Dementia | Progress in Treatments for Severe Neurological Disease

Session Focus

Selkoe opened by setting the clinical baseline. Lecanemab was approved in 2023 and donanemab in 2024, the first disease-modifying agents for Alzheimer’s disease, now approved in many countries. Mass General Brigham neurology is pursuing two missions in parallel: learning to use the current agents most effectively and safely, and developing new ones.

He noted the scale of that clinical experience: more than 30 cognitive neurologists have now treated over 1,100 to 1,200 patients across all ApoE genotypes.

The conversation then moved to blood-based biomarkers, what presymptomatic treatment would require operationally, and where the field’s remaining gaps sit.

Why the Field Acts Too Late

John Sims MD offered the framing that anchored much of the session. Treating Alzheimer’s disease today is like waiting for everyone to have a first heart attack or stroke before anyone checks blood pressure or cholesterol. The clinic acts once a patient is symptomatic.

What he most wants in the next two years is for the field to look for disease well before symptoms, with therapies capable of preventing people from becoming symptomatic.

Reisa Sperling MD agreed and went further, saying she expects a disease-modifying therapy approved for people without symptoms within two years. She pointed to three large trials underway: Trailblazer, led by Lilly; the AHEAD study, an NIH and public-private partnership; and Proventron from Roche. She described them as complementary and hopes all three succeed, because consistency across them is what would change how the world thinks about the disease.

Donna Masterman MD, a behavioral neurologist with more than 30 years in the field split between academic medicine and biotechnology, identified the convergence as what excites her: biomarkers allowing earlier and more accurate biological detection arriving alongside therapies directed at underlying pathophysiology rather than symptoms.

She recalled wishing in clinic for a simple blood test, particularly because the earliest cases are often the hardest to figure out. The goal she named is moving care from reactive late-stage management to proactive early diagnosis and treatment.

Blood-Based Biomarkers and Risk Prediction

Sims described p-tau 217 as already approved to help physicians diagnose and treat symptomatic Alzheimer’s disease, with active research using it to identify people without symptoms. Lilly is running a trial enrolling exclusively on the basis of the blood test.

His analogy was again cardiovascular. The disease starts 10 to 20 years before symptoms for some people, much as cholesterol accumulates silently. He believes p-tau 217 contains information the field does not yet understand, including signal about who progresses in the short versus long term.

A figure for perspective: roughly 3 percent of 55-year-olds already have the disease process underway, rising from there, with age the dominant risk factor and family history contributing genetic risk on top. He expects the field to reach a 10-year risk algorithm comparable to the cardiovascular one.

Sperling said that algorithm is already coming into focus. A paper her group published this summer in JAMA found that people with very high levels on the blood test carry a 78 percent risk of developing mild cognitive impairment or dementia within 10 years. She noted how unusual that is: very few people on cardiovascular risk scales exceed 50 percent.

Her proposal was concrete. Test everyone at 65 on entering Medicare, with repeat intervals of three or five years depending on family history. She considers it a better predictor over a decade than cholesterol or hemoglobin A1c.

She attached an important caveat. The testing only makes sense if something can be done about the result, so she does not want people running out for the test now. The intervention would start with an anti-amyloid antibody, possibly dosed more efficiently in people without symptoms, and ultimately move to less expensive options including active immunization, oral agents or combinations.

Whether to Genotype

Tanzi raised the audience question of whether ApoE genotyping should be performed at a certain age, noting the bipartisan ASAP bill, the Alzheimer’s Screening and Prevention Act, under consideration in Congress, which would address availability of p-tau 217 and potentially ApoE genotyping.

Sperling gave a qualified no. ApoE tells you that you are at risk, but once a p-tau 217 level is known, ApoE adds only a little about whether a person will progress within a given timeframe. Knowing ApoE status matters for therapeutic decisions. If she could order only one test, it would be p-tau 217.

Selkoe pressed on primary prevention. An E4 homozygote will develop Alzheimer’s pathology, even presymptomatically, at close to 100 percent. If safe and affordable preventive drugs existed, would it not be worth knowing?

Sperling agreed in that case, and added the timing data. P-tau 217 begins rising about seven years before a person becomes amyloid positive. E4 homozygotes rise roughly 14 years earlier and heterozygotes about nine years earlier than non-carriers. Her conclusion was that both pieces of information are needed, because roughly 92 percent develop pathology at a given age and the early accumulation is what you want to reach.

Masterman agreed primary prevention is the direction, while saying the field is not there yet. Moving earlier into populations living healthy, active and productive lives raises the bar on safety and ease of administration.

Remaining Gaps

Sperling was most animated about shuttle technology for crossing the blood-brain barrier, which she expects to reshape every central nervous system therapy rather than anti-amyloid drugs alone.

She also named what the field is not currently doing: making neurons and synapses work better. Clearing amyloid and tau leaves that gap untouched.

Selkoe raised the question of whether ARIA risk will prove lower in prevention trials than it was in the MCI and mild Alzheimer’s trials, which bears on how strong the efficacy signal can be when treatment begins presymptomatically.

Tanzi raised the cost problem directly. In a perfect world everyone over 60 would get a p-tau 217 test and those needing amyloid reduction would receive immunotherapy, but that is currently cost prohibitive, which makes safer and more affordable options the practical path to early detection and intervention.

Key Takeaways

1. The field still waits for symptoms. Sims compared it to waiting for a first heart attack before checking cholesterol.

2. Presymptomatic approval may come within two years. Three large prevention trials are running in parallel.

3. P-tau 217 is a strong decade-scale predictor. High levels carry a 78 percent 10-year risk of MCI or dementia.

4. Testing only makes sense with an available intervention. Panelists were careful not to encourage testing ahead of treatment options.

5. ApoE adds less than expected once p-tau is known, though it remains relevant for therapeutic decisions and for primary prevention timing.

6. Crossing the blood-brain barrier is the next platform shift, with implications well beyond Alzheimer’s disease.

7. Restoring neuronal and synaptic function remains unaddressed. Clearing pathology does not by itself repair what is lost.