Alzheimer's disease (AD), a progressive neurodegenerative disorder (NDD) impacting 45 million people globally, poses a significant and pressing global concern (1). With no definitive cure and given its insidious nature, the economic burden of AD and related dementias is predicted to soar to $4.7 trillion by 2030, highlighting the urgency for targeted therapies to slow or halt the disease progression (2).
The search for effective AD treatments has been marked by disappointments, evident in the staggering 98% failure rate in Phase II and III clinical trials (3). Recent developments, like the FDA’s approval of Biogen/EISAI’s drug Leqembi and positive results coming from Elli Lilly’s Donanemab trial provide some hope (4). However, challenges remain, such as: only modest delay of disease progression; a need for real-world validation of clinical effects; and the high prevalence of serious adverse events (5). Addressing the complexity of NDDs like AD in clinical research and patient care requires a multi-target approach, innovative trial designs and tooling. For example, using AI-based multimodal co-variate adjustment when planning a trial and implementing adaptive trial designs can help handle the complex nature of AD, improve the success rate of its trials and address the shortcomings of approved therapies.
As the first disease modifying treatment options for AD patients after decades of failed attempts, monoclonal antibodies (mAb) against amyloid aggregates represent a scientific breakthrough. But their effect on slowing the disease progression is only moderate, and their risk of inducing devastating side-effects like brain edema or bleeding amyloid-related imaging abnormalities, short ARIA, is substantial (6).
Aducanemab, marketed under the brand name Aduhelm, was the first mAb targeting amyloid to be approved by the FDA in 2021. Its approval was based on its ability to reduce amyloid load, however, there are doubts regarding its clinical effectiveness. Aduhelm only showed significant benefits for the subset of patients that received the highest dose and all planned infusions (7,8). Another mAb targeting amyloid, Lecanemab (marketed as Leqembi), was approved earlier this year and was found to reduce cognitive decline by 27% after 18 months (9). Most recently a third anti-amyloid mAb – Donanemab – made the headlines for its positive results in May this year. Donanemab targets both the soluble and aggregate forms of Aβ42 and demonstrated a decrease in amyloid load, as well as a 35% slowing of cognitive decline (10). Although Lecanemab and Donanemab show great promise, uncertainties remain regarding the effectiveness of single target amyloid therapies to forestall cognitive and functional decline.
The pathophysiology of AD is complex, and experts believe that the focus on a single target – amyloid – is a reason for the moderate clinical efficiencies of approved therapies (3). The primary pathological changes in AD in the brain are characterized by the accumulation of amyloid-beta (Aβ) plaques, mainly derived from the Aβ42 variant type, and the formation of tau that cause neuronal loss and neurodegeneration. However, the development of AD is thought to be influenced by multiple complex biological pathways, including cholesterol transportation and metabolism mediated by apolipoprotein E (APOE), neuroinflammatory response, energy utilization, and vascular burden (11).
The interplay of the different pathways in AD offer potential targets for neuroprotective and neurorestorative treatments. These targets include tau, addressing neuroinflammation, and mitochondrial dysfunction (12, 13 ,14, 15). The complexity of AD demands a multifaceted treatment approach. Simultaneously targeting different areas of AD pathophysiology may lead to synergistic therapeutic effects, and hence better overall efficacy.
Although a multi-target approach may bring many benefits, it also presents many challenges, like multiple drug interactions, compounded adverse events, and the need to identify and monitor multiple biomarkers. Selecting meaningful biomarkers with high diagnostic and prognostic value and using them in the right context for patient screening and therapy monitoring is essential yet extremely challenging. The Alzheimer’s Disease Neuroimaging Initiative (ADNI) is tackling this challenge by collecting and facilitating the sharing of longitudinal data on clinical, imaging, genetic, and biochemical biomarkers among scientists. This initiative aims to help detect AD at the earliest possible stage and identify disease progression with diagnostic and prognostic biomarkers. However, to fully unlock the value contained in powerful datasets like ADNI advanced analytical tools are required.
The recent and rapid development of highly flexible, reusable artificial intelligence (AI) models has the potential to revolutionize drug development (16). These AI models incorporate multimodal data from imaging, electronic health records, laboratory results and genomics. This makes them well suited to deal with complex longitudinal data, such as that collected in ADNI. With the help of these AI models, patient populations can be precisely characterized. Also, the diagnostic, prognostic, and predictive value of biomarkers can be evaluated, established biomarkers can be enhanced, or new ones can be identified. This helps fast-track clinical trials in AD.
The AICURA Platform emerges as a pivotal tool for biopharma looking to accelerate and de-risk clinical drug development in AD. In the platform, we have combined three key technologies: a graph-based data layer that ensures distributed and interoperable data management; foundational models in neurodegeneration, capable of handling unlabeled and multimodal data; and a ready to use AI development kit, that employs these foundational models as a base to build powerful AI solutions.
AICURA’s innovative technologies can help biopharma future-proof its clinical programs by using data-driven clinical trials designs that effectively de-risk and accelerate the drug development process. For instance, AICURA can help improve the diagnosis of AD, predict disease prognosis, and treatment effectiveness, which helps in identifying subgroups of patients for clinical trials. Additionally, the platform enables biopharma to assess the predictive value of every biomarker, which can be very useful in the planning phases of trials.
1. de Aquino CH., et al. Methodological Issues in Randomized Clinical Trials for Prodromal Alzheimer's and Parkinson's Disease. Front Neurol. 2021. DOI: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8377160
2. Nandi A., et al. Global and regional projections of the economic burden of Alzheimer's disease and related dementias from 2019 to 2050: A value of statistical life approach. The Lancet. 2022. DOI: https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(22)00310-8/fulltext
3. Kim CK., et al. Alzheimer's Disease: Key Insights from Two Decades of Clinical Trial Failures. J Alzheimers Dis. 2022. DOI: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9198803/
4 Sims JR., et al. Donanemab in Early Symptomatic Alzheimer Disease: The TRAILBLAZER-ALZ 2 Randomized Clinical Trial. JAMA. 2023 DOI: https://jamanetwork.com/journals/jama/article-abstract/2807533
5. Osborne OM., et al. Anti-amyloid: An antibody to cure Alzheimer's or an attitude. iScience. 2023. DOI: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10425904/
6. Salloway S., et al. Amyloid-Related Imaging Abnormalities in 2 Phase 3 Studies Evaluating Aducanumab in Patients With Early Alzheimer Disease. JAMA Neurol. 2022. DOI: https://pubmed.ncbi.nlm.nih.gov/34807243/
7. Biogen News Releases. Biogen Plans Regulatory Filing for Aducanumab in Alzheimer’s Disease Based on New Analysis of Larger Dataset from Phase 3 Studies. 2019. DOI: https://investors.biogen.com/news-releases/news-release-details/biogen-plans-regulatory-filing-aducanumab-alzheimers-disease
8. Ebell MH., et al. Why Physicians Should Not Prescribe Aducanumab for Alzheimer Disease. Am Fam Physician. 2022. DOI: https://pubmed.ncbi.nlm.nih.gov/35426626/
9. van Dyck CH., et al. Lecanemab in Early Alzheimer's Disease. N Engl J Med. 2023. DOI: https://pubmed.ncbi.nlm.nih.gov/36449413/
10. Taylor MP. Lilly shares full pivotal data on Alzheimer's drug, revealing how it shapes up against Leqembi. Fierce Biotech. 2023. DOI: https://www.fiercebiotech.com/biotech/lilly-shares-full-pivotal-data-alzheimers-drug-revealing-how-it-shapes-against-leqembi
11. Masters CL., et al. Alzheimer's disease. Nat Rev Dis Primers. 2015. DOI: https://pubmed.ncbi.nlm.nih.gov/27188934/
12. Osborne OM., et al. Brain endothelium-derived extracellular vesicles containing amyloid-beta induce mitochondrial alterations in neural progenitor cells. Extracell Vesicles Circ Nucl Acids. 2022. DOI: https://pubmed.ncbi.nlm.nih.gov/36649440/
13. Sun N., et al. Single-nucleus multiregion transcriptomic analysis of brain vasculature in Alzheimer's disease. Nat Neurosci. 2023. DOI: https://pubmed.ncbi.nlm.nih.gov/37264161/
14. Kim J., el al. Evolutionarily conserved regulators of tau identify targets for new therapies. Neuron. 2023. DOI: https://pubmed.ncbi.nlm.nih.gov/36610398/
15. Lonnemann N., et al. The NLRP3 inflammasome inhibitor OLT1177 rescues cognitive impairment in a mouse model of Alzheimer's disease. Proc Natl Acad Sci U S A. 2020. DOI: https://pubmed.ncbi.nlm.nih.gov/33257576/
16. Moor M., et al. Foundation models for generalist medical artificial intelligence. Nature. 2023. DOI: https://www.nature.com/articles/s41586-023-05881-4#citeas