Summary: Tau PET scans are a powerful tool for visualizing Alzheimer’s biology in living patients, but the most commonly used tracer—Flortaucipir (Tauvid)—can produce signals that do not always correspond to tau tangles. In a high-precision study, researchers used AI-enabled, voxel-by-voxel alignment to compare living PET scans directly with postmortem brain tissue, revealing that off-target factors such as iron deposits and MAO-B–related neuroinflammation can produce PET signals that mimic tau.
The study shows that Flortaucipir PET remains valuable for detecting tau in Alzheimer’s disease, where tau contributes substantially to the signal, but it also demonstrates that non-tau biology can explain much of the tracer’s signal in several non-Alzheimer’s disorders and even in some control cases. Recognizing these off-target sources improves diagnostic interpretation, helps avoid overestimating tau pathology, and points the way toward more specific future tracers.
Key Findings
- AI-driven precision: The team used artificial intelligence to align PET images with microscopic tissue sections on a voxel-by-voxel basis, producing thousands of matched points per case and far greater anatomical accuracy than conventional PET–autopsy comparisons.
- Primary off-target culprits: In several non-Alzheimer’s tauopathies and a non-tau control, most of the PET signal was better explained by ferric iron deposits and by elevated monoamine oxidase B (MAO-B), a marker associated with reactive astrocytes and neuroinflammation, rather than by phospho-tau.
- Alzheimer’s versus other disorders: For Alzheimer’s disease cases, phospho-tau remains a major contributor to Flortaucipir binding, though not the only one. In other tauopathies, the tracer frequently cannot distinguish tau from inflammation-related or iron-related signals.
- Clinical impact: These insights can reduce over-interpretation of borderline PET signals, improving prognosis, patient selection for trials, and clinical decision-making.
- Tracer development: Identifying specific off-target sources provides clear targets for designing next-generation PET tracers with greater specificity for tau pathology.
Source: UCSF
Background on tau and PET imaging: Tau proteins normally stabilize neuronal microtubules, but in Alzheimer’s disease tau can misfold and form intracellular tangles. These aggregated, hyperphosphorylated tau species spread across brain regions, disrupt neuronal function, and contribute to cell death. Tau PET imaging offers one of the few methods to observe these pathological processes in living people, informing diagnosis, prognosis conversations, and enrollment and monitoring in clinical trials.
Although Flortaucipir was developed to bind tau tangles, it also interacts with other brain constituents. That off-target binding can produce a diffuse, low-level signal that may be misinterpreted as tau pathology—especially in non-Alzheimer’s tauopathies where tau aggregates have different biochemical structures than Alzheimer’s-type tau.
The study, published in Acta Neuropathologica, combined ante-mortem tau PET imaging with extensive postmortem histology. Using an AI-enabled voxel-wise registration method, investigators matched PET intensity values to the exact microscopic tissue locations, enabling thousands of direct comparisons for each subject. They quantified three biological measures in the tissue sections: phospho-tau (using CP-13 immunostaining), ferric iron (Perls’ Prussian blue), and MAO-B (as an index of reactive astrocytes and neuroinflammation).
Results showed that in Alzheimer’s disease the Flortaucipir signal correlated moderately with histological tau, confirming tau’s prominent role in the tracer signal. In contrast, cases of non-Alzheimer’s tauopathies and a frontotemporal lobar degeneration case driven by TDP-43 pathology demonstrated weak or negligible correlations between PET signal and tau. In those cases, the PET signal correlated more strongly with ferric iron and MAO-B. The authors also note that iron and MAO-B do not explain all off-target signal, indicating additional unknown contributors remain to be identified.
“These findings clarify why tau PET sometimes lights up beyond what tau pathology alone predicts,” said co-senior author Lea T. Grinberg, MD, PhD. “For Alzheimer’s disease, tau is a key driver of the signal, but in other disorders iron and inflammation can dominate the image.”
Co-senior author Gil Rabinovici, MD, added that the study is intended to improve how clinicians and researchers interpret Flortaucipir PET, not to discourage its use. By distinguishing when the tracer reflects tau versus other biology, clinicians can make better-informed prognostic judgments and investigators can design tracers and analytic methods with higher specificity.
Additional authors: Yuheng Chen, Renaud La Joie, Felipe L. Pereira, Ganna Blazhenets, Lucile Zhu, Salvatore Spina, William W. Seeley, Helmut Heinsen, Daniela Ushizima, Duygu Tosun.
Funding: Supported by a research grant from Eli Lilly (Lilly Research Award Program), NIA R01AG070826, NIA K24 AG053435 (Grinberg), P30 AG062422 (GDR), U01 AG057195, P01AG09724, and the Rainwater Charitable Foundation (GDR).
Key Questions Answered
A: Not necessarily. In patients with Alzheimer’s disease, Flortaucipir signal is still largely driven by tau. But this study gives clinicians a clearer framework to determine when a modest or atypical PET signal might reflect inflammation or iron rather than tau spread.
A: Flortaucipir targets structural features of aggregated tau, but the chemical environments created by ferric iron deposits and the MAO-B enzyme can partially mimic those binding sites. As a result, the tracer can attach to those non-tau components and produce PET signal.
A: No. The study refines our understanding of what Flortaucipir images and supports more accurate interpretation. It also identifies targets for improved software filtering and next-generation tracers that will better distinguish tau from other brain biology.
Editorial Notes
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full.
- Additional context and clarification added by staff.
About this neuroimaging and Alzheimer’s disease research news
Author: Melinda Krigel ([email protected])
Source: UCSF
Contact: Melinda Krigel – UCSF
Image credit: Neuroscience News
Original Research (open access): Disentangling on and off-target binding in flortaucipir PET: a voxel-to-voxel P-tau, ferric iron, and MAO-B histology-to-flortaucipir PET comparison. Acta Neuropathologica. DOI: 10.1007/s00401-026-02983-x
Abstract (summary)
Flortaucipir PET has enabled in vivo visualization of tau pathology, but interpretation is complicated by off-target binding that can be prominent in non-Alzheimer’s tauopathies and other non-tau disorders. Using a voxel-to-voxel correlation method, this study quantitatively compared thousands of matched histology–Flortaucipir pairs from individual cases to evaluate relationships between PET signal and three tissue measures: phospho-tau (CP-13), ferric iron (Perls’ stain), and MAO-B. In Alzheimer’s disease, Flortaucipir signal correlated significantly, though moderately, with tau pathology, indicating tau is an important but not exclusive driver of signal. In non-Alzheimer’s tauopathies and a TDP-43 case, Flortaucipir signal correlated weakly with tau and more strongly with iron and MAO-B. Not all off-target signal was accounted for, implying additional contributors. A fuller understanding of these mechanisms is essential to improve diagnostic accuracy, analytic methods, and development of next-generation tau tracers.