TRAILBLAZER-ALZ-2-PTAU217
Plasma P-tau217 for detecting amyloid clearance after donanemab in Alzheimer's disease
Year of Publication: 2026
Authors: Emily C Collins, Ming Lu, Rose Beck, ..., Oskar Hansson
Journal: Alzheimer's & Dementia
Citation: Alzheimers Dement. 2026 Aug 7;22(8):e71740. doi: 10.1002/alz.71740
PDF: Clinical Question
Can a blood test (plasma p-tau217) replace amyloid PET to confirm amyloid clearance after donanemab treatment in Alzheimer's disease?
Bottom Line
Plasma p-tau217 — alone or combined with p-tau181, GFAP, NfL, age, and APOE ε4 status — cannot currently be used to detect treatment-related amyloid clearance (<24.1 Centiloids) at the individual level after donanemab; amyloid PET remains necessary to confirm plaque clearance, and novel plaque-specific plasma biomarkers are needed.
Major Points
- Plasma p-tau217 measured by mass spectrometry showed low diagnostic performance for detecting TRAC (<24.1 CL on PET) in donanemab-treated participants: AUROC 0.61 at 52 weeks, 0.64 at 24 weeks, 0.63 at 76 weeks.
- The Elecsys p-tau217 immunoassay performed similarly (AUROC 0.67 at 24 weeks, 0.64 at 52 and 76 weeks), showing the finding is not assay-specific.
- Change from baseline in p-tau217 correlated only weakly with change in PET amyloid plaque level at the individual level: R=0.34 at 52 weeks, R=0.26 at 24 weeks, R=0.22 at 76 weeks (all P<0.0001), despite a clear group-level relationship.
- A multivariate model combining p-tau217 with p-tau181, GFAP, NfL, age, and APOE ε4 status reached AUROC only 0.71 at 76 weeks — below the >0.80 threshold generally considered clinically valuable.
- TRAC was achieved by 29.7% of participants at 24 weeks, 66.1% at 52 weeks, and 76.4% at 76 weeks.
- The authors suggest amyloid-targeting therapy may disrupt the association between amyloid deposition and hyperphosphorylated tau accumulation, so post-treatment p-tau217 may depend more on non-amyloid factors such as cerebral neurofibrillary tangles or peripheral tau forms.
- There is an urgent need for novel plaque-specific plasma biomarkers to monitor amyloid clearance after anti-amyloid therapy.
Design
Study Type: Biomarker sub-analysis of a phase 3, multicenter, randomized, double-blind, placebo-controlled trial (TRAILBLAZER-ALZ 2); analysis restricted to donanemab-treated participants
Randomization: 1
Blinding: Double-blind (parent trial); blinded switch to placebo after amyloid clearance criteria met
Allocation: 1:1 donanemab vs placebo in the parent trial; this analysis used the donanemab arm only
Follow-up Duration: 76 weeks (assessments at baseline and weeks 24, 52, 76)
Centers: Multicenter (number of centers not reported)
Countries: Australia, Canada, Czech Republic, Japan, Netherlands, Poland, United Kingdom, United States
Sample Size: 830
Analyzed: 830
Analysis: Exploratory biomarker analyses, unadjusted for multiplicity; Spearman rank correlation between plasma biomarkers and PET amyloid; ROC analysis (univariate p-tau217 and multivariate model with p-tau217, p-tau181, GFAP, NfL, age, APOE ε4) for detecting TRAC (<24.1 CL); biomarkers log10-transformed; SAS 9.4; two-sided P<0.05
Registration: ClinicalTrials.gov NCT04437511
Inclusion Criteria
- Age 60 to 85 years
- Diagnosis of early symptomatic AD (prodromal AD or AD with mild dementia)
- Elevated amyloid and tau levels on PET imaging
- MMSE score of 20 to 28
Baseline Characteristics
Donanemab (N=830):
- Female: 474 (57.1%)
- Male: 356 (42.9%)
- Age, mean (SD), years: 72.9 (6.2)
- White: 752 (90.7%)
- Asian: 56 (6.8%)
- Black or African American: 19 (2.3%)
- Hispanic or Latino (US/Puerto Rico only): 34 (5.7%)
- United States: 599 (72.2%)
- APOE ε4 carrier: 578 (69.8%)
- APOE ε3/ε4: 419 (50.6%)
- APOE ε4/ε4: 138 (16.7%)
- AChE inhibitor/memantine use at baseline: 507 (61.1%)
- iADRS, mean (SD): 104.4 (14.2)
- CDR-SB, mean (SD): 3.9 (2.1)
- ADAS-Cog13, mean (SD): 28.6 (8.8)
- ADCS-ADL, mean (SD): 66.5 (8.6)
- ADCS-iADL, mean (SD): 47.9 (7.9)
- MMSE, mean (SD): 22.5 (3.8)
- MMSE category: MCI (≥27): 144 (17.3%)
- MMSE category: mild AD (20-26): 685 (82.5%)
- Whole brain volume, mean (SD), cm3: 971.7 (101.9)
- Hippocampus volume, mean (SD), cm3: 6.2 (1.1)
- Ventricle volume, mean (SD), cm3: 49.6 (22.3)
- Amyloid plaque level at screening, mean (SD), Centiloids: 103.0 (34.2)
- Tau PET (MUBADA) at screening, mean (SD), SUVR: 1.3 (0.3)
- High tau at screening: 258 (31.1)
- Low-intermediate tau at screening: 571 (68.9)
Arms
| Field | Donanemab |
|---|---|
| N | 830 |
| Intervention | Donanemab IV every 4 weeks for up to 72 weeks (700 mg for first 3 doses, then 1400 mg); blinded switch to placebo if amyloid <11 CL on one PET scan or 11 to <25 CL on two consecutive scans at weeks 24/52. Plasma p-tau217 (C2N mass spectrometry assay and Elecsys immunoassay), p-tau181, and GFAP measured at baseline and weeks 24, 52, 76 alongside amyloid PET (florbetapir/florbetaben F18) |
| Duration | 76 weeks |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Diagnostic performance (AUROC) of plasma p-tau217 (mass spectrometry) in detecting treatment-related amyloid clearance (TRAC, <24.1 Centiloids on amyloid PET) after donanemab treatment | Primary | AUROC 0.61 at 52 weeks (N=670); AUROC 0.64 at 24 weeks; AUROC 0.63 at 76 weeks | AUROC 0.61 at 52 weeks — suboptimal diagnostic performance (clinically valuable threshold generally >0.80) | ||
| Diagnostic performance of p-tau217 by Elecsys immunoassay in detecting TRAC | Secondary | AUROC 0.67 at 24 weeks; 0.64 at 52 weeks; 0.64 at 76 weeks | |||
| Correlation between change from baseline in plasma p-tau217 (mass spectrometry) and change in PET amyloid plaque level | Secondary | R=0.34 at 52 weeks (P<0.0001); R=0.26 at 24 weeks (P<0.0001); R=0.22 at 76 weeks (P<0.0001) — statistically significant but weak | |||
| Diagnostic performance of multivariate model (p-tau217, p-tau181, GFAP, NfL, age, APOE ε4 status) in detecting TRAC | Secondary | AUROC 0.71 at 76 weeks — no meaningful improvement over p-tau217 alone; still below clinically valuable threshold | |||
| Proportion of donanemab-treated participants achieving TRAC (<24.1 CL) | Secondary | 29.7% at 24 weeks; 66.1% at 52 weeks; 76.4% at 76 weeks | |||
Criticisms
- Only two plasma p-tau217 assays were evaluated; other assays could theoretically yield different results (though both assays used have excellent diagnostic performance and gave concordant findings)
- Analysis limited to a clinical trial cohort of early symptomatic AD, precluding generalization to more diverse real-world populations
- All biomarker analyses were exploratory and unadjusted for multiplicity
- Industry-conducted analysis: all authors are Eli Lilly employees and shareholders
Funding
Eli Lilly and Company (responsible for design, conduct, data collection, analysis, interpretation, manuscript preparation, and decision to publish); medical writing and editorial assistance by Syneos Health, funded by Eli Lilly
Based on: TRAILBLAZER-ALZ-2-PTAU217 (Alzheimer's & Dementia, 2026)
Authors: Emily C Collins, Ming Lu, Rose Beck, ..., Oskar Hansson
Citation: Alzheimers Dement. 2026 Aug 7;22(8):e71740. doi: 10.1002/alz.71740
Content summarized and formatted by NeuroTrials.ai.