More than a hundred years have passed since Alois Alzheimer first described the pathological hallmarks of what would become the most common form of dementia. The journey from microscopic plaques and tangles to today’s molecularly-informed diagnosis tells a story of persistent scientific progress. In the late 1990s, cerebrospinal fluid (CSF) biomarkers—Aβ42, total tau (t-tau), and phosphorylated tau (p-tau)—emerged as the first clinical tools to identify Alzheimer’s Disease (AD) pathology in living individuals. This laid the foundation for the introduction of amyloid PET imaging in the 2000s, followed by tau PET in the 2010s1,2.
In the 2020s, we find ourselves entering a new diagnostic era—where plasma biomarkers promise to democratize access to early detection by offering minimally invasive, scalable, and cost-effective solutions.
2024 Diagnostic Framework: A New Model Anchored in Biology
At the heart of the revised 2024 criteria for Alzheimer’s diagnosis lies a conceptual shift: from syndromic diagnosis to biology-based classification3. This framework introduces Core 1 and Core 2 biomarkers. Core 1, consisting of Aβ42 and p-tau variants, reflects the early pathogenic processes and is critical for confirming AD pathology, even in asymptomatic individuals. Core 2 biomarkers, like MTBR-tau243 and tau PET, reflect later-stage tau aggregation and offer insights into disease severity3.
Notably, the criteria adopt a dual-axis staging approach: one based on biological progression (stages A–D) and the other on clinical-functional status (stages 0–6). This orthogonal system acknowledges a key truth in AD: pathology and symptoms do not always progress in tandem. By separating these axes, clinicians can better personalize prognosis and therapeutic strategies3.
Bridging Bench and Bedside: Clinical Guidelines and Biomarker Readiness
Recognizing the transformative potential of blood biomarkers, the Alzheimer’s Association has convened a panel of international experts to craft clinical practice guidelines. These are grounded in rigorous systematic reviews and the GRADE methodology3. The goal is not just to validate biomarkers in ideal research conditions but to ensure they are actionable in everyday practice, from primary care to specialized neurology clinics3 .
This is crucial because today’s clinicians face a new set of patient questions:” Am I at risk? What stage is my disease? Can it be slowed? Is my treatment working?” Biomarkers, particularly when integrated into decision-making tools, will play a central role in answering them.
Therapeutic Milestones: From Symptom Management to Disease Modification
Historically, treatments for AD were symptomatic—acetylcholinesterase inhibitors (donepezil, rivastigmine, galantamine) and the NMDA antagonist (memantine) provided modest benefit. However, the FDA approvals of lecanemab (Leqembi) and donanemab (Kisunla) mark the beginning of an era focused on disease modification4,5.
Lecanemab, an anti-amyloid monoclonal antibody, is now approved in the U.S., EU, and numerous countries across Asia and the Middle East. It is being transitioned from biweekly infusions to more patient-friendly monthly IV or subcutaneous options4. Donanemab, likewise, has shown promising outcomes with a modified dosing strategy that reduces ARIA-E events, particularly in APOE4 carriers—a critical consideration in personalized treatment5.
The TRAILBLAZER and AHEAD studies reflect a new preventive focus, enrolling cognitively normal individuals with biomarker-confirmed pathology. These trials aim not just to treat AD—but to delay or even prevent it6,7.
New Frontiers: Metabolic Modulation and GLP-1 Agonists
Semaglutide, known for its use in diabetes and obesity, is now being repurposed for Alzheimer’s in the EVOKE and EVOKE+ trials8. These trials span 40 countries, enrolling participants aged 55–85 with mild cognitive impairment or early AD. Preliminary data from over 15,000 pooled participants show fewer dementia events among those treated with GLP-1 receptor agonists8. If successful, this could signal a paradigm shift— leveraging immune modulation and metabolic regulation to alter neurodegenerative trajectories.
Therapy Pipeline: Expanding Beyond Amyloid
According to a 2024 review by Jeff Cummings, the global AD pipeline now includes 127 unique drugs in 164 trials. While 76% are still focused on disease modification, there's increasing diversification: 42% are small molecules, 34% are biologics, 12% target cognitive enhancement, 13% address behavioral symptoms, and 31% are repurposed compounds. This evolution reflects a growing recognition that AD is a multifactorial disease, driven not only by amyloid and tau, but also by inflammation, synaptic dysfunction, oxidative stress, mitochondrial abnormalities, and vascular contributions9.
Genotype-Informed Care and Real-World Data Networks
The ALZ-NET platform, a real-world data registry, reveals critical genotype trends: over a quarter of enrolled participants carry the E3/E4 APOE genotype, reinforcing the need for genetic stratification in trials and clinical practice10.
Complementing this is the ALZ-NET Imaging Network, launched in 2025, which aims to streamline imaging data collection, enable AI validation, and support a more integrated ecosystem for AD care11,12.
The Prevention Paradox: Why Early Action is Still Rare
Despite a rising global awareness of Alzheimer’s Disease (AD), prevention remains elusive in clinical practice. About 90% of Mild Cognitive Impairment (MCI) and 75% of dementia cases go undiagnosed13. This underdiagnosis is not due to a lack of tools but a lack of public and clinical engagement. Individuals often delay seeking evaluation with perceptions such as: “I’m still healthy,” “I have no time,” “It’s too expensive,” or “Nothing can be done.”
This inertia stems from gaps in motivation, insight, education, time, and access—all of which limit screening uptake even among at-risk individuals. Without systemic solutions to these barriers, early-stage interventions risk remaining underutilized13.
Screening Gaps: A Diagnostic Dilemma
Current diagnostic modalities for AD, while biologically robust, are poorly suited for large-scale screening. CSF biomarkers and amyloid PET, though highly accurate, are invasive or costly. Cognitive tests are accessible but often unreliable in the prodromal stage14.
Blood biomarkers (e.g., plasma p-Tau217) have emerged as a promising middle ground— balancing cost and diagnostic yield—but further validation and infrastructure are needed for widespread adoption. What’s missing is a rapid, affordable, and accurate tool deployable at scale in primary care or community settings14.
A New Frontier: Retinal Imaging for Alzheimer’s Disease
To address these limitations, researchers introduced retinal imaging as a non-invasive, scalable screening tool. By leveraging advances in AI and ophthalmic imaging, this novel approach analyzes the microvascular architecture of the retina, where early neurodegenerative changes mirror those in the brain.
Pioneered by Professor Carol Cheung and collaborators, AI-based models trained on over 16,000 fundus photographs across four countries (HK, SG, UK, USA) achieved internal validation with AUC 0.93 and external validation with AUC of 81%15,16.
This supports the eye-brain axis hypothesis in AD, where changes in vascular tortuosity, venule/arteriole diameter, and branching angles are detectable even in early disease stages15,16.
The Holistic Path Forward: Integration Across Domains
In light of these findings, the emerging AD paradigm calls for a multi-modal diagnostic and care strategy:
- Biomarkers (CSF, plasma, PET)
- Cognitive tools (e.g., MoCA, MMSE)
- Ophthalmic screening (retinal AI analysis)
- Genetics and risk profiling (e.g., APOE4 status)
Combined, these tools can inform individualized prevention, stratify risk, and support early, targeted interventions.
Moreover, real-world data registries like ALZ-NET and imaging networks can enhance longitudinal tracking, therapeutic optimization, and population-level surveillance, vital for translating innovations into health system practice.
Conclusions
ASAD 2025 highlights a major shift in the Alzheimer’s Disease (AD) landscape, driven by breakthroughs in diagnostics, therapeutics, and prevention. The 2024 diagnostic framework introduces a biology-based classification using core biomarkers and dual-axis staging to better guide prognosis and treatment. Innovative tools such as plasma biomarkers and AI-powered retinal imaging offer scalable, non-invasive options for early detection. Meanwhile, therapies like lecanemab, donanemab, and GLP-1 agonists represent the transition toward disease modification. Yet, underdiagnosis and limited public engagement, particularly in primary care, remain challenges. A multi-modal strategy integrating diagnostics, cognitive tools, genetics, and real-world data is crucial to deliver personalized, accessible, and equitable AD care.
Our correspondent’s highlights from the symposium are meant as a fair representation of the scientific content presented. The views and opinions expressed on this page do not necessarily reflect those of Lundbeck.
