The Next Drug

Robert Cheng has read the informed consent document twice. He is 69, a retired civil engineer, and he approaches legal documents the way he approached structural calculations: line by line, noting what is stated and what is omitted. Diane, his wife of 41 years, has read it once and set it down on the kitchen table between the coffee cups and the pen she brought in case they decided to sign tonight. Their daughter Lisa is on the phone from Seattle, her voice on speaker, filling the silence that formed after the last paragraph.

Robert was diagnosed with early Alzheimer’s eight months ago. He carries the APOE4/APOE4 genotype, the highest genetic risk configuration. The consent document in front of him is for a Phase III anti-tau trial at a university medical center 140 miles from their home. The trial involves biweekly infusions for eighteen months, MRI monitoring every three months, cognitive testing every six weeks, and a one-in-three chance of receiving a placebo. Robert has made his list of questions. He has checked them off, one by one, as his neurologist answered them. The last question, which Lisa asked from Seattle, is the one nobody at the medical center answered: “What if it works and you can’t afford it afterward?”

The question is real. It deserves a real answer. This piece is, in part, that answer.

Where the Amyloid Story Stands

#

Lecanemab and donanemab are the first FDA-approved therapies that target the amyloid plaques that accumulate in the brains of people with Alzheimer’s disease. They work. They remove amyloid. The clinical question is what that removal accomplishes.

Both drugs have demonstrated statistically significant slowing of cognitive and functional decline in Phase III trials. Lecanemab slowed decline by approximately 27 percent over eighteen months compared to placebo in the Clarity AD trial. Donanemab slowed decline by approximately 35 percent in the TRAILBLAZER-ALZ 2 trial among patients who achieved amyloid clearance. These are real effects measured by validated instruments. They are not cures. They are not stabilization. They are a measurable reduction in the rate of decline.

The side effect profile is significant. Amyloid-related imaging abnormalities, known as ARIA, occur in a meaningful fraction of patients, with higher rates in APOE4 carriers. ARIA can involve brain swelling or microbleeds, is usually asymptomatic and resolves, but can be serious. Robert’s APOE4/APOE4 genotype places him at higher ARIA risk, which some trials treat as an exclusion criterion and others treat as a reason to include him with enhanced monitoring.

The cost is $26,500 per year for lecanemab before insurance. Medicare covers it with restrictions. The infusion requirement means regular visits to an infusion center, which for Robert and Diane means a 140-mile drive.

The research community now understands what two decades of amyloid-focused trials have taught: amyloid accumulation is one part of the pathological process, not the whole of it. Tau tangles, neuroinflammation, synaptic loss, and vascular changes each contribute independently to the neurodegeneration that produces symptoms. Removing amyloid slows the process. It does not stop it. The next generation of therapeutics targets the rest.

Anti-Tau Therapies

#

Tau tangles are the second major hallmark of Alzheimer’s pathology. Where amyloid plaques accumulate between neurons, tau tangles form inside them. The correlation between tau burden and cognitive symptoms is stronger than the correlation between amyloid burden and symptoms, which is why the research community has long suspected that tau-targeting therapies might produce more noticeable functional benefits than amyloid-targeting ones.

Multiple anti-tau therapies are in Phase II and Phase III trials. The mechanisms vary: some aim to reduce tau production, others to prevent tau aggregation, others to clear existing tangles. The most advanced programs expect primary results within the next one to two years.

The honest uncertainty is real. The amyloid trials took decades from first hypothesis to FDA approval. The tau trials are earlier in that trajectory. The biology is promising. The clinical translation has not yet been proven at scale. Robert’s trial is one of the programs that will produce the answer.

The Metabolic Connection

#

The connection between metabolic health and Alzheimer’s risk has been documented across dozens of studies over two decades. Type 2 diabetes approximately doubles the risk of developing Alzheimer’s. Insulin resistance, even without diabetes, is associated with accelerated amyloid accumulation and cognitive decline. The mechanism involves insulin signaling in the brain, which affects both amyloid clearance and tau phosphorylation.

GLP-1 receptor agonists, the class of drugs that includes semaglutide (marketed as Ozempic for diabetes and Wegovy for weight loss), are now in clinical trials for cognitive protection. The EVOKE trial is testing semaglutide directly in patients with early Alzheimer’s. Other trials are testing GLP-1 drugs for cognitive protection in people with metabolic risk factors who do not yet have a dementia diagnosis.

The implication if results are positive is different from every other drug in this piece. GLP-1 drugs are already approved. They are already manufactured at scale. Their safety profile is established across millions of patients. If semaglutide proves effective for cognitive protection, it would not require the decade-long development timeline of a new molecular entity. It would require a label expansion and a coverage decision. The timeline from positive trial results to availability could be measured in months rather than years.

The cost and access question would still matter. Semaglutide is expensive, though less so than lecanemab. Supply constraints have affected GLP-1 availability for diabetes patients. But the infrastructure for manufacturing, distributing, and prescribing the drug already exists. This is Lisa’s question, partially answered: a repurposed drug with existing infrastructure would reach patients faster and at lower cost than a novel biologic.

Neuroinflammation and Combination Therapy

#

The brain’s immune response to Alzheimer’s pathology is emerging as a third therapeutic target. Microglia, the brain’s immune cells, activate in response to amyloid and tau accumulation. The inflammatory cascade they trigger may drive more neurodegeneration than the plaques and tangles themselves. TREM2-targeting therapies and complement inhibition approaches are in early trials. These are genuinely early-stage: Phase I and early Phase II. They are included here not because they are near but because they represent the direction.

The combination therapy argument draws on oncology’s hard-won lesson. Cancer treatment moved from single-agent to combination therapy after recognizing that single-pathway targeting allowed disease escape. A tumor that shrinks under one drug often finds an alternative growth pathway. Attacking multiple pathways simultaneously closes the escape routes.

Alzheimer’s researchers are applying the same logic. The modest effects of amyloid-only therapy may reflect the same dynamic: removing amyloid while tau, neuroinflammation, and vascular pathology continue unchecked allows the disease to progress through alternative mechanisms. The first combination trials, targeting amyloid plus tau or amyloid plus neuroinflammation, are in early stages. Results are years away. The concept is sound and the precedent from oncology is strong.

The Trial Question

#

Robert’s specific situation: a Phase III anti-tau trial, APOE4/APOE4 genotype, 140-mile drive, eighteen months of biweekly infusions. What participation involves and what it provides.

Screening takes one to two visits. Randomization means a one-in-three chance of receiving placebo in his trial, which is standard for Phase III studies where no approved comparator exists for the specific mechanism. He will not know whether he is receiving the drug or the placebo until the trial concludes or until unblinding occurs. The biweekly infusions take two to four hours each. The MRI monitoring every three months adds a day trip to the medical center. The cognitive testing every six weeks takes approximately ninety minutes.

What participation provides: access to a potentially effective treatment that is not otherwise available. Careful, protocol-driven monitoring that exceeds standard clinical care. The knowledge that his data, whether he receives drug or placebo, will inform the decision for every family that follows. For Robert specifically, some trials exclude APOE4/APOE4 carriers due to elevated ARIA risk. Others specifically include them because the genetic risk population is where the unmet need is greatest and where the treatment, if effective, produces the most meaningful benefit. His trial is one of the latter.

The practical burden is real. The 140-mile drive, biweekly, for eighteen months, is 280 miles round trip, 26 times per year, plus monitoring visits. Diane will drive. They have budgeted the gas. They have not budgeted the fatigue.

Lisa’s Question

#

“What if it works and you can’t afford it afterward?”

The question has an honest answer, and the honest answer is that the pipeline and the access system are separate questions.

Lecanemab costs $26,500 per year. Medicare Part B covers it with prior authorization, and most Medicare Advantage plans have followed. The out-of-pocket cost for a Medicare beneficiary is roughly $5,300 per year before supplemental coverage. For a family like the Chengs, with retirement savings and supplemental insurance, the cost is manageable but not trivial. For a family without supplemental coverage, the cost is a barrier.

If Robert’s anti-tau trial succeeds and the drug reaches FDA approval, its pricing will follow the biologic pricing model that produced lecanemab’s $26,500. Novel biologics are expensive because the development costs are high, the patient populations are defined, and the pricing power exists in a market where the alternative is progression.

If GLP-1 drugs prove effective for cognitive protection, the pricing picture changes. Semaglutide’s current list price is roughly $1,000 per month, but it is manufactured at scale, and generic competition for the class is approaching. A repurposed drug with existing manufacturing infrastructure would cost less than a new biologic. It would still not be cheap.

The honest answer to Lisa’s question: the pipeline is advancing with genuine momentum. The access system has not kept pace. If the next drug works, whether it reaches her father depends on coverage decisions that have not been made, pricing negotiations that have not occurred, and a political willingness to prioritize access that has not been tested. The pipeline and the access system are separate problems. Lisa’s family is facing the first one now. The second one will come when the first one is answered.

Robert picks up the pen. He and Diane have discussed it. He signs the consent document. Lisa, still on speaker from Seattle, says nothing for a moment. Then she says she will drive down for his first infusion. Diane says she was going to ask.