Highlights
– In this post hoc analysis of TRAILBLAZER‑ALZ 2, lower posttreatment amyloid plaque levels after donanemab were closely correlated with slower clinical progression (iADRS and CDR‑SB) over 76 weeks.
– Lower posttreatment amyloid also correlated with declines in plasma p‑tau217, p‑tau181, and GFAP, but not with neurofilament light chain (NfL), suggesting specific downstream effects on tau pathology and astrocytic activation rather than broad mitigation of neuroaxonal injury within 76 weeks.
– Findings support amyloid plaque removal as the likely mechanism of action of donanemab and raise the possibility that achieved amyloid level could serve as a surrogate biomarker for clinical benefit—pending prospective validation.
Background and Clinical Context
Alzheimer disease (AD) is characterized by extracellular amyloid-β (Aβ) plaque deposition, intracellular tau aggregation, neuroinflammation, and progressive neurodegeneration. The amyloid cascade hypothesis posits Aβ accumulation as an early and upstream driver of downstream tau pathology and neuronal injury—a framework that underpins the development of anti‑amyloid immunotherapies.
The recent wave of anti‑amyloid monoclonal antibodies (mAbs) has shown that removal of fibrillar amyloid from the brain is possible and can be associated with slowing of clinical decline in early AD. However, questions persist regarding the extent to which amyloid removal mediates clinical benefit, what level of posttreatment amyloid reduction is necessary, and whether achieved amyloid burden can act as a surrogate predictor of downstream clinical outcomes.
Study Design
This article summarizes a post hoc, exploratory analysis of the randomized, placebo‑controlled phase 3 TRAILBLAZER‑ALZ 2 trial (ClinicalTrials.gov: NCT04437511), conducted at 277 centers across eight countries. From June 2020 to April 2023, participants aged 60–85 with early symptomatic AD and biomarker evidence of amyloid and tau pathology were randomized 1:1 to receive donanemab (intravenous 700 mg for three doses then 1,400 mg every 4 weeks) or placebo for up to 72 weeks, with outcomes through 76 weeks.
The present analysis included 1,582 participants (766 donanemab, 816 placebo) who had baseline and at least one posttreatment amyloid assessment. Each participant was assigned to one of 10 groups defined by the decile of their lowest posttreatment amyloid value. Primary clinical outcomes for this analysis were changes in the integrated Alzheimer Disease Rating Scale (iADRS) and Clinical Dementia Rating–Sum of Boxes (CDR‑SB). Plasma biomarkers included phosphorylated tau 217 (p‑tau217), p‑tau181, glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL). Correlations between median amyloid in each decile and 76‑week least‑squares mean changes in outcomes were calculated, with R2 values reported for the decile‑level associations.
Key Findings
Participant characteristics were balanced across groups: mean age 72.9 years (SD 6.2), 56.9% female. As expected, donanemab recipients achieved substantially lower posttreatment amyloid values than placebo-treated participants.
Clinical Outcomes
Across the trial population, lower posttreatment amyloid levels were strongly correlated with slower clinical progression at 76 weeks. The decile‑level correlations yielded R2 values of 0.73 (95% CI, 0.37–0.97) for iADRS change and 0.87 (95% CI, 0.70–0.97) for CDR‑SB change, indicating that decile median amyloid explained a large proportion of between‑decile variance in clinical outcomes over the study period.
Plasma Biomarkers
Lower achieved amyloid levels correlated strongly with reductions in plasma markers of tau pathology and astrocytic response. R2 values were 0.86 (95% CI, 0.73–0.97) for p‑tau217, 0.88 (95% CI, 0.77–0.97) for p‑tau181, and 0.87 (95% CI, 0.76–0.97) for GFAP at 76 weeks. These findings suggest that lowering fibrillar amyloid leads to measurable downstream biochemical changes consistent with reduced tau phosphorylation and glial activation.
In contrast, there was no meaningful correlation between posttreatment amyloid level and NfL (R2, 0.03; 95% CI, 0.00–0.54), a marker reflecting neuroaxonal injury. This dissociation suggests that, within the 76‑week timeframe of the study, amyloid removal produced early favorable effects on tau‑related and glial processes but not on markers of ongoing neurodegeneration detectable by plasma NfL.
Interpretation of Effect Sizes
The high decile‑level R2 values imply a strong association between the magnitude of amyloid clearance and short‑term clinical and biomarker outcomes at the group level. However, these correlations are derived from an exploratory, post hoc analysis that aggregates participants into deciles; they do not equate to individual‑level surrogacy or confirm causality beyond the randomized treatment effect.
Expert Commentary and Mechanistic Insights
These results strengthen the biological plausibility that amyloid plaque removal is a proximal mechanism leading to downstream biochemical changes (reduced p‑tau and GFAP) and clinical stabilization. The observed pattern—amyloid reduction with concordant declines in p‑tau and GFAP but not immediate reductions in NfL—fits a mechanistic model in which amyloid is upstream of tau phosphorylation and astrocytic activation, while neuroaxonal loss (reflected by NfL) represents a later, less reversible stage.
From a regulatory and translational perspective, the findings support the argument that achieved amyloid level might serve as a candidate surrogate endpoint for clinical benefit in amyloid‑targeting therapies. However, formal surrogacy requires demonstration that treatment-induced changes in the biomarker reliably predict the clinical effect across multiple trials and settings, ideally at both trial and individual levels. The approach used here—decile‑level correlations—provides encouraging, hypothesis‑generating evidence but falls short of formal validation.
Safety Considerations
Anti‑amyloid antibodies as a class have been associated with amyloid‑related imaging abnormalities (ARIA), including edema and microhemorrhages, which require monitoring and management algorithms in clinical practice. TRAILBLAZER‑ALZ 2 reported typical class‑related safety signals for donanemab in its primary publications; clinicians must weigh benefits against risks such as ARIA and the logistics of serial amyloid imaging and monitoring.
Limitations and Generalizability
Key limitations of this analysis include its post hoc exploratory nature and the aggregation of individuals into deciles, which may obscure within‑decile heterogeneity and does not establish individual‑level surrogacy. Correlation at the decile (group) level does not confirm that the biomarker change mediates the treatment effect in every individual. The follow‑up period (76 weeks) may be insufficient to observe effects on more slowly evolving neurodegenerative markers such as NfL or long‑term clinical outcomes.
Additionally, the trial’s inclusion criteria—early symptomatic AD with PET evidence of amyloid and tau—limit applicability to patients without biomarker confirmation or with more advanced disease. External factors such as APOE genotype, baseline tau burden, comorbidities, and concomitant medications could modify both biomarker responses and clinical trajectories and warrant further study.
Clinical and Research Implications
For clinicians, this analysis provides supportive evidence that achieving lower amyloid levels with donanemab is associated with favorable short‑term clinical and biomarker changes. It underscores the value of biomarker‑guided therapy and monitoring in early AD but does not yet mandate specific posttreatment amyloid targets in routine practice.
For researchers and regulators, the study highlights the potential for posttreatment amyloid burden to be developed as a surrogate marker, subject to prospective validation. Future research priorities include: (1) prospective, prespecified validation of achieved amyloid thresholds predictive of clinical benefit; (2) assessment of individual‑level surrogacy across multiple trials and agents; (3) longer follow‑up to determine effects on NfL and sustained clinical outcomes; and (4) exploration of combination approaches targeting both amyloid and tau or neuroinflammation.
Conclusion
This secondary analysis of TRAILBLAZER‑ALZ 2 demonstrates a strong decile‑level correlation between lower posttreatment amyloid plaque levels after donanemab and slower clinical progression, accompanied by reductions in plasma p‑tau217, p‑tau181, and GFAP but not NfL over 76 weeks. The results support amyloid plaque removal as a mechanistic mediator of benefit and nominate achieved amyloid burden as a promising surrogate biomarker candidate—while underscoring the need for prospective, formal validation and longer‑term data to establish its role in regulatory decision‑making and clinical practice.
Funding and Trial Registration
ClinicalTrials.gov Identifier: NCT04437511. Funding and disclosures are reported in the primary trial publications.
References
1. Lu M, Kim MJ, Collins EC, et al. Posttreatment Amyloid Levels and Clinical Outcomes Following Donanemab for Early Symptomatic Alzheimer Disease: A Secondary Analysis of the TRAILBLAZER‑ALZ 2 Randomized Clinical Trial. JAMA Neurol. 2025 Oct 13. doi:10.1001/jamaneurol.2025.3869. Epub ahead of print. PMID: 41082199.
2. Jack CR Jr, Bennett DA, Blennow K, et al. NIA‑AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement. 2018;14(4):535–562.
3. van Dyck CH, Swanson CJ, Aisen P, et al. Lecanemab in Early Alzheimer’s Disease. N Engl J Med. 2023;388(1):9–21.
4. Sevigny J, Chiao P, Bussière T, et al. The antibody aducanumab reduces Aβ plaques in Alzheimer’s disease. Nature. 2016;537(7618):50–56.
