Four Years to Get a Memory Test: An Early-Onset Alzheimer's Diagnosis at 60 | Sean Terwilliger

Sean Terwilliger noticed memory and word-finding problems after a mini-stroke in 2018. He was 56. Over the next four years, he saw four primary care physicians across three states. All of them told him he was fine. No cognitive screening, no referral.

What finally got him tested: a doctor in Massachusetts administered the MoCA, a standard 30-point cognitive screening that takes about ten minutes. Terwilliger failed by one point. That single point was the entire gateway to a neurologist referral. The neurologist ordered thyroid labs, an MRI, and an amyloid PET scan. The PET showed significant amyloid plaque. He was diagnosed with early-onset Alzheimer's at 60.

After the diagnosis, the neurologist told him the prognosis was "eight to 10 years from diagnosis," told him to come back in two weeks to start infusions, and sent him home. No social worker. No elder-law attorney referral. No financial guidance. Terwilliger describes it as being handed a death sentence and a parking validation.

He went on lecanemab (brand name Leqembi), an FDA-approved drug that removes amyloid plaque from the brain. Two infusions a month for 18 months, 39 infusions total. When the standard course ended, he and his neurologist decided against a monthly maintenance dose. Terwilliger's reasoning drew on data from donanemab (Kisunla, Eli Lilly's similar drug), where maintenance dosing showed no appreciable additional benefit. He acknowledges that's a layman's inference applied across two different drugs, not an established clinical equivalence. His neurologist agreed with the decision for him specifically.

Unable to get a follow-up amyloid PET scan covered by insurance, Terwilliger paid out of pocket for a p-tau217 blood test, a newer biomarker test that measures the ratio of two proteins associated with Alzheimer's pathology. The result: he no longer meets the threshold used to diagnose the disease. If tested today for the first time, he would not qualify for a diagnosis. He calls it remission, not a cure, and plans to retest at least twice a year, expecting his levels to rise again over time.

There's a practical wrinkle in that outcome. Because his biomarkers improved so much, he no longer qualifies for most clinical trials targeting amyloid, tau, inflammation, or microglial activity. All of them require measurable disease burden at entry. Successful treatment has made him ineligible for the research that might help him and others next.

Three things worth pulling out for anyone navigating this:

The MoCA is a ten-minute test your primary care doctor can run in the office. Terwilliger's point is that doctors routinely skip it. If you or someone close to you has raised memory concerns and been told "you seem fine," ask for the MoCA by name. Failing by even one point can be enough to trigger a neurologist referral and a real workup.

The p-tau217 blood test is becoming available outside clinical trials. It's not yet a standard of care and insurance coverage varies, but Terwilliger's case is an early real-world example of using it to track treatment response when a follow-up PET scan is unavailable or unaffordable.

Post-diagnosis support is often absent, and the legal and financial clock starts immediately. When a family member gets a dementia diagnosis, ask the neurologist that day for referrals to a social worker, an elder-law attorney, and a financial planner who handles disability planning. Power of attorney, disability claims, and long-term care decisions all carry deadlines that most neurology practices don't proactively flag.

Sean Terwilliger, diagnosed with early-onset Alzheimer's at 60, spent four years being dismissed by doctors before a single-point failure on a standard cognitive screening finally got him tested properly. After 18 months and 39 infusions of the drug lecanemab (brand name Leqembi), a blood test shows he no longer meets the pathology threshold for Alzheimer's — a remarkable outcome that illustrates both what the new disease-modifying drugs can do and how hard it remains to get diagnosed in the first place.

What was covered

• The four-year diagnostic delay. Terwilliger noticed symptoms after a TIA (mini-stroke) in 2018 — word-finding problems, difficulty with numbers, balance issues. Three states, four primary care physicians, and years of reassurance that he was "fine" followed before a doctor in Massachusetts administered the MoCA (Montreal Cognitive Assessment), a 30-point screening test. He failed by one point, which was sufficient to trigger a neurologist referral.
• The diagnostic workup. The neurologist ordered thyroid blood work (to rule out metabolic causes of confusion), an MRI (to rule out lasting stroke damage), and an amyloid PET scan. The PET showed significant amyloid plaque buildup, leading to an Alzheimer's diagnosis.
• A blunt prognosis and little else. After diagnosis, Terwilliger says he was told the prognosis was "eight to 10 years from diagnosis," instructed to return in two weeks to start infusions, and sent home — no social worker referral, no legal or financial guidance, no roadmap.
• Leqembi treatment — 39 infusions over 18 months. Terwilliger completed the standard course (two infusions per month) and then, in consultation with his neurologist, chose not to continue to a monthly maintenance dose. His reasoning drew partly on data from Kisunla (donanemab, Eli Lilly's similar drug), where maintenance dosing showed no appreciable additional benefit.
• The p-tau217 blood test result. Unable to get a follow-up amyloid PET scan — his neurologist didn't think it was warranted and insurance wouldn't cover it — Terwilliger paid out of pocket for a p-tau217 blood test. The result showed he no longer meets the amyloid/tau ratio threshold used to diagnose Alzheimer's pathology. He describes this as "remission," not a cure, and plans to retest at least twice a year.
• A paradox: too healthy for clinical trials. Because his biomarkers have improved so significantly, Terwilliger no longer qualifies for most clinical trials targeting amyloid, tau, inflammation, or microglial activity — all of which have entry criteria requiring measurable disease burden.
• Writing and advocacy as a survival strategy. Terwilliger credits starting the ALZBlog and writing the book ALZ Fired Up! with pulling him out of post-diagnosis despair. He also coined the Finnish word sisu — meaning one who perseveres stoically when all hope is lost — to describe spousal caregivers, whom he argues face a uniquely heavy burden.

Notable claims & predictions

• Sean Terwilliger: "I was told, 'Generally the prognosis is eight to 10 years from diagnosis. Come back in two weeks and I'll start the process of getting you onto the infusions. Have a good rest of your day.' And that was it." — on the near-total absence of post-diagnosis support.
• Sean Terwilliger: Failing the MoCA by a single point — a 30-point test — was the entire gateway to seeing a neurologist. Without that one point, he says, no further testing would have followed. His implication: countless people who are "still really good" at their worst are being turned away without even this simple screen.
• Sean Terwilliger: His p-tau217 blood test shows he no longer meets the pathological threshold for Alzheimer's after Leqembi treatment — meaning that if tested today for the first time, he would not qualify for a diagnosis. He frames this explicitly as remission, not cure, and expects his levels to rise again over time.
• Sean Terwilliger: After his treatment course, his amyloid levels are so low that he cannot qualify for current clinical trials — a practical consequence of successful drug treatment that the field hasn't fully addressed yet.
• Sean Terwilliger: He applied data from Kisunla's maintenance-dose trial to his own Leqembi decision — reasoning that if one anti-amyloid drug showed no appreciable benefit from maintenance dosing, the same might apply to the other. He acknowledges this is a layman's inference, not a clinical equivalence.

Fact check

The MoCA as a gateway test. Terwilliger's description of the MoCA as a 30-point test that, if failed, routes a patient to a neurologist is accurate. It is a widely validated, brief cognitive screening tool. His broader claim — that doctors routinely decline to offer it — is consistent with well-documented patterns in primary care, where cognitive screening remains inconsistent, though this is a systemic observation rather than a verifiable single fact.

Leqembi dosing. The 18-month, two-infusions-per-month course Terwilliger describes is consistent with the approved Leqembi (lecanemab) regimen. The subsequent maintenance option he mentions (monthly infusions) is also accurate.

The Kisunla maintenance-dose reasoning. Terwilliger acknowledges this is his own layman's inference — applying Kisunla trial findings to his Leqembi decision. This is not an established clinical equivalence. The two drugs have different dosing schedules, mechanisms of amyloid removal, and trial designs. His neurologist apparently agreed with the decision for this patient, but readers should not treat the logic as generalizable without their own physician's input.

"Eight to 10 years from diagnosis." Prognosis in Alzheimer's is highly variable and depends on age at diagnosis, rate of progression, overall health, and other factors. A blanket "eight to 10 years" figure at diagnosis is a rough population-level estimate, not an individual prognosis. Terwilliger doesn't present it as precise; he presents it as what he was told and found inadequate.

APOE genetic testing and Leqembi safety. Terwilliger mentions that APOE (apolipoprotein E — a gene variant associated with elevated Alzheimer's risk) testing was offered to assess bleeding-brain risk from Leqembi. This is accurate: APOE ε4 carriers have a higher risk of ARIA (amyloid-related imaging abnormalities, a potentially serious side effect), and genetic testing is part of the prescribing discussion. He chose to proceed without it, which is a personal decision, not a recommended protocol.

No claims that clearly fail scrutiny — the main caveats are the Kisunla-to-Leqembi inference (his own, flagged as such) and the bluntness of the prognosis figure (presented as anecdote, not medical guidance).

Why this matters for you

• If you or someone close to you has been told "you seem fine" after raising memory concerns, Terwilliger's story is a direct argument for asking your primary care physician — by name — for a MoCA. It takes about 10 minutes. Failing by even one point can be the threshold that unlocks a neurologist referral and a real workup. Don't wait for the doctor to offer it.
• The p-tau217 blood test is new and increasingly available. Terwilliger's case is an early real-world example of it being used outside a clinical trial setting to track treatment response. If you or a family member is on an anti-amyloid therapy (lecanemab or donanemab) and can't get a follow-up PET scan approved by insurance, this blood test may be worth asking about — though it's not yet a standard of care and coverage will vary.
• Post-diagnosis support is often absent. Terwilliger's account of being handed a prognosis and sent home is not unusual. If a family member receives a dementia diagnosis, it is worth immediately asking the neurologist for referrals to a social worker, an elder-law attorney, and a financial planner experienced in disability planning — the diagnosis creates legal and financial urgency (disability claims, power of attorney, long-term care planning) that most neurology practices do not proactively address.
• The spousal caregiver burden is real and distinct. Terwilliger's framing of spousal caregivers as carrying a qualitatively