Newsletter · 2026-09-27 · Ashutosh Agarwal

The Alzheimer's Blood Test Gets Its User Manual - The Neuro & Alzheimer's Pipeline - Week of September 27, 2026

A synthesis of what podcasts and clinicians said about Alzheimer's diagnostics and drug pipelines for the week of September 20-27, 2026, built around a CME masterclass that laid out the blood-test playbook: two FDA-cleared tests for two jobs, p-tau-217 as the recommended marker, an intermediate zone that keeps PET in play, and an academic read on tau versus inflammation.

The Neuro & Alzheimer's Pipeline

Week of September 27, 2026: The Alzheimer's Blood Test Gets Its User Manual

Two weeks ago the story here was tau. Last week it was cholesterol, when a heart-drug CEO said his pill moves Alzheimer's blood markers. This week the podcasts brought no new drug and no new mechanism. What they brought was less exciting and probably more useful: a clear, step-by-step guide to the Alzheimer's blood test, the tool that decides how many patients ever reach one of these drugs.

First, a note on sources, because this newsletter depends on them. It was a very quiet week on our core topics. None of the big Alzheimer's companies (Lilly, Biogen, Eisai, Roche, AbbVie) came up by name in any podcast episode released in the window. None of the dedicated biotech shows had an in-window Alzheimer's episode that surfaced. We found five relevant episodes. One is squarely on topic: a Medscape-style teaching session on blood tests. One is a solid scientific interview about inflammation and dementia. One is about the weight-loss drug class moving into addiction. One is a regulatory chat about CNS drug development. The fifth is a cautionary tale, and we treat it that way.

(Quick vocabulary, used throughout. Amyloid and tau are the two toxic proteins that build up in an Alzheimer's brain. The two approved drugs, Eisai/Biogen's Leqembi and Eli Lilly's Kisunla, are antibodies that clear amyloid. p-tau-217 and p-tau-181 are pieces of the tau protein that can be measured in an ordinary blood sample. They rise as soon as amyloid starts to build up, so they work as a stand-in for a brain scan. A PET scan is the expensive brain scan that shows amyloid directly. CSF is the spinal fluid collected with a lumbar puncture, another way to check for amyloid. Sensitivity is how often a test catches people who have the disease. Specificity is how often it correctly clears people who don't.)

TL;DR

The blood-test playbook is now spelled out. A specialist on Keeping Current CME described two FDA-cleared blood tests with two different jobs. A p-tau-181 test lets primary-care doctors rule Alzheimer's out. A p-tau-217/amyloid-ratio test lets specialists rule it in. More tests are under FDA review. This is the route to finding patients earlier, which is what the drug makers need.
PET scans aren't going away yet. The same podcast said 10–20% of blood results fall into an unclear middle zone and need a scan or spinal tap. So do patients with severe kidney disease, and cases where the result doesn't fit the clinical picture. That takes some air out of last week's claim that PET could soon become "archaic." Reimbursement is also still "inconsistent."
An independent scientist backed the tau excitement, and warned on inflammation drugs. On BBC's Inside Health, Edinburgh's Prof. Tara Spires-Jones called tau-lowering drugs she saw at a recent conference "super exciting and promising." She also said "all of the anti-inflammatories tested have failed" in Alzheimer's.

What's New

The blood test now has an instruction manual, and it's built for primary care. On Keeping Current CME ("Alzheimer's Disease Precision Diagnostics: A 3-Part Live Masterclass – Part 3 – Blood-Based Biomarkers," Sep 22). EXPERT/CLINICAL: a continuing-medical-education session for doctors. The presenting specialist isn't named in the parts of the recording we reviewed. The key point is that the field has settled on two tests for two jobs:

The primary-care triage test (p-tau-181). It's cleared for patients aged 55 or older with signs of cognitive trouble. "If the test is negative, that essentially rules out amyloid pathology." A positive result does not mean Alzheimer's. It means the patient needs a confirmation test. The speaker said plainly: "we don't recommend this p-tau 181 test in specialty care."
The specialist confirmation test (p-tau-217 to amyloid-beta-42 ratio). It's cleared for patients aged 50 or older. "If the result is clearly positive, the patient is very likely to have amyloid pathology, and it is not necessary in at least some cases to perform an additional test." In a typical older patient with a clearly positive result, "the likelihood of amyloid pathology can approach 100%."
More are coming. "There are more tests that are currently under review and may be cleared soon."

(Our note, not from the podcast: these two descriptions match the FDA clearances for Fujirebio's Lumipulse p-tau-217/Aβ42 ratio test and Roche's Elecsys pTau181 primary-care test.)

Why it moves numbers: the drugs work best when started early, but most patients show up first at a family doctor, not a memory clinic. The speaker was direct: "We really need primary care providers to become more comfortable in diagnosing Alzheimer disease early." A cheap blood draw that a GP can order with routine lab work, and that screens out most people who don't have amyloid, is how the pool of patients who might get Leqembi or Kisunla grows. The speaker linked it to treatment directly: blood tests "have the potential to reduce the cost and time to diagnosis," which matters "because we think that initiating treatment as early as possible is helpful."

"Plasma p-tau 217 is the most accurate blood test for Alzheimer's disease pathology." (presenting specialist, Keeping Current CME)

The fine print that protects PET scans and slows the funnel. Still on Keeping Current CME, the limits were spelled out just as clearly. Blood results land in an intermediate zone "about 10 to 20% of the time," and those patients need a second test, usually a PET scan or spinal tap. The speaker deliberately switches to "a different modality," for example "from a blood test to an amyloid PET scan." Chronic kidney disease raises p-tau levels, so "if a patient has severe chronic kidney disease, I just order a different type of test, usually an amyloid PET scan." Younger patients, unusual symptoms, and any result that "does not align with your clinical suspicion" also send people to a scan. On money: "There can be inconsistent reimbursement… these tests are sometimes paid for, but they sometimes may not be paid for."

The speaker also used a simple example to show why who gets tested matters. A 75-year-old with classic Alzheimer's symptoms has about an 85% chance of amyloid before any test. A 60-year-old who only feels forgetful has maybe 20%. The same positive result means very different things in those two people. So the guidelines say to test only patients with objective signs of cognitive impairment, not worried healthy people. Why it matters: this is a gatekeeper as well as an accelerator. Blood testing widens the front of the funnel, but it won't be a mass screening tool for healthy adults any time soon, and PET keeps a real role as the tie-breaker.

A practical warning for the lab-test companies: order one test, not five. A small but telling detail from the same podcast: clinicians are "ordering a whole slew of tests, so not just p-tau 217, but p-tau 217, A-beta 42 to 40, NfL, GFAP, and then are quite confused about how to interpret the results." The advice: "p-tau 217 is the analyte that's best associated with amyloid pathology… these other tests are much less accurate." The speaker also said, "I would not recommend relying on lower performing tests that include A-beta 42 to 40 as the sole analyte," and asked for tests that reach roughly 90% sensitivity and specificity. Why it matters: a guideline-level voice is telling doctors to use p-tau-217. That favors platforms built around that marker over older amyloid-ratio-only tests and over add-on panels.

An independent scientist backs tau, and warns off inflammation drugs. On BBC's Inside Health ("Is inflammation behind all chronic disease?", Sep 22). EXPERT/ACADEMIC: Prof. Tara Spires-Jones, Director of the Centre for Discovery Brain Sciences at the University of Edinburgh and a division lead in the UK Dementia Research Institute. She said the field "has completely turned on its head" on inflammation. It is now seen as "a huge part" of how Alzheimer's develops. The brain's immune cells help early on by clearing amyloid and tau, then "become really toxic to neurons." The genes linked to higher Alzheimer's risk "are actually expressed in these immune cells." (That is the biology behind programs such as TREM2, which appeared on our watch list. She didn't name any specific program.) But she was blunt about drugs: "so far, all of the anti-inflammatories tested have failed." In contrast, "we're much, much closer with treatments that target both the amyloid pathology and the tau pathology… there are a couple that I saw recently at a conference, tau-lowering drugs that are looking super exciting and promising." She also cited "really convincing data from a study in Wales" linking the shingles vaccine to lower dementia risk, and "14 really well substantiated" modifiable risk factors (depression, inactivity, obesity, diabetes, social isolation), many of which run through inflammation. Why it matters: this is a neutral academic, not a company. She agrees with the view that tau is the next real frontier, and she says broad anti-inflammatory approaches remain unproven. Put simply: the next wave looks like tau, not inflammation.

The weight-loss drug class moves further into psychiatry, this time addiction. On Futureproof with Jonathan McCrea ("Could ozempic cure addiction?", Sep 20). EXPERT/ACADEMIC: Prof. Paul Kenny, Director of the Friedman Brain Institute at Mount Sinai. He explained that GLP-1 drugs (the Ozempic/Wegovy class) act on a brain "that's enough" circuit that also controls the appetite for drugs. "At least in animals unambiguously… you stimulate GLP-1 signaling in the brain, animals do not want to use addictive drugs." In people, the evidence so far is mostly anecdotal: "I don't drink as much," or the craving for cocaine "is really diminished." That's partly because until recently "no one actually asked the question." He called it "potentially a new class of treatment for substance use disorders." He was candid about the gaps. On 10-year safety, "very little" is known. On mood side effects, "the reports are mixed," and likely dose-dependent. Why it matters: last week we reported that semaglutide's EVOKE trials failed in Alzheimer's. The brain story for this drug class is moving away from dementia and toward addiction and psychiatry, which matters for Lilly and Novo's pipelines outside obesity.

A former FDA insider on how regulators keep up with CNS science (short version: slowly). On Citeline Podcasts ("Advancing CNS Drug Development Through AI, New Approach Methodologies and Regulatory Innovation," Sep 24). OPERATOR/REGULATORY CONSULTANT: Marcus DeLatt, VP of Regulatory Strategy at Ellucent, with prior FDA experience. "The reality is the agency is always behind… the agency relies very, very largely on following the lead of sponsors." He said AI is speeding up data analysis and report writing, but "AI tools may also hallucinate," so "you still need humans." Lab alternatives to animal testing, such as organ-on-a-chip devices and patient-derived "organoids," can "de-risk products before expensive investments," but "you can't use them alone to assess and manage product risks." Why it matters: this is background, not a catalyst. It fits a broader theme: in CNS, companies push new science and new endpoints (like blood biomarkers) to the FDA, and approval frameworks follow with a lag.

A cautionary tale from the unregulated edge of "Alzheimer's reversal." On Alzheimer's Breakthrough with Dr. Josh Helman, MD ("Mazos' (Giorgos Mazonakis) Passing: The Warning Signs We Should All Know," Sep 24). PRACTITIONER, WITH A CLEAR COMMERCIAL INTEREST: the host offers therapeutic plasma exchange (TPE, a dialysis-like blood-filtering procedure) in his own clinics and promotes his own supplement line during the episode. The discussion followed the death of the Greek singer Mazonakis, which the episode links to a blood-filtering procedure at a clinic in Greece. The host said the "double filtration" technology used there is "not FDA approved" in the US, and warned that "if you think that you're having a stroke or a heart attack… please go directly to the emergency room," not a wellness clinic. He also said TPE clinics are proliferating for "generalized anti-aging." He pointed to the AMBAR study as evidence for plasma exchange in Alzheimer's, and claimed symptoms "are getting reversed." Treat that last claim with real skepticism. AMBAR (a Grifols-sponsored trial of plasma exchange with albumin replacement) is a real study, but "reversal" is not an established finding. The source here sells the procedure. Why it matters: not an investable signal. It's a reminder that where approved drugs are modest and access is hard, patients look elsewhere, and that end of the market carries real safety and reputational risk.

The Debate

Does a clear, primary-care-friendly blood-test pathway finally open the patient funnel wide enough to make anti-amyloid (and next-generation tau) drugs a multi-billion-dollar franchise? Or do intermediate results, patchy reimbursement, comorbidities and the "test only the symptomatic" rule keep the funnel narrow, while tau bets are still unproven?

Bull: The biggest structural complaint about this category has always been diagnosis. Too many patients are never diagnosed, and confirming amyloid meant a costly PET scan or a lumbar puncture. That problem now has a practical answer. The two blood tests are cleared, more are under review, and the guidelines describe a simple path: GP rules out, specialist rules in. In a typical patient with a clear positive, amyloid "can approach 100%" with no scan needed. Blood tests are "cost effective, certainly more cost effective than amyloid PET," and patients "like it much more than a lumbar puncture" (Keeping Current CME). Add last week's Medicare GUIDE program, which pays to find the "up to 50%" never diagnosed, and the front of the funnel is widening. Meanwhile a neutral academic calls the next wave, tau drugs, "super exciting" (Inside Health).

Bear: Read the guidelines as a constraint, not a green light. Testing is only for patients with objective impairment, so there's no screening of healthy seniors. One in ten to one in five results is unclear and needs a scan or spinal tap anyway. Kidney disease, which is common in elderly patients, distorts the result. Reimbursement is "inconsistent." Doctors are still ordering the wrong panels and getting "quite confused." None of this is a sales figure: there were zero Kisunla or Leqembi launch numbers on any podcast again this week, and zero subcutaneous-dosing ramp data. The "exciting" tau drugs were described in one sentence, unnamed, from a conference. On inflammation, "all of the anti-inflammatories tested have failed."

Net: A plumbing week. The diagnostic pipe is getting better documented and more practical, which is a real long-term positive for patient volume. But the podcasts gave us nothing new on drug uptake or dollars. The things that will settle the debate are still ahead: actual launch numbers, blood-test reimbursement, and named tau Phase 3 data.

Stocks in Play

No company was discussed by name with fresh company news on any in-window podcast. Everything below is a read-through from what was said, labeled as such.

Roche (RHHBY / ROG SW), diagnostics. Read-through, the clearest positive. Bull: guideline-level validation that a p-tau-181 primary-care triage test has a defined, recommended job ("as a triaging test in primary care"), and that p-tau-217 is the gold-standard marker (Keeping Current CME). Bear: the same speaker said p-tau-181 should not be used in specialty care, and a positive result always needs confirmation, which limits the test's value per patient. Nothing on the trontinemab or bepranemab drug programs. Next catalyst: further FDA clearances among the tests "currently under review," and any trontinemab update.
Fujirebio (subsidiary of H.U. Group, 4544 JP). Read-through, positive. Bull: the specialist p-tau-217/Aβ42 ratio test described (matching Fujirebio's Lumipulse clearance, our note) is positioned as a stand-alone confirmation test in clear cases. Bear: intermediate results (10–20%) still send patients to PET or CSF. Next catalyst: volume and reimbursement data.
Quanterix (QTRX), C2N (private), Labcorp (LH), Quest (DGX). Read-through, mixed. Bull: strong expert push toward p-tau-217, and blood testing is "very accessible… often you're sending them to the lab anyway." Bear: explicit warnings against ordering "a whole slew" of markers (NfL, GFAP, Aβ42/40), and against relying on Aβ42/40 "as the sole analyte." Tests that aren't built around p-tau-217 are on the wrong side of this advice. "Inconsistent reimbursement." Next catalyst: payer coverage decisions for blood-based biomarkers.
LLY (Eli Lilly). Read-through only. Bull: every step that shortens the path to diagnosis feeds Kisunla eligibility. Separately, the GLP-1 class (Lilly's tirzepatide franchise included) is gaining scientific credibility in addiction, with "unambiguous" animal data (Futureproof). Bear: no Kisunla sales, patient-start or infusion-capacity commentary on any podcast, for another week. Addiction evidence in humans is still anecdotal. Next catalyst: actual Kisunla uptake numbers.
NVO (Novo Nordisk). Read-through. Bull: Ozempic/Wegovy were named directly as the drugs behind the anecdotal addiction reports (Futureproof). Bear: after the EVOKE Alzheimer's failure, the brain upside now rests on psychiatry indications that have no controlled human data yet, and long-term safety is "very little" known. Next catalyst: controlled trials of GLP-1 drugs in substance use disorders.
BIIB (Biogen) / Eisai (4523 JP). Read-through. Bull: independent support for tau-lowering drugs ("super exciting and promising") fits Biogen's tau story from two weeks ago (Inside Health). Earlier diagnosis helps Leqembi. Bear: the speaker named no program. There was no Leqembi or IQLIK ramp data. Next catalyst: tau Phase 3 readouts and Leqembi subcutaneous uptake.
Grifols (GRFS). Read-through, low quality. Bull: its AMBAR plasma-exchange study is being cited by practitioners and remains part of the discussion. Bear: the only source this week was a clinic owner promoting the procedure, alongside a death linked to an unregulated version of it (Alzheimer's Breakthrough). That is reputational noise, not a demand signal. Next catalyst: none from this week.
No podcast coverage this week (flagged, not invented): Kisunla/Leqembi launch and sales figures; Leqembi IQLIK subcutaneous ramp; Roche trontinemab/bepranemab; AbbVie neuroscience (emraclidine, Vyalev, Cerevel assets); named tau and non-amyloid programs (E2814, remternetug, BIIB080, ACI-35, TREM2, complement); PET imaging companies (Lantheus, GE HealthCare, Avid); a CMS coverage decision, registry or prior-authorization change; infusion centers/specialty pharmacy (OPCH); Cobenfy/KarXT, Parkinson's, Huntington's.

Read-Throughs

PET imaging (LNTH, GEHC): better than last week's scare suggested. Last week a CEO mused that PET could become "archaic." This week a guideline-level clinician described PET's lasting role: the 10–20% of intermediate blood results, patients with severe kidney disease, younger or unusual patients, and any result that clashes with clinical judgment (Keeping Current CME). The fair read: PET shifts from the first test to the tie-breaker. Per-patient use falls, but a much bigger diagnosed pool could keep total scan volume steady.
Blood diagnostics (Roche, Fujirebio, QTRX, C2N, LH, DGX): the winner is p-tau-217, stated as plainly as a clinician can: "the most accurate blood test." The investable question is no longer "will doctors order it?" but "which p-tau-217 platform, and will payers cover it?" Reimbursement was named as a live problem.
Infusion centers (OPCH): nothing direct. Indirectly, a faster diagnostic path means more patients reaching infusion chairs over time, while the move to self-injected dosing remains the longer-term headwind.
Partners (Eisai, Roche): Eisai benefits from earlier diagnosis like Biogen. Roche is the rare company on both sides, since a stronger diagnostics business partly hedges its drug-pipeline risk.
Broader CNS pipeline: two academic signals. Inflammation-targeted Alzheimer's drugs have "all… failed" so far, a caution for immune and complement programs. GLP-1 drugs are gaining serious scientific attention in addiction. The weight-loss leaders are quietly building psychiatry options.
The unregulated "reversal" market: plasma-exchange and "anti-aging" clinics are multiplying (Alzheimer's Breakthrough). Not investable, but it shows how much unmet demand approved treatments are leaving.

What Changed vs Last Week

Quiet week. No company-level news. Last week a heart-drug CEO put cholesterol into the Alzheimer's conversation, and a Mayo Clinic neurologist gave us the cleanest efficacy scorecard yet. This week no drug company spoke, and nothing we covered last week (obicetrapib, the revised Kisunla dosing, GUIDE) got a follow-up.

One real correction to last week's optimism. Last week New Amsterdam's CEO said PET scans could become "archaic relatively soon." This week's diagnostics specialist effectively walked that back, independently: PET remains the standard second test for the 10–20% intermediate zone, for severe kidney disease, and for results that don't match the clinical picture (Keeping Current CME). Blood is the volume play. PET is not dead.

Two threads continued. On tau: Biogen's CEO said two weeks ago that lowering tau moved cognition for "the very first time." This week an unaffiliated Edinburgh scientist called the tau-lowering drugs she saw at a recent conference "super exciting and promising" (Inside Health). That's independent support, though still without names or numbers. On the diagnosis gap: last week's GUIDE podcast said "up to 50%" of dementia is never diagnosed. This week showed the tool built for that gap: a primary-care blood test that rules Alzheimer's out.

A tension worth noting. Last week's cholesterol theory ran partly through brain inflammation. This week a leading neuroinflammation researcher said "all of the anti-inflammatories tested have failed" in Alzheimer's. That isn't a direct contradiction, since obicetrapib works through cholesterol transport rather than suppressing inflammation, but it is a reminder of how many promising "upstream" theories have failed in this disease.

The GLP-1 brain story moved on. Last week the semaglutide-for-Alzheimer's door stayed shut (EVOKE). This week the brain conversation for that drug class moved to addiction, with strong animal data and anecdotal human reports (Futureproof).

The honest gap list: no Kisunla or Leqembi launch or sales figures; no IQLIK subcutaneous ramp; no Roche drug data; no AbbVie neuroscience update; no named tau or non-amyloid program; no PET-company commentary; no CMS coverage or prior-authorization news; nothing on infusion centers, Cobenfy, Parkinson's or Huntington's.