The findings provide rare evidence from the human brain that clearing amyloid plaques may help protect against other harmful processes involved in Alzheimer’s disease. They also suggest that thorough amyloid removal, particularly early in the disease, could help limit tau accumulation and slow the progression of brain damage.

Researchers from the Perelman School of Medicine at the University of Pennsylvania presented the findings at the 2026 Alzheimer’s Association International Conference.

“Seeing both disease patterns side by side in the same brain gives us a unique opportunity to understand how amyloid clearance affects other proteins that contribute to Alzheimer’s disease,” said co-senior author David Wolk, MD, PhD, co-director of the Penn Memory Center and director of the Penn Alzheimer’s Disease Research Center.

“This study provides some of the clearest human evidence to date that anti-amyloid therapy can limit tau accumulation and slow brain changes that lead to memory loss and cognitive decline.”

How amyloid and tau contribute to Alzheimer’s disease

Two proteins play major roles in Alzheimer’s disease. Beta-amyloid collects outside nerve cells and forms sticky deposits known as plaques. Tau accumulates inside neurons and forms abnormal tangles. Both are associated with progressive brain-cell damage, memory loss, personality changes, and difficulty thinking.

Treatments designed to remove amyloid plaques have changed how Alzheimer’s disease is managed. Aducanumab was the first drug of its type to receive accelerated approval from the U.S. Food and Drug Administration, although it is no longer commercially available. Newer amyloid-targeting drugs, including donanemab (Kisunla) and lecanemab (Leqembi), have since received full FDA approval.

Researchers have long believed that amyloid buildup helps create the conditions for harmful tau tangles to develop. However, it has been much harder to determine whether removing amyloid can limit tau accumulation in the human brain.

Direct evidence remains limited because relatively few people who received amyloid-targeting drugs have undergone postmortem brain examinations.

Rare postmortem brain examination reveals a striking pattern

Researchers investigated the question using brain tissue donated by a former Alzheimer’s clinical trial participant.

The man, who was in his 50s, had been diagnosed with mild cognitive impairment due to Alzheimer’s disease. He participated in a clinical trial of aducanumab and received 30 doses over four and a half years.

He died approximately four years after his final dose and more than a decade after his first symptoms appeared. Before his death, he expressed a desire to donate his brain for scientific research. The tissue was provided to the Neurodegenerative Disease Research Center in the Department of Pathology and Laboratory Medicine.

When researchers examined the donated brain, they found dramatically different levels of Alzheimer’s-related damage in adjacent regions.

The raised ridges on the folded surface of the brain, known as gyri, had lost most of their amyloid plaques and contained little accumulated tau. In contrast, the adjacent grooves between the ridges, called furrows, still contained substantial amyloid. These regions also had significantly more tau tangles, inflammation, and degeneration of brain tissue.

This patchwork distribution was consistent with brain scans taken while the participant was alive. Those scans had already shown that amyloid clearance can vary considerably across different parts of the brain.

“Many previous reports have shown either widespread amyloid clearance or limited clearance,” said co-senior author Edward B. Lee, MD, John Q. Troyanowski, MD, Professor of Neurodegenerative Disease Research in Pathology and Laboratory Medicine.

“This case represented a unique ‘Goldilocks’ scenario in which amyloid was cleared in some areas but not in others. This allowed us to directly compare what happened next in adjacent brain regions and better understand the relationship between amyloid, tau, and neurodegeneration.”

Could amyloid-clearing treatments protect the brain for years?

The discovery could have important implications for newer Alzheimer’s drugs that remove amyloid through a mechanism similar to the treatment received by this participant.

Although the study involved only one individual, the contrast between different brain regions suggests that substantial amyloid clearance may affect other disease processes long after treatment ends.

The findings also raise important questions about how researchers measure the effectiveness of Alzheimer’s treatments. Clinical trials typically follow patients for about 18 months, but some protective effects may take much longer to become apparent.

However, the study cannot establish how much amyloid must be removed to provide lasting protection. It also cannot prove that plaque removal directly caused the differences in tau accumulation and brain damage between the regions.

Researchers now hope to understand why some parts of the brain respond more effectively to amyloid-removal treatments than others. Identifying and potentially strengthening the biological mechanisms involved in plaque clearance could lead to improved therapies.

Another major question is whether starting treatment before widespread tau tangles develop can significantly slow Alzheimer’s progression or even prevent its onset.

Why early Alzheimer’s treatment may matter

Amyloid can build up silently in the brain for years before memory loss begins. As a result, researchers are increasingly studying whether treatment during the earliest stages of Alzheimer’s disease, before noticeable symptoms appear, could provide greater benefits.

“Patients can accumulate amyloid for years before symptoms appear,” said lead author Christopher A. Brown, MD, assistant professor of neurology. “This case suggests that removing amyloid early may help limit changes that ultimately damage brain cells. Ongoing trials will help determine whether starting treatment before symptoms begin can provide even greater benefits.”

This research was supported by the National Institutes of Health through grants P30-AG072979, RF1-AG069474, R01-AG056014, R01-AG055005, R01-AG072796, R01-AG065341, and P01-AG084497.