive decline caused by her early Alzheimer’s disease recently died from a massive brain hemorrhage that some researchers link to the drug. The clinical trial death, described in an unpublished case report Science has obtained, is the second thought to be associated with the antibody called lecanemab. The newly disclosed fatality intensifies questions about its safety and how widely lecanemab should be prescribed if ultimately approved by regulators.

The woman, who received infusions of the antibody as part of the trial, suffered a stroke and a type of swelling and bleeding previously seen with such antibodies, which bind to and remove forms of amyloid-beta, a protein widely theorized to cause Alzheimer’s. After the stroke was diagnosed in an emergency room at Northwestern University Medical Center in Chicago, she was given a common intervention, the powerful blood-clot busting medication tissue plasminogen activator (tPA). Substantial bleeding immediately followed throughout her brain’s outer layer immediately followed, and the woman died a few days later, according to the case report.

Rudolph Castellani, a Northwestern neuropathologist who studies Alzheimer’s and conducted an autopsy at the request of the patient’s husband, called the case “very dramatic.” The report, co-authored by Castellani, concluded that the woman, like the other person whose death was linked to lecanemab, had amyloid deposits surrounding many of her brain’s blood vessels. This pre-existing condition, found in both Alzheimer’s patients and to a lesser degree in the general population, frequently goes undetected other than by autopsy. It likely contributed to her brain hemorrhage after biweekly infusions of lecanemab inflamed and weakened the blood vessels. The vessels apparently burst when exposed to tPA—known to cause brain bleeds even in some conventional stroke cases.

“It was a one-two punch,” Castellani says. “There’s zero doubt in my mind that this is a treatment-caused illness and death. If the patient hadn’t been on lecanemab she would be alive today.” (Castellani says his comments reflect personal views and were not reviewed or approved by Northwestern. The patient’s husband told Science he authorized Castellani to speak publicly about his wife’s case. Science agreed to withhold both names to protect the family’s privacy.)

Castellani, his co-authors, and other researchers say the newly disclosed death suggests that tPA and perhaps other, less potent blood thinners pose safety considerations for Alzheimer’s patients receiving the antiamyloid antibody drugs, including lecanemab. The 30-page consent form for trial participants, obtained by Science, carries this warning about blood thinners: “You may continue with these medications, but you and the investigator should discuss the risk of bleeding since medications which prevent clots and [lecanemab] are both associated with a slight risk of bleeding in the brain.” It does not address tPA directly.

The woman’s husband says the events surrounding her death were fully disclosed to Great Lakes Clinical Trials, the Chicago-based contract research organization that administered lecanemab to his wife as part of the antibody’s international, multicenter study. He attended a postmortem meeting with Great Lakes principal investigator and psychiatrist Jeffrey Ross, and a Northwestern physician who had participated in his wife’s stroke care. According to the husband, Ross said during the meeting that he had shared details of the case with Eisai Co., the Japanese company that originally developed lecanemab with the Swedish firm BioArctic and sponsored the trial with its U.S. biotech partner Biogen. Ross did not respond to requests for comment.

Eisai declined to comment on the woman’s case, including whether it knew about the death. “All the available safety information indicates that lecanemab therapy is not associated with an increased risk of death overall or from any specific cause,” the company said in a statement to Science. It also declined to describe any other deaths in the woman’s trial, citing the need to protect the privacy of participants, although it said there was no greater frequency of deaths among lecanemab-treated people in an earlier, smaller trial that had a placebo group. Any safety information is reported to regulators and the trial’s principal investigators, Eisai’s statement noted.

This week, Eisai is scheduled to provide the first detailed account of the phase 3 trial, known as Clarity AD, which enrolled about 1800 people with signs of early Alzheimer’s disease. Outside scientists have been eager to scrutinize the data to assess whether the modest benefits announced in September in a brief press release—people on lecanemab had less amyloid in the brain and 27% less cognitive decline than participants who received a placebo over an 18-month period—hold up and outweigh any safety concerns.

If approved, lecanemab would be the second antiamyloid drug to reach the clinic. Aducanumab, a recently approved drug for Alzheimer’s disease also from Eisai and Biogen, is now on the market as Aduhelm. But many Alzheimer’s researchers question the evidence that aducanumab works and were surprised it received U.S. Food and Drug Administration (FDA) approval.  

Now some urge caution about its potential successor. “[Regulators] should take this case report seriously into account, because we’re talking about significant side effects,” says Andreas Charidimou, a neuroscientist at Boston University who examined the report on the woman’s death for Science. “When there’s so many unknowns it’s better to be more conservative.”

“Her body was on fire”

When speaking to Science last week, the dead woman’s spouse was at times overcome by emotion as he described a shattering and chaotic scene after his wife entered the emergency department with stroke symptoms. He informed the doctors about the lecanemab trial and contacted Great Lakes, which provided a web link to information about the antibody. After the attending physician reviewed the website, the care team proposed infusing tPA, describing its risks as relatively small.

“As soon as they put it in her, it was like her body was on fire. She was screaming, and it took like eight people to hold her down,” the husband says. “It was horrific. Everybody’s running in and [asking] ‘What the hell is going on?’” His wife was sedated and moved into intensive care, he adds. A priest came to deliver the “Anointing of the Sick” prayer.

Soon after, the husband says, his wife suffered a series of seizures and was placed on a ventilator. A few days later, the family approved disconnecting the device, and she died. Her doctors told him they had never before seen such massive bleeding under those circumstances, and that they wanted to write up the details of the case for a medical journal.

Lecanemab targets a soluble, “protofibril” version of amyloid-beta, and also binds—albeit more weakly—to the extracellular amyloid deposits known as plaques that are a hallmark of Alzheimer’s disease. Other antibodies, including Aduhelm, bind to those plaques more strongly. Many physicians and Alzheimer’s experts already advise against combining Aduhelm and blood thinners.

STAT recently reported that an 80-year-old man in lecanemab’s phase 3 trial died of a brain hemorrhage linked to possible interaction between the experimental antibody and the blood thinner apixaban, sold under the name Eliquis. Doctors commonly prescribe the drug for atrial fibrillation—an irregular heart beat that can lead to stroke or heart failure. According to STAT, Eisai conceded in adverse event reports for the trial that its drug could have played a part in the fatal brain hemorrhage. (Such reports, submitted to FDA by family members, doctors, or others, are not regarded as proof that a tested therapy caused the event.) But the company later called the death unrelated and STAT said the case remains under investigation.

The woman who died after receiving tPA for her stroke had remained physically active throughout the lecanemab trial. She had retired about a year ago from a professional job that required sophisticated communication and analytic skills, but she had maintained civic involvement.

During the 18-month core trial, the woman might have received either the antibody or the placebo. But there was no question that she was given the antibody during the month preceding her death as part of an open-label trial extension, in which participants who want to take the experimental medication can do so.

Several physicians and researchers not involved in the trial or the woman’s care reviewed the case report at Science’s request and concurred with its findings that lecanemab likely contributed to her death. They said the woman probably received the placebo during the first part of the trial, because the inflammation seen in her blood vessels typically occurs within the first weeks of treatment with antiamyloid antibodies. Her husband agrees, recalling that during the core trial, she did not develop the headaches that she experienced after each infusion of the drug during the trial extension.

In its statement, Eisai said, “It can be difficult to determine what contributed to the death in any given patient, in particular when they are elderly, have multiple medical problems and may have recently received a concomitant treatment or intervention for an acute condition.”

James Nicoll, a neuropathologist at the University of Southampton and a Biogen consultant, was among those asked to review the report. He says although no single case provides proof of a harmful side effect, this death exposed “a very legitimate concern.” Nicoll called the combined use of lecanemab and blood thinners “something you would want to keep a close eye on” should the antibody achieve approval and wide use.

What will FDA decide?

Eisai and Biogen made headlines in September when they announced that lecanemab had more clearly slowed cognitive decline in people with early Alzheimer’s than any previous amyloid-targeted therapy. The conclusion was based on a standardized clinical measure of dementia that draws on observations of memory, judgment, personal care, and other factors by patients, family members, caregivers, and health professionals. But clinicians disagreed about whether the modest slowing would translate into benefits perceptible by patients or their loved ones.

As in other trials of antiamyloid antibodies, many of the people who received lecanemab in the phase 3 trial experienced amyloid-related imaging abnormalities (ARIA)—a term that describes brain swelling and bleeding. ARIA occurred in more than 21% of those who were on the drug; 17% experienced brain bleeds, but none of the ARIA cases was life-threatening, according to the Eisai and Biogen press release.

Still, one reason to think lecanemab contributed to the woman’s death is that her autopsy revealed widespread cerebral amyloid angiopathy (CAA), a condition in which amyloid deposits gradually replace the smooth muscle of blood vessel walls. Castellani, Nicoll, and others who reviewed her case suspect CAA made her blood vessels vulnerable to weakening when lecanemab did what it is expected to do: strip amyloids from the brain. The tPA treatment then likely ruptured those weakened vessels, leading to serious ARIA—and apparently fatal brain bleeding, according to the Northwestern report authors and independent CAA or Alzheimer’s experts.

Nearly half of Alzheimer’s patients also have CAA—including, according to STAT, the man whose death was previously linked to combining lecanemab and blood thinners. Eisai screened potential trial participants with tests often used to detect moderate or serious CAA. For example, applicants had MRI brain scans and anyone whose scan showed more than four “microhemorrhages”—tiny bleeds—or other signs of possibly serious CAA were not allowed to enroll. But the condition can be hard to detect says Charidimou, who studies CAA. The two deaths show that even in the trial population, some patients with serious CAA slipped through.

Many Alzheimer’s patients with CAA also suffer from other ailments, such as atrial fibrillation, that are normally treated with blood thinners, says Matthew Schrag, a Vanderbilt University physician and neuroscientist who specializes in CAA and assessed the report on the woman’s death for Science. That means both risk factors are likely to be present in a significant fraction of the patients who might be prescribed lecanemab if it is approved and marketed widely. Training physicians to interpret tests for CAA will be key to ensure that vulnerable Alzheimer’s patients and their caregivers are properly informed and warned about the possible hazards of lecanemab, Charidimou says.

Assuming a close look at Clarity AD’s data validates the companies’ press release, lecanemab could still help early Alzheimer’s patients who don’t have moderate or severe CAA, says University of Kentucky neuroscientist Donna Wilcock, who also reviewed the report on the death. “Even if that [only] means 6 or 12 more months of knowing who their kids are … it’s meaningful to dementia patients and their families.”

But FDA should require meticulous screening for CAA and a warning against concurrent tPA use, Wilcock adds. “If lots of people [on lecanemab] who go into the hospital with garden-variety strokes end up dying, that sets the field back decades.” FDA has promised to make a decision on whether to approve lecanemab, and impose any conditions on its use, by 6 January 2023.

This story was supported by the Science Fund for Investigative Reporting.