In-vitro interaction studies between the amyloid PET tracer florbetaben and the amyloid-beta targeting antibodies lecanemab and donanemab on AD brain samples reveal no interferences

Abstract

Background

Amyloid PET imaging enables the in vivo visualization and quantification of amyloid-beta deposits in the brain. In research and clinical settings, it is further used to monitor amyloid-beta burden as well as the biological response to disease-modifying therapies. Amyloid-beta targeting monoclonal antibodies (mAbs) such as lecanemab and donanemab were designed to reduce brain amyloid-beta burden. If the PET tracer and the mAbs would bind to the same site, the PET signal may be impacted in patients undergoing therapy. Binding interaction studies were conducted to verify the reliability of PET readouts in this setting. The aim of this study was to investigate whether lecanemab or donanemab interfere with the binding of the amyloid PET tracer florbetaben to aggregated amyloid-beta deposits in vitro.

Methods

Human Alzheimer’s disease (AD) brain tissue from various sources was used to assess potential interactions between florbetaben and lecanemab or donanemab. Three complementary approaches were used to confirm target binding and to study potential interactions: (1) competitive immunohistochemistry (IHC), (2) competitive autoradiography (ARG), and (3) ligand binding assays (LBA).

Results

Across all three sets of experiments, no evidence of competition or inhibition of florbetaben binding to amyloid-beta deposits by lecanemab or donanemab was observed. Autoradiography demonstrated robust tracer binding to amyloid-beta plaques that was unaffected by incubation with excess antibody. Similarly, IHC and LBA experiments confirmed that florbetaben and the tested mAbs target distinct binding sites on amyloid-beta aggregates.

Conclusions

These results demonstrate that neither lecanemab nor donanemab interfere with florbetaben binding to amyloid-beta plaques in vitro, further validating its use in this setting. While the current data are limited to in-vitro experiments, they further support the use of florbetaben PET for monitoring amyloid-beta changes during treatment with amyloid-beta targeting therapies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13195-026-02001-y.

Introduction

Alzheimer’s disease (AD) is characterized by the progressive accumulation of amyloid-beta aggregates in the brain. This pathological hallmark occurs first and plays a central role in current concepts of disease pathogenesis, it is followed by deposition of tau neurofibrillary tangles and cognitive decline [1]. Positron emission tomography (PET) with radiolabeled tracers targeting amyloid-beta or tau deposition allows direct and specific visualization and quantification of these protein deposits. Amyloid-beta peptides are generated through sequential cleavage of the amyloid precursor protein (APP) and are washed out of the brain. However, they can self-assemble into aggregated species that differ in size, conformation, and solubility leading to neurotoxic species and accumulation in the brain [2]. Many therapeutic strategies were investigated aiming to interfere with amyloid-beta aggregation or to confer plaque clearance. Importantly, therapeutic monoclonal antibodies (mAbs) have been developed to target different epitopes and specific species along the amyloid-beta aggregation pathway, allowing amyloid-beta targeting immunotherapy [3, 4].

Recent years have witnessed the regulatory approval of the first amyloid-beta targeting mAbs, including lecanemab and donanemab, which have demonstrated efficacy in reducing amyloid-beta burden in the brain and slowing cognitive decline in early symptomatic AD patients [5, 6].

Target engagement in these trials, i.e. amyloid-beta lowering during treatment, was demonstrated using amyloid PET imaging [7–10]. The robustness and generalizability of PET based measurements of plaque density across different tracers has been shown [11]. Indeed, all three amyloid PET tracers currently approved for clinical use in many countries, florbetaben, florbetapir, and flutemetamol, have been employed in most of the trials.

Amyloid PET imaging enables the in vivo visualization and quantification of amyloid-beta deposits in the brain and provides means to directly monitor changes in amyloid-beta levels over time. While these findings underscore the utility of amyloid PET as an essential biomarker, they also raise questions about possible interactions between amyloid-beta targeting therapeutics, such as lecanemab or donanemab, and amyloid PET tracers, such as florbetaben. Both therapeutic mAbs and amyloid tracers bind to amyloid-beta deposits in the brain and interference could occur if they target overlapping sites, particularly when administered in close temporal proximity. Because therapeutic mAbs reach the brain at much higher concentrations than amyloid tracers, they could theoretically occupy or sterically block epitopes that serve as binding sites for florbetaben and have an effect on the measured amyloid PET signal.

Lecanemab and donanemab show similarities and differences in their binding profiles and mechanisms of action. Lecanemab is a recombinant humanized IgG1 mAb derived from the murine antibody mAb158, specifically developed to recognize the conformational epitopes of N-terminal soluble protofibrillar amyloid-beta and binding to amyloid-beta plaques was confirmed on post-mortem AD brain tissue [3, 12–14]. Donanemab is a humanized IgG1 mAb that targets a modified form of the amyloid-beta peptide containing the N-terminal pyroglutamate (N3pG) epitope with high affinity [15]. Florbetaben was developed to visualize and quantify amyloid-beta plaques in vivo. Florbetaben is an established tool in the diagnosis of AD, for studying the natural history of AD and for evaluating patients undergoing amyloid-beta targeting therapy [16–18]. Autoradiographic analyses and in vivo PET studies have shown that florbetaben reliably binds to amyloid-beta plaques across cortical regions in the AD brain, with binding patterns closely correlating with histopathological markers such as immunohistochemistry and Bielschowsky silver staining [19]. Furthermore, distinct high- and low-affinity binding sites were identified in AD cortical homogenates and reported for amyloid PET tracers including PiB, florbetaben and florbetapir [20]. Strong selectivity of florbetaben binding to amyloid-beta deposition over other pathological proteins such as tau or α-synuclein was demonstrated ex vivo [21]. The ability of florbetaben PET to detect treatment-induced changes in amyloid-beta burden has been studied in preclinical models. Reductions in florbetaben PET signal corresponded with decreased fibrillar amyloid-beta load and plaque density [22]. These findings are consistent with the use of amyloid PET in trials of amyloid-beta targeting therapies such as lecanemab [9] and donanemab [10], where PET imaging was used to monitor reduction of amyloid-beta levels in response to treatment. Moreover, this aligns with recent appropriate use updates for amyloid PET [11] and treatment-specific recommendations, such as those for donanemab [23], which both highlight the role of PET in assessing amyloid-beta levels during and after therapy.

Despite the established role of amyloid PET as a biomarker of treatment-induced amyloid-beta changes in clinical trials, potential interactions between amyloid-beta targeted treatments and amyloid PET tracers when binding to amyloid-beta plaques have not been systematically investigated in humans. Hypothetically, the binding of amyloid tracers to amyloid-beta plaques could be influenced during treatment with amyloid-beta targeting mAbs. This could have implications for the interpretation of PET signal during treatment and conclusion about amyloid-beta clearance.

The aim of this study was to investigate whether binding interactions exist in vitro between florbetaben and the amyloid-beta targeting mAbs lecanemab and donanemab and ultimately if lecanemab or donanemab reduce florbetaben signal on histopathology tissue. Because human brain tissue contains endogenous human IgG, immunostaining with human or humanized monoclonal antibodies, such as lecanemab or donanemab, can result in substantial background signal [24], complicating the interpretation of true antigen-specific binding. Although immunohistochemical detection of lecanemab in human tissue has been reported, most recently in brains from individuals with Down syndrome [25], these studies also required optimized blocking and detection strategies to overcome endogenous IgG interference.

To circumvent this, three complementary in vitro methods were applied to investigate potential interactions: competitive immunochemistry (IHC), competitive autoradiography (ARG) and ligand binding assays (LBA). Firstly, IHC assays were used to confirm the presence of the target by using established techniques and standard amyloid-beta targeting antibodies. Confirmation of lecanemab and donanemab binding to these targets was obtained by competitive IHC against standard antibodies. Secondly, ARG and LBA were used to assess florbetaben binding after pre-incubation of AD brain-sections or homogenates with lecanemab or donanemab or when co-incubated with these antibodies.

Methods

Antibodies

Leqembi (lecanemab-irmb) with 100 mg/mL and Kisunla (donanemab-azbt) with 17.5 mg/ml were both purchased in the UK as clinical samples from the respective marketing authorization holders, Eisai Europe Limited or Eli Lilly Nederland B.V., respectively, and solely used for performing preclinical experiments.

The antibody mAb158 (Antibody Systems Inc.), an amyloid-beta protofibril-selective murine antibody (mouse lecanemab), the antibody mouseDonanemab (mDona, mE8-IgG2a, MedChemExpress LLC), murine monoclonal antibody which binds to N-terminal truncated, pyroglutamate-modified at amino acid position 3, and the monoclonal 6E10 antibody (Absolute Antibody Ltd), which is selective for the N-terminal part of amyloid-beta, were used for immunohistochemistry experiments.

Antibody concentrations in the in vitro experiments were chosen to span and exceed estimated CNS levels achieved during therapeutic dosing of lecanemab and donanemab. Based on typical brain penetration of ~ 0.1% of systemic dose [26–28], cumulative CNS concentrations of lecanemab and donanemab over 18 months are estimated to be ~ 434 nM and ~ 370 nM. In vitro, equimolar concentrations of up to ~ 3.4 µM (corresponding to 500 µg/mL) were used in the histochemistry experiments and up to 6.8 µM for competition assays, ensuring that any measurable effects could be detected well above physiologically relevant levels.

Florbetaben

Radiolabeled florbetaben was obtained according to established procedures used for commercial manufacturing [29]. Mean molar activity across florbetaben syntheses was 111.9 GBq/µmol ± 74.4 GBq/µmol (n = 8).

Human brain tissue samples

AD brain tissue from the florbetaben phase 3 histopathology study [19], the Netherland Brain Bank (NBB), Netherlands Institute for Neuroscience, Amsterdam (www.brainbank.nl) and the Neurosciences Victoria (NSV) brain bank were used for the IHC and autoradiographic evaluation (see supplementary Table 1). For the ligand-binding assay a brain homogenate was used that was pooled from four AD brain frontal cortex samples previously obtained from NSV (see supplementary Table 1).

Written informed consent for brain autopsy and for the use of tissue and clinical information for research purposes has been obtained by the tissue source. All experiments were performed in the Lantheus research labs in Berlin, Germany.

Immunohistochemistry

Air-dried frozen brain tissue sections were fixated with acetone for 20 min at −20 °C. Afterwards, slides were washed for 10 min with 50 mM Tris buffer, pH 7.4 and incubated for 20 min with 2.5% normal horse serum (NHS). For the pre-incubation experiments, slides were incubated with lecanemab or donanemab at an equimolar concentration of ~ 3.4 µM (corresponding to 500 µg/mL) in NHS for 1 h. Afterwards, sections were incubated overnight with 6E10 (mouse anti-beta-amyloid, Absolute Antibody Ltd) monoclonal antibodies diluted (1:1750 in NHS, 0.57 µg/mL) in a humidified chamber at 4 °C. Alternatively, sections were incubated overnight with lecanemab or donanemab at an equimolar concentration of ~ 3.4 µM (corresponding to 500 µg/mL) and 6E10 monoclonal antibodies diluted (1:1750 in NHS, 0.57 µg/mL) in a humidified chamber at 4 °C. Following the incubation, slides were washed twice for 5 min in 50 mM Tris buffer, pH 7.4 following an incubation with ImmPRESS-AP Horse Anti-Mouse IgG Polymer Reagent (Vector Labs) for 60 min in a humidified chamber at room temperature. Sections were washed twice for 5 min each with 50 mM Tris buffer, pH 7.4 and incubated with ImmPACT Vector Red Solution (Vector Labs) until desired stain intensity developed. Following the staining, the sections were washed for 5 min in fluent water and mounted with Roti Mount Aqua (Carl Roth) before scanning the different sections with the Pannoramic Midi scanner (Sysmex Deutschland GmbH). The same experiment was repeated using mAb158 (mouse lecanemab 1:1200; 1.15 µg/mL, Antibody Systems Inc.) or mouse Donanemab (mE8-IgG2a, MedChemExpress LLC) instead of 6E10.

Displacement analysis

Displacement was assessed by comparing the staining of 6E10, mAb158, and mDona on tissue sections with and without co- or pre-incubation of excess lecanemab or donanemab. Displacement reflects the competitive effect of excess lecanemab or donanemab on the binding of the primary antibodies used to characterize amyloid-beta plaques (6E10, mAb158, and mDona). Reduced staining intensity of the primary antibodies served as the semi-quantitative visual readout and was formalized on a three-point scale: no displacement, partial displacement, or complete displacement. In addition, staining intensity was quantified using ImageJ (ImageJ,.JS https://ij.imjoy.io/, [30]) using a standard pipeline [31] by applying a threshold-based approach to identify pixels with positive labeling. For each set of experiments (see supplementary Fig. 2), tissue characterization (without co- or pre-incubation) was scaled to 100% based on a optimized threshold, that was applied to the corresponding displacement experiments.

Autoradiography

Frozen 10–18 μm thick human brain slices were examined by autoradiography (ARG). All slides were equilibrated for 5 min in 25 mM Hepes, 0.1% BSA, pH 7.4 prior to use in the experiment. Each brain section was covered with 200 μL of [¹⁸F]florbetaben at 0.15 kBq/µL, to determine total binding (TOTB). For the non-specific binding (NSB), an excess of non-radioactive ¹⁹F-florbetaben was added (10 μM) to the tracer solution. For co-incubation a dilution series of non-radioactive ¹⁹F-florbetaben, lecanemab or donanemab (for each series: 3.4 nM, 34.2 nM, 342 nM, 3.4 µM and 6.8 µM) was added to the brain sections for studying competition. Similarly, for pre-incubation a dilution series of lecanemab and donanemab (for each series a 1 h pre-incubation with lecanemab in Hepes or with donanemab in Hepes was performed at following concentrations: 3.4 nM, 34.2 nM, 342 nM, 3.4 µM and 6.8 µM). The brain sections were then incubated with tracer solution at room temperature (RT) for 1 h in a humidity chamber, drained thereafter and placed in a slide holder. Slides were washed sequentially in ice-cold buffers: 1 min (H₂O), 2 min (70% EtOH in H₂O), 1 min (30% EtOH in H₂O), and 1 min (H₂O). Slides were allowed to air-dry before exposing the imaging plates overnight. Imaging plates were scanned using the Amersham Typhoon (GE) and images analyzed using AIDA Image Analyzer 5.1 (Raytest). Quantification of autoradiography signal was performed by applying three regions of interest (ROIs) per sample, with ROIs defined equivalently across corresponding tissue sections incubated with different competition conditions. Presence of amyloid-beta was confirmed by IHC as described above, which was performed on consecutive sections of the same brain.

Ligand-binding assays (LBA)

Human AD brain homogenates (20 μg total protein/well), prepared in-house for this study, were incubated with [¹⁸F]florbetaben and non-radioactive ¹⁹F-florbetaben, lecanemab or donanemab ranging from 0.061 nM to 1000 nM for 60 min at 37 °C in 96-well plates. For the pre-incubation series, AD brain homogenates were incubated for one hour in Hepes containing either lecanemab or donanemab at concentrations ranging from 0.061 nm to 1000 nm. In general, assays were performed in 25 mM Hepes containing 0.1% BSA and 2% DMSO. Non-specific binding (NSB) was determined by addition of 1 µM ¹⁹F-florbetaben. After incubation, samples were filtered under vacuum on equilibrated GF/C UniFilter plates (PerkinElmer) using the FilterMate 196 (PerkinElmer). Afterwards, filters were washed twice with 200 μL chilled buffer. Top and bottom sides of filter plates were sealed, and an imaging plate was placed on top of the filter plates and exposed overnight. Imaging plates were scanned and assessed as described above. Specific binding was calculated by subtracting the NSB signal from measured sample signals. The unblocked radiolabeled tracer signal was defined as total binding (TOTB). IC₅₀ values were calculated using Prism V10 (GraphPad).

Results

Tissue characterization and antibody displacement studies

Brain samples from three deceased AD patients were examined. Presence of significant levels of amyloid-beta plaques was confirmed on consecutive tissue slices. The established anti-amyloid-beta antibody 6E10, the mouse-version of lecanemab, mAb158 and the mouse version of donanemab, mDona (as exemplary shown in Fig. 1A-D and supplementary Fig. 1A-D) were used for the detection of human amyloid-beta depositions, because direct detection of humanized antibodies such as lecanemab and donanemab on human tissue samples is challenging as endogenous human IgG causes substantial background signal, requiring optimized blocking and detection strategies [24].

Fig. 1.

Tissue characterization (Case #2 in supplementary Table 1, hippocampus) with immunohistochemistry (6E10, mAb158 and mDona) targeting amyloid-beta deposits: A-D Amyloid-beta deposition on brain slices as shown by pink staining by either 6E10 (A and B), mAb158 (C) or mDona (D). E–H Corresponding displacement experiments using co-incubation with excess amyloid-beta targeting therapies (lecanemab or donanemab): E 6E10 + excess lecanemab, F 6E10 + excess donanemab, G mAb158 + excess lecanemab, H mDona + excess donanemab. Displacement of the primary antibodies used for tissue characterization (6E10, mAb158 and mDona) was semi-quantitatively assessed with a three-point scale: no displacement, partial displacement, or complete displacement

Subsequently, competitive IHC assays were performed to test binding of either lecanemab or donanemab to the same binding epitopes of 6E10, mAb158 or mDona, respectively. Co-incubation with excess lecanemab showed both a qualitative reduction in 6E10 staining (see Fig. 1E) and a strong statistically significant reduction in 6E10 staining intensity (100% vs 39 ± 7.6%, p = 0.03, see supplementary Fig. 2 A) confirming the binding epitopes of both antibodies are overlapping. Pre-incubation experiments with excess lecanemab confirmed this result (see supplementary Fig. 1E) and led to a statistically significant reduction in 6E10 staining intensity (100% vs 19.7 ± 2.6%, p = 0.001, see supplementary Fig. 2 A). Co-incubation and pre-incubation with excess donanemab showed no statistically significant change in 6E10 staining intensity (Co-incubation: 100% vs 146 ± 15.0%, p = 0.16; Pre-incubation: 100% vs 115 ± 37.0%, p = 0.92, see supplementary Fig. 2B) confirming the binding epitopes of both donanemab and 6E10 are not overlapping (see Fig. 1F, supplementary Fig. 1 F).

The binding epitopes of the corresponding murine and human antibody versions are identical. Control displacement experiments were performed with mAb158 and lecanemab and with mDona and donanemab. A complete signal blocking of mAb158 was observed with excess lecanemab (Co-incubation: 100% vs 1 ± 0.58%, p < 0.0001; Pre-incubation: 100% vs 0.67 ± 0.34%, p < 0.0001; see Fig. 1G, supplementary Fig. 1G and supplementary Fig. 2 C) and correspondingly excess donanemab blocked mDona completely (Co-incubation: 100% vs 1 ± 0.58%, p < 0.0001; Pre-incubation: 100% vs 3.67 ± 2.67%, p = 0.0003; see Fig. 1H, supplementary Fig. 1H and supplementary Fig. 2D), confirming identical binding epitopes of the corresponding murine and human antibody versions.

No statistical differences between pre- and co-incubation experiments were observed in all experiments (supplementary Fig. 2).

Autoradiography

Florbetaben autoradiography was performed on frozen human brain tissue of three deceased AD patients to detect amyloid-beta plaques. A strong autoradiography signal was observed when the slices were incubated only with ¹⁸F-labelled florbetaben (see Fig. 2 top row). The autoradiography signal on AD tissue was found to be displaceable in the presence of excess non-radioactive ¹⁹F-florbetaben. Specific binding of florbetaben was demonstrated by addition of various concentrations of non-radioactive ¹⁹F-florbetaben, leading to a concentration-dependent signal reduction (see Fig. 2, left column). Co-incubation with lecanemab showed no reduction in the florbetaben autoradiography signal at all tested concentrations up to 6.8 µM (1000 µg/mL) of lecanemab (see Fig. 2, middle column). Co-incubation of donanemab showed no reduction in the florbetaben autoradiography signal at all tested concentrations with up to 6.8 µM (1000 µg/mL) of donanemab (see Fig. 2, right column). The same experiments were repeated using pre-incubation of tissue with various concentrations of lecanemab or donanemab (see supplementary Fig. 3) and showed no impact on the florbetaben autoradiography signal. Quantitative analysis of the autoradiography data was performed using three ROIs per tissue sample, and the results are shown in Supplementary Fig. 4. A clear concentration-dependent decrease in signal intensity was observed for incubation with non-radioactive ¹⁹F-florbetaben, whereas no concentration-dependent effect was seen for either lecanemab or donanemab. The same binding patterns were observed for both co-incubation and pre-incubation experiments.

Fig. 2.

Top row: total binding of [¹⁸F]florbetaben, without co-incubation with competitor. Left column: Competitive florbetaben autoradiography on AD tissue (Case #2 in supplementary Table 1, hippocampus) with addition of different concentrations (3.4 nM, 34.2 nM, 342 nM, 3.4 µM, 6.8 µM) of non-radioactive ¹⁹F-florbetaben; Middle column: co-incubation with lecanemab at different concentrations; or right column: co-incubation with donanemab at different concentrations. Co-incubation with excess non-radioactive.¹⁹F-florbetaben showed concentration-dependent reduction of the autoradiography signal (left column). No signal reduction of florbetaben autoradiography was observed for different concentrations of lecanemab (middle column) or donanemab (right column)

Ligand-binding assays

Ligand-binding assays were performed on a pooled AD frontal cortex homogenate (n = 4 brains). No inhibition of florbetaben binding was observed during co-incubation or pre-incubation at all tested concentrations with up to 3.4 μM (500 μg/mL) of lecanemab or donanemab. The results are shown in Supplementary Fig. 5. In contrast, non-radioactive ¹⁹F-florbetaben displaced radioligand binding with an IC₅₀ of 17.8 nM.

Discussion

This study investigated potential binding interactions between florbetaben and the amyloid-beta targeting mAbs lecanemab and donanemab using human AD brain tissue. Across all in-vitro interaction assays, neither lecanemab nor donanemab competed with or inhibited florbetaben binding to amyloid-beta plaques.

Tissue characterization by immunohistochemistry was conducted on selected AD brain samples to confirm presence of amyloid-beta. All three antibodies used (6E10, mAb158, mDona) demonstrated target engagement, confirming binding to amyloid-beta deposits. These findings validated the suitability of the selected AD samples and standard amyloid-beta targeting antibody 6E10, as well as the murine versions of lecanemab (mAb158) and donanemab (mDona), for subsequent competition assays. Although all three antibodies bind amyloid-beta, their binding epitopes differ. 6E10 was traditionally thought to bind to residues 1–16, however, recent high-resolution epitope mapping refined its binding site to residues 4–10 and demonstrated that 6E10 binds in conformation- and sequence-dependent manner [32]. Lecanemab and its murine analogue mAb158 were specifically developed to recognize conformational epitopes in the N-terminal region of amyloid-beta.

Competitive IHC with 6E10 for the detection of plaques revealed that incubation with excess lecanemab resulted in a strong reduction in 6E10 staining, consistent with overlapping binding epitopes. Competitive IHC with mAb158 and excess lecanemab showed complete abolishing of staining, indicating a complete blocking of mAb158 binding, as expected due to the identical binding epitopes of the human and murine versions of lecanemab [2].

In contrast, competitive IHC of donanemab and 6E10 did not reduce the 6E10 signal. This finding is consistent with the distinct binding epitope of donanemab, which targets pyro-glutamate containing amyloid-beta aggregates, rather than epitopes recognized by 6E10. In control experiments, excess donanemab completely blocked mDona binding, as expected due to the identical binding epitopes of the human and murine versions of donanemab. Both pre- and co-incubation experiments yielded same results.

Competitive autoradiography using florbetaben confirmed specific tracer binding to beta-amyloid-beta plaques in AD brain tissue, consistent with previously published findings [21]. Increasing concentrations of the non-radioactive florbetaben completely abolished specific tracer binding. In contrast, neither co-incubation nor pre-incubation with lecanemab or donanemab at any tested concentration resulted in a detectable reduction of the florbetaben autoradiography signal. These results indicate that neither antibody interferes with florbetaben binding to fibrillar amyloid-beta deposits, suggesting distinct binding sites and absence of steric blocking or masking effects.

Ligand-binding assays performed on pooled AD frontal cortex homogenate further supported these findings. Florbetaben binding remained unchanged during both co-incubation and pre-incubation with lecanemab or donanemab at any used concentration. Together, these data support the conclusion that neither antibody interferes with florbetaben binding to fibrillar amyloid-beta, reinforcing the tracer’s suitability for detection of amyloid-beta-plaques in the presence of therapeutic antibodies.

Overall, these findings indicate that florbetaben binds to sites on amyloid-beta plaques that are distinct from those of lecanemab and donanemab, and that its binding to amyloid-beta plaques is not affected by either of these amyloid-beta-targeting drugs. The present results are in line with a recent preclinical study, which found no evidence of an interaction between amyloid PET and amyloid-beta targeting drugs in vivo [33]. The investigators treated tg-ArcSwe mice with the murine analogue of lecanemab (mAb158), and showed that these mice had [¹¹C]PiB binding levels both in vivo and on autoradiography that are comparable to controls, indicating that lecanemab did not interfere with [¹¹C]PiB binding to amyloid-beta plaques [33].

Therefore, any observed reduction in [¹⁸F]florbetaben PET signal in patients treated with lecanemab [9] or donanemab [10] are unlikely to result from a direct masking effect caused by these therapeutic antibodies. Recent use recommendations for donanemab [23] and the updated appropriate use criteria for amyloid PET [11] suggest that amyloid PET is suitable to monitor amyloid-beta changes during treatment with amyloid-beta targeting therapies, as also outlined in recent practical recommendations [34]. In this context, the presented in-vitro results confirm suitability of [¹⁸F]florbetaben PET for such monitoring.

While potential direct interactions between florbetaben and amyloid-beta targeting drugs have been excluded in this work, another potential indirect confounder to be considered is ARIA (Amyloid-related Imaging Abnormality). ARIA has been observed in some patients undergoing amyloid-beta targeting therapy and can co-localize with sites of amyloid removal on PET [35]. This raises the theoretical concern that ARIA induced by amyloid-beta targeting antibodies might alter local cerebral blood flow and thereby affect PET tracer uptake and clearance independent of amyloid-beta burden. However, several lines of evidence indicate that global amyloid PET is not affected by ARIA. First, ARIA is localized to limited brain regions only, whereas antibody therapies induce global amyloid PET reductions. Any local effect would not substantially influence the overall assessment of the amyloid burden as measured by PET. Of note, there have been heterogeneous histopathological findings when describing the amyloid-beta clearance mechanism, ranging from superficial plaque removal [36] to robust plaque clearance [37, 38], which underscore that the exact patterns and mechanisms of amyloid-beta removal using ATTs remain incompletely understood and warrant further investigation. Second, ARIA observed from treatment with lecanemab or donanemab typically occurs early and resolves within months, while amyloid-beta reductions measured by PET persist over time [39]. Third, donanemab trials reported comparable plaque reductions with standard and modified titration schemes despite markedly different ARIA rates [40]. Finally, the second-generation amyloid-beta targeting antibody trontinemab achieved rapid and marked amyloid-beta lowering, as measured with PET, without ARIA [41]. Collectively, these data indicate that amyloid PET signal decreases during therapy reflect amyloid-beta removal rather than ARIA-related tracer effects.

Limitations

Although this study provides crucial insights, several limitations must be acknowledged. The immunohistochemistry and autoradiography experiments were designed as proof-of-principle studies to qualitatively demonstrate signal displacement and were not intended to support high-resolution spatial, regional, or plaque-type–specific analyses. The microscopic data were not evaluated by a qualified pathologist, and detailed neuropathological characterization (e.g., plaque subtyping or regional profiling) was therefore not performed, as it was beyond the scope of the study. While a threshold-based quantitative analysis was applied to provide an objective assessment of overall staining and displacement patterns, this approach does not capture fine-grained spatial or intensity distribution features.

Conclusions

In conclusion, these findings indicate that [¹⁸F]florbetaben binds to sites on amyloid-beta plaques that are distinct from those of lecanemab and donanemab, and that its binding is not affected by either of these antibodies. Therefore, reductions in [¹⁸F]florbetaben PET signal in patients treated with amyloid-beta targeting therapies are unlikely to be driven by masking effects at the tracer binding site. Interpretation of in vivo PET signal changes must consider additional physiological and therapy-related factors; however, the present data argue against masking as an explanation for the observed signal reductions. While the current data are limited to in-vitro experiments, they further support the use of florbetaben PET for monitoring amyloid-beta changes during treatment with amyloid-beta targeting therapies.

Authors’ contributions

AM, JS, MR, EB, NKi performed the experimental analyses, including immunohistochemistry, autoradiography and ligand binding assays, and contributed to data interpretation and manuscript drafting. AM, NK, AJ led the manuscript writing, coordinated all correspondence and revisions, and contributed significantly to the study design and data integration. MC contributed to manuscript preparation, interpretation of findings, and contextualization within the broader clinical landscape of amyloid PET imaging and therapeutic monitoring. EM, AS provided clinical input and guidance on the application of amyloid PET in therapeutic monitoring. NK conceived the study, oversaw scientific coordination, and contributed to manuscript writing and data interpretation. All authors read and approved the final manuscript.

Funding

This research received no external funding.

Data availability

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study involved the use of previously collected, de-identified human brain tissue samples. Ethical approval for the use of these materials was obtained from the appropriate institutional review board, in accordance with applicable laws and regulations.

Consent for publication

Not applicable. This study did not involve individual participant data in any form (e.g., images, videos, or identifiable personal details).

Competing interests

Authors are employed by Lantheus or by Re:cognition Health (EM).