Aims

Lecanemab (BAN2401) is a humanized IgG1 monoclonal antibody that selectively targets soluble aggregated Aβ species. Clarity AD is an 18-month, multicenter, double-blind, placebo-controlled, parallel-group Phase 3 study in participants with early AD. The impact of treatment with lecanemab on the association of change from baseline (CFB) in tau SUVR and baseline tau and amyloid levels, in a tau sub-study of Clarity AD, are described.

Methods

During the Core Clarity AD study, patients had the option to participate in sub-studies that evaluated longitudinal changes in amyloid and tau PET. Tau ([¹⁸F]-MK-6240) and amyloid (Florbetaben) SUVR from 8 composite regions were available at baseline for 333 amyloid-positive subjects (169 treated with lecanemab). Correlation matrices using Spearman's rank correlation coefficients were generated to explore those associations across brain regions.

Results

In this sub-study of 203 MCI and 130 mild AD participants, median age was 72 years (range 50-88 years), 51.4% were female and 56.8% were APOE4 carriers, comparable to demographics of all participants in the Core study. Baseline associations confirm the diffuse nature of amyloid deposition and the spatio-temporal distribution of tau across brain regions, with higher tau levels in early temporal brain regions, and higher tau levels with higher baseline amyloid levels. Higher tau levels at baseline were associated with higher tau accumulation across brain regions with disease progression (placebo group). This association is stronger between early temporal tau regions and the rest of the brain. Lecanemab treatment slowed progression of tau in medial temporal regions relative to placebo, and also disrupts the correlations between baseline tau and CFB in tau levels suggesting disruption of tau accumulation across brain regions.

Conclusions

Lecanemab treatment impacts tau accumulation across brain regions.

Abstract Body

Abstract Body

Donanemab, an IgG1 monoclonal antibody directed at clearing brain amyloid plaques, has shown promising results in slowing clinical progression of early symptomatic Alzheimer’s disease (AD). This presentation will provide an overview of the ongoing donanemab clinical trial program, synthesizing evidence of clinical efficacy and safety across the TRAILBLAZER-ALZ trials. Findings will be compared across completed phase 2 and 3 trials, and the latest available data will be reported on progress in trials for preclinical AD and on risk of amyloid-related imaging abnormality-edema/effusion (ARIA-E). Assessing evidence across these trials will yield new insight into amyloid targeting therapies, including biomarker indications of efficacy, and safety in different subpopulations as defined by patient and clinical characteristics. This information will be useful for advancing our understanding of the efficacy and safety of amyloid targeting therapies for AD.

Aims

To identify baseline characteristics associated with achieving rapid amyloid clearance determined as < 24.1 centiloids after the first 24 or 52 weeks of treatment with donanemab.

Methods

Probability of rapid amyloid clearance based on 16 baseline characteristics, including demographics, APOE ε4 status, Mini-Mental State Examination (MMSE) score, imaging (amyloid PET, tau PET, volumetric MRI) and plasma measurements were examined with three models (Random Forest, Gradient Boosting, and Penalized Regression). Analyses included longitudinal 24 weeks florbetapir images in TRAILBLAZER-ALZ (NCT03367403, N=116) phase 2 and TRAILBLAZER-ALZ4 (NCT05108922, N=66) phase 3 trials in early symptomatic Alzheimer’s disease. Participants with amyloid pathology and low-medium tau level were enrolled in the phase 2 trial while amyloid pathology presence only was required in TRAILBLAZER-ALZ4.

Results

Robust and consistent (across methods and datasets) predictions of amyloid clearance at week 24 are exhibited with Areas under the Receiver Operating Characteristic Curve (AUC) ranging between 0.68 and 0.71. Baseline amyloid level emerged as the most significant predictor. Baseline tau level was also associated with amyloid clearance at week 24 in TRAILBLAZER-ALZ4 dataset with the whole spectrum of tau pathology.

Conclusions

Analysis of amyloid clearance at week 24 suggests that baseline amyloid level is the strongest predictive factor while baseline tau level and age may also be considered to understand the individual amyloid clearance. Analysis of amyloid clearance at week 52 will also be presented. Further research with larger and diverse cohorts is warranted.

Abstract Body

Several anti-amyloid immunotherapies have been shown to be associated with greater brain volume and ventricular enlargement compared to placebo. Originally described in the first vaccination trial (AN1792) excess volume loss has also been seen with a number of monoclonal antibodies including most recently donanemab and lecanemab. This “paradoxical” volume loss or “pseudo-atrophy” is an area of great interest and controversy, with potential explanations ranging from it being a marker of effective amyloid removal, through to it being an indication of treatment-related toxicity. I will review these different hypotheses and consider their plausibility and their compatibility with recent data.

Aims

SKYLINE (NCT05256134) was a secondary prevention study in Alzheimer's disease enrolling amyloid-positive, cognitively unimpaired 60- to 80-year-old individuals at risk of or at the earliest stages of AD, with subcutaneous gantenerumab or placebo treatment planned for four years – the study was terminated early and available data analyzed. We assessed the impact of blood-based prescreening, using Elecsys® Phospho-Tau (181P) and Apolipoprotein E4 plasma prototype assays to screen-out participants with a low likelihood of amyloid positivity by positron emission tomography (PET) or cerebrospinal fluid (CSF) testing.

Methods

Participants either underwent blood-based prescreening, or proceeded directly to main screening: Irrespective, plasma biomarkers were measured using Elecsys^® prototype immunoassays for all participants that provided a blood sample. Amyloid pathology confirmation by PET or CSF was undertaken for participants who did not fail other screening criteria. Performance metrics were calculated for the two different screening populations, including positive and negative predictive value for amyloid pathology (where applicable), rule-out rate and observed amyloid positivity prevalence.

Results

The amyloid positivity prevalence by CSF/PET was as low as expected (approximately 12%) and underlines the utility of a prescreener in this population. In summary, prescreening enriched the amyloid positivity prevalence from 12.3% to 19.3% while avoiding more than 40% of unnecessary downstream screening assessments with a very high negative predictive value (98%), i.e. without falsely prescreening out many true amyloid positive subjects.

Conclusions

Data provides insights into the potential value of implementing a prescreener using Elecsys^® Phospho-Tau (181P) and Apolipoprotein E4 in a clinical trial such as SKYLINE by reducing unnecessary downstream screening assessments such as cognitive batteries, magnetic resonance imaging, and PET/CSF testing and thereby reducing participant burden, improving enrolment timelines, and reducing screening costs.

Abstract Body

Objectives: Historical comparison data and simulated placebo data may be a viable option to improve clinical trials and reduce the number of patients who are required to receive placebo when reproducible and well characterized data are available. Herein, we review some of the relevant regulatory history relevant of these approaches and to highlight the challenges of this approach.

Methods: Relevant regulatory history and recent historical/simulated placebo literature were reviewed. Key criteria for replacing some or all of a placebo group with a historical/simulated placebo along with challenges of this approach were summarized. Modelling and simulation approaches to producing a simulated placebo were explored.

Results: FDA has accepted properly applied historical controls and simulated placebo. These require understanding the natural disease course. Examples where these approaches have been used include infectious disease trials and in neuroscience epilepsy development programs. In epilepsy, for example, nine monotherapy trials with reproduceable placebo response were utilized to generate a historical placebo for new monotherapy indications in epilepsy studies of agents with add-on approvals. Key criteria for simulated placebo use include robust efficacy thresholds, well-established placebo progression rates, low data variability, robust effectiveness evidence, and reproducible treatment response. Simulated placebo data can be generated through multiple approaches, including model-based Bayesian methods, artificial intelligence approaches, and Quantitative Systems Pharmacology modelling. Key challenges to implementation exist, including changes in standard of care, lack of applicable data, variability in existing data, selecting a predictable group to study, and lack of placebo safety data.

Conclusions: The use of historical data and simulated placebo in clinical study design may help in efficiently move drug development programs forward and expose fewer individuals to placebo, while expediting availability of new therapies.

Abstract Body