Pull up the enrollment criteria for any Phase 2 or Phase 3 Alzheimer’s trial running today. Somewhere in the inclusion criteria, you will find a p-tau threshold — a blood-based cut point designed to confirm amyloid pathology before a patient enters a trial. That cut point was built on a single disease assumption. A study published March 11, 2026 in Nature Medicine just broke it.
The research, a multicenter cross-sectional study, found that elevated serum phosphorylated tau is not exclusive to Alzheimer’s disease. Patients with the most common forms of systemic amyloidosis — including AL amyloidosis and ATTR amyloidosis — also showed elevated serum p-tau levels. More precisely, the investigators demonstrated that serum p-tau could differentiate amyloidosis-related polyneuropathy from polyneuropathy caused by other etiologies. One biomarker. Multiple diseases. One enormous enrollment problem.
The Stratification Assumption That Just Cracked
The old operating assumption was clean: elevated p-tau in blood means Alzheimer’s pathology. That logic drove the design of large blood-based screening programs, including Lilly’s donanemab Phase 3 program (TRAILBLAZER-ALZ 2), which enrolled 1,736 participants using amyloid-related biomarker confirmation as a gatekeeping criterion. It underpinned the FDA’s accelerated approval framework for lecanemab — Eisai and Biogen’s anti-amyloid antibody — which relied on amyloid beta reduction as a surrogate endpoint, with p-tau trajectories used as supporting evidence in the clinical package. The entire architecture of modern Alzheimer’s trial enrollment rests on the premise that these biomarkers are disease-specific.
They are not. Or at least, not exclusively.
If systemic amyloidosis patients carry elevated serum p-tau, sponsors running blood-based prescreening programs are now looking at a contaminated funnel. A patient with ATTR amyloidosis — a disease with an estimated prevalence of 300,000 to 500,000 people in the United States, according to Pfizer’s epidemiological estimates for its tafamidis program — could screen into an Alzheimer’s trial on p-tau criteria alone. Conversely, an Alzheimer’s patient with undiagnosed systemic amyloidosis could be miscategorized, or more dangerously, enrolled without clinicians recognizing the comorbidity. Neither scenario produces clean data.
FDA’s 2023 guidance on clinical trials enrichment strategies explicitly warns against single-biomarker enrichment when that biomarker lacks disease specificity. The agency’s position in that document: enrichment tools must demonstrate that the selected population is more likely to show a detectable drug effect — a standard that presupposes you have correctly identified who is in the population in the first place. The Nature Medicine findings suggest that for serum p-tau, that presupposition now requires direct challenge.
Who Feels This First
Rare disease sponsors building ATTR or AL amyloidosis trial programs are operating in a landscape where Alnylam’s patisiran Phase 3 APOLLO trial — 225 patients with hereditary ATTR amyloidosis with polyneuropathy — showed a 34-point improvement on the mNIS+7 composite score, establishing the disease’s responsiveness to intervention. But that trial used nerve conduction and neuropathy staging as its primary stratification tools. The question now is whether the next generation of ATTR trials — particularly those chasing earlier-stage intervention — will incorporate blood-based biomarker screening. If they do, and if they reach for p-tau as a convenient off-the-shelf tool borrowed from the Alzheimer’s toolkit, the Nature Medicine data suggests they will need orthogonal confirmation. Serum p-tau alone will not hold.
CNS sponsors face a mirror-image problem. The biomarker validation packages underpinning most current Alzheimer’s IND submissions were built when p-tau was treated as a CNS-specific signal. Regulatory affairs teams that submitted those packages under FDA’s guidance on early Alzheimer’s disease drug development did not need to account for systemic amyloidosis cross-reactivity — because the cross-reactivity was not characterized. It is characterized now. Any sponsor planning a pre-IND meeting or a Type B meeting in the next 12 months should expect the agency to ask whether their p-tau screening algorithm accounts for this population overlap, particularly if the trial is enrolling patients with peripheral neuropathy comorbidities, cardiac involvement, or any organ-system profile consistent with undiagnosed systemic amyloidosis.
The eCOA and DHT implications compound the problem. Several sponsors have begun validating serum p-tau as a remote or decentralized screening tool — drawing blood at home collection sites or regional labs, running it through centralized assay platforms, and making enrollment decisions without an in-person clinical assessment. Under that model, the clinical context that might flag a systemic amyloidosis presentation — the cardiomyopathy, the carpal tunnel history, the unexplained weight loss — never enters the decision loop. A blood draw and a cut point. Clean, fast, and now potentially wrong.
The Validation Gap No One Budgeted For
What clinical ops leaders need to do now is straightforward, even if it is not cheap: biomarker qualification packages for serum p-tau used in enrollment decisions need a specificity addendum. That means adding orthogonal confirmation — imaging-based amyloid PET, CSF analysis, or cardiac biomarker panels such as NT-proBNP, which differentiates cardiac ATTR from other cardiomyopathies — to any blood-based p-tau screen used as a primary gating criterion. If you are running a decentralized Alzheimer’s trial using serum p-tau as a remote inclusion criterion, your current protocol may be enrolling a mixed population without knowing it. That is not a hypothesis. The Nature Medicine data, drawn from a multicenter cohort design, gives it empirical weight.
Sponsors should also watch what happens at the FDA’s Peripheral and Central Nervous System Drugs Advisory Committee and the Oncologic Drugs Advisory Committee — the two panels most likely to encounter submissions touching this biomarker overlap. The next PDUFA action date for any Alzheimer’s asset relying on blood-based p-tau stratification will be the first real test of whether the agency treats the Nature Medicine findings as a signal requiring protocol-level response or as background noise. Given the FDA’s current posture on biomarker-driven enrichment — increasingly rigorous since the 2021 accelerated approval reform discussions — the former is the more likely posture. Sponsors who wait for the agency to raise this in a complete response letter will have lost a year they did not need to lose.
Moe Alsumidaie is Chief Editor of The Clinical Trial Vanguard. Moe holds decades of experience in the clinical trials industry. Moe also serves as Head of Research at CliniBiz and Chief Data Scientist at Annex Clinical Corporation.
