Emerging Treatments | Alzheimer’s Disease

Session Focus

Zemansky noted at the outset that panelist Brad Dickerson had a family event arise and could not attend, leaving Dickson to carry the full session.

The discussion covered what the first generation of disease-modifying therapies has actually demonstrated in practice, the operational burden they impose, what would define a better next generation, and how far treatment can move toward prevention.

What the First Generation Demonstrated

Dickson described lecanemab and donanemab, both with full FDA approval, as removing amyloid from the brain in clinical trials and in routine use across the Mass General Brigham system.

On efficacy he was careful. Trials demonstrated, and the patients he follows are beginning to show, some response in slowing cognitive decline. It does not stop or reverse cognitive impairment, but it does appear to slow it in at least some patients. He called that a paradigm shift, the first disease-modifying therapy for a devastating illness.

The lessons learned concern patient selection, particularly using biomarkers to confirm that a patient has accumulating amyloid available to remove.

On ARIA, the amyloid-related imaging abnormalities involving swelling or bleeding in the brain, he was direct about both the concern and the real-world experience. Patients are very concerned when it is discussed in clinic, and considerable worry has come from trials and case reports. In routine practice, with careful patient selection and appropriate monitoring, it has proven quite safe. He stated that there have been no patient deaths from this therapy in the Mass General Brigham system.

The Operational Burden

Dickson identified administration and monitoring complexity as the major limitation of current therapy.

Both drugs launched as infusion-only, requiring extensive MRI monitoring particularly in the first six months to evaluate for ARIA.

The recent approval of subcutaneous lecanemab for initiation changes the capacity constraint, since home administration removes the dependence on infusion center availability.

ARIA monitoring remains complex, requiring schedule management across therapy timing and MRI availability, which he called a significant challenge in coordination.

His two wishes for improvement:

Easier monitoring, ideally a blood test rather than MRI.

Better risk stratification, so that a particular combination of genetics and MRI features could identify patients at low enough ARIA risk to forgo routine MRI unless symptoms develop, while concentrating close monitoring on genuinely high-risk patients.

What a Better Next Generation Looks Like

Asked what would define success for the next generation, Dickson ranked the possibilities.

Faster amyloid removal would be ideal. Improved efficacy would be outstanding. But the most realistic expectation, and in his view the most desirable, is an improved safety profile, because that reduces the monitoring burden on patients.

He pointed to the transferrin receptor shuttle as early cause for hope.

On why that matters beyond Alzheimer’s, he framed the blood-brain barrier as a significant problem for delivering biological agents including antibodies and oligonucleotides into the central nervous system. Shuttle methods could ensure drugs reach the right place within the CNS while reducing systemically required doses, which in turn reduces potential systemic side effects.

He added a specific mechanistic link. Given the dose-dependent relationship between anti-amyloid treatment and ARIA, a lower dosing regimen may itself be partly responsible for reduced ARIA risk.

Have Anti-Amyloid Antibodies Reached Their Limit

Dickson reported that most patients respond at least from a biomarker standpoint, with substantial amyloid reduction. A handful in clinical practice show only partial response or are non-responders for amyloid clearance, and the field does not yet understand why. Identifying the patient characteristics that limit clearance would help route those patients toward other therapies.

On the more important limit, he identified timing. Current FDA approval covers early symptomatic Alzheimer’s with mild cognitive impairment or mild dementia severity. Several prevention trials are underway in pre-symptomatic patients who are cognitively normal or have very mild subjective symptoms.

How far the boundary can be pushed toward a preclinical stage and primary prevention is, in his assessment, the most significant open question in the field right now.

Tau and the ASO Data

Zemansky raised the mid-year phase two results for the tau antisense oligonucleotide, noting the observation that while some people with Alzheimer’s have limited amyloid burden, it is difficult to find anyone with the disease who lacks substantial tau burden.

Dickson acknowledged the formal endpoints were not met in terms of dose response for cognitive outcomes.

His reading of why was appropriately uncertain. One possibility is that a maximum response is already achieved at the lower dose, which would explain a positive cognitive signal even there. If so, that provides reason for hope about the phase three trial.

He was clear about what remains unproven. The agent demonstrates target engagement, reducing both CSF tau and tau PET levels. Whether that translates into clinical benefit requires a well-powered phase three trial.

Key Takeaways

1. The first approved therapies slow cognitive decline without stopping or reversing it, which still represents a paradigm shift.

2. ARIA has proven manageable in practice, with no patient deaths from the therapy in the Mass General Brigham system.

3. Monitoring complexity is the limiting operational factor, dependent on MRI scheduling rather than drug availability.

4. Subcutaneous administration removes the infusion capacity constraint by allowing treatment at home.

5. Safety, not efficacy, is the most realistic improvement target for the next generation.

6. Blood-brain barrier shuttles could lower required doses, which may itself reduce ARIA given the dose-dependent relationship.

7. How early treatment can begin is the field’s most significant open question, with prevention trials now underway.