DRUG DEVELOPMENT

Promising Alzheimer’s therapies shrink brains

Effects of antiamyloid antibody drugs worry some scientists, who call for more study

By Jennifer Couzin-Frankel many consider a driver of the disease. About shrinkage was associated with so-called
half were trials of secretase inhibitors, tra- amyloid-related imaging abnormalities

A
class of drugs meant to slow the ditional small-molecule drugs targeting an (ARIA), a type of brain swelling and bleed-
cognitive decline of Alzheimer’s dis- enzyme that produces beta amyloid from ing that is a well-documented side effect
ease, including one that appears to a larger protein. (Those drugs have been of the antibodies. ARIA occurred in 21% of
have real but modest benefits, can largely abandoned due to ineffectiveness.) the 898 people taking lecanemab in Eisai’s
cause brain shrinkage, a new analy- The others were trials of monoclonal anti- pivotal trial (as well as 9% on a placebo);
sis shows. Although scientists and bodies like lecanemab that directly target most had no symptoms, but some became
drug developers have documented brain various forms of beta amyloid. Another severely ill and at least two died. Ayton and
volume loss in trials of the drugs for years, antiamyloid antibody in the analysis, adu- his colleagues found the experimental ther-
the scientific review, published last week in canumab, was approved in 2021 amid much apies with a higher rate of ARIA also gener-
Neurology, is the first to look at data across controversy, and still others are in trials. ated a bigger average increase in the size of
numerous studies. It also links the the ventricles.

Downloaded from https://www.science.org on November 14, 2024

brain shrinkage to a better known side “I was just shocked when we put
effect of the drugs, brain swelling. these data together,” Ayton says, but he
“We don’t fully know what these sees a possible connection. ARIA shows
changes might imply,” says Jonathan up on brain scans as inflammation, and
Jackson, a cognitive neuroscientist at generally, “it’s not controversial that
Massachusetts General Hospital. But, neuroinflammation would lead to neu-
he says, “It’s likely these changes are rodegeneration,” he says. As a result,
detrimental.” They cast a shadow on the patients who develop ARIA—which
promise of one drug, the antibody lec- tends to happen early in treatment—
anemab, which was granted accelerated might later be more prone to brain vol-
approval in the United States in January ume loss, he speculates.
after a trial showed it slowed the rate of FDA officials who approved Eisai’s ap-
cognitive decline by 27% after 18 months plication for lecanemab noted the brain
compared with people on a placebo. volume changes. Like the company,
The analysis alarmed Scott Ayton, a though, they weren’t unduly concerned.
neuroscientist at the Florey Institute A lead FDA reviewer “questions the clin-
of Neuroscience and Mental Health ical relevance of the changes to whole
in Melbourne, Australia, who led the brain volume and total ventricular vol-
work and has advised Eisai, which ume,” FDA wrote in its summary review
led the development of lecanemab. of the drug, in part because the drug
“We’re talking about the possibility of met its goal of slowing cognitive decline.
brain damage” from treatment, Ayton But researchers note that individual
says. “I find it very peculiar that these patients may have had cognitive or other
data, which are very important, have A CT scan shows the brain of a 78-year-old woman effects not noted across the trial popu-
been completely ignored by the field.” with Alzheimer’s disease. lation. Madhav Thambisetty, a neuro-
A spokesperson for Eisai suggested logist at Johns Hopkins University and
the brain shrinkage could be benign. The Alzheimer’s disease frequently causes the chief of the clinical and translational neuro-
company said that although participants in brain to shrink. But the researchers found science section at the National Institute on
its pivotal trial did experience “greater cor- both types of antiamyloid drugs generally Aging, says it’s not clear “whether brain vol-
tical volume loss on lecanemab relative to caused trial participants to lose brain vol- ume loss and increase in ventricular volume
placebo,” those reductions may be due to the ume faster. In two large lecanemab trials, are associated with worsening clinical and
antibody clearing the protein beta amyloid people on the highest dose—which is the cognitive outcomes” in individual patients.
from the brain and reducing inflammation. one the U.S. Food and Drug Administra- The “incomplete reporting of results … is
But that’s sheer speculation, some scien- tion (FDA) approved—recorded, on aver- very concerning to me.”
IMAGE: ZEPHYR/SCIENCE PHOTO LIBRARY

tists worry. “We don’t know what [the shrink- age, a 28% greater brain volume loss than Ayton shares the sentiment. “The lack
age] means,” says Lon Schneider, director of people on placebo after about 18 months. of information increases my worry,” he
the California Alzheimer’s Disease Center at This translated to a loss of an extra 5.2 milli- says, calling on companies, including Eisai,
the Keck School of Medicine of the Univer- liters (mL) in brain matter. The authors also and drug regulators to examine this issue
sity of Southern California. It’s crucial, he reported that the antiamyloid antibodies— and share more trial data. “Tell us why we
says, to “try to seriously understand this.” but not the secretase inhibitors—led to a shouldn’t be concerned.”
Ayton and two colleagues identified greater increase in the size of the brain’s FDA is planning an advisory commit-
31 published clinical trials of drugs that aim fluid-filled ventricles. tee meeting by July to consider whether to
to eliminate beta amyloid, whose buildup Ayton’s group then studied whether the grant lecanemab full approval. j

SCIENCE science.org 7 APRIL 2023 • VOL 380 ISSUE 6640 19