Highlights from AD/PD 2026
At the International Conference on Alzheimer’s and Parkinson’s Diseases and Related Neurological Disorders (AD/PD) meeting (17–21 March, Copenhagen, Denmark), more than 6,000 attendees shared data from over 140 clinical trials investigating more than 15 biological mechanisms. PSI’s Valentin Sakhovsky, Therapeutic Area Lead, Neurology, Gyorgy Andor, Director of Medical Monitoring in Neurology, and Luke Shelton, Lead Scientific Writer, attended to identify what matters for trial execution in chronic neurodegenerative diseases, drawing on the latest developments from basic research, clinical research, and industry perspectives.
In its most comprehensive program to date, AD/PD emphasized four main themes: building on recent disease-modifying therapies in Alzheimer’s disease, using those lessons to accelerate truly disease-modifying therapies in Parkinson’s, implementing AI strategies for decision-making and hypothesis generation in clinical development and trial operations, and revolutionizing the use of biomarkers for patient stratification and endpoint measurement. The question for sponsors is: How will recent developments and the discussions held at AD/PD impact Alzheimer’s and Parkinson’s trials of the future?
Alzheimer’s disease: Great advances, persistent challenges
The drug discovery ecosystem in AD: From fragmentation to integration
In early AD, disease-modifying therapies (principally lecanemab and donanemab, both anti-amyloid monoclonal antibodies) are now a reality. The next step is refinement and optimization: when and how to use disease-modifying therapies in patients with AD, how to treat patients with later-stage disease, and how to address disease where anti-amyloid monoclonals show limited utility are the main questions.
The move towards diversity and integration (across diagnosis, biomarkers, and therapeutic approaches) was evident. AD/PD attendees are embracing multimodal frameworks, recognizing that neurodegenerative diseases are driven by multiple interacting pathways rather than by a single mechanism.
Sponsors must prepare for a landscape where combination approaches and sequential strategies replace monotherapy assumptions. Trial designs that assume homogeneity will underperform; those that stratify by biomarker profile, genetic risk, and disease trajectory will identify responders faster and with greater confidence.
We expect a new wave of development of AD therapies going beyond anti-amyloid: gamma secretase modulators, tau antisense oligonucleotides, or anti-tau antibodies for moderate disease, and in late-stage AD, perhaps cell therapies.
Safety concerns: Managing ARIAs
- Amyloid-related imaging abnormalities (ARIAs) are the key safety risk with anti-amyloid therapies
- They present as brain edema and/or microhemorrhages
- ARIAs can be symptomatic (or asymptomatic) and can be life-threatening
- Patients require regular MRI monitoring, creating significant operational burden, and ARIAs are associated with treatment interruption, dose adjustments, or discontinuation
- The general consensus at AD/PD 2026 is that ARIAs are no longer viewed as a reason to halt development of potentially viable therapeutics – they are a known and manageable class effect
- Identifying who is at risk of ARIAs (through genetic testing – primarily APOE4) could mean biomarker-driven patient selection, helping overcome operational challenges (MRI capacity and delivering treatment at scale)
Blood-based biomarkers: Revolutionary, but work still needed
Throughout AD/PD, multiple presentations showcased data on plasma-derived phosphorylated tau 217 (pTau217). It is widely considered the key breakthrough in Alzheimer’s biomarker development: accurately and non-invasively identifying AD pathology in the preclinical stage with great specificity. Quantification of pTau217 can also provide earlier indications of treatment benefit compared to amyloid-β PET and CSF biomarkers – for the longest time, the best methods in evaluating AD in patients [1]. Developments in blood-based biomarkers are transforming how patients are evaluated. For primary care doctors, they enable simple, rapid, scalable testing. However, standardization and partnership with commercial vendors still need improvement.Commercially available tau biomarkers are able to assess different components of multidimensional AD. Increased clinical trial complexity is emerging, but also greater accuracy in patient and disease stratification, target engagement, and the detection of meaningful treatment effects. Sponsors must partner closely with diagnostic companies, contribute to validation efforts, and design protocols that account for current limitations while positioning for future regulatory acceptance.
We can identify the existing gaps in evaluating neuroinflammation, an area often overshadowed in AD research. Two decades ago, there was an expectation that advancements in drug discovery would outpace biomarker development; however, currently we have the complexity of trying to translate scientific discoveries into tangible patient insights without a full complement of tools to do so.
Latest data: AD snapshots
- Novo Nordisk’s phase 3 evoke and evoke+ trials investigated the efficacy and safety of oral semaglutide in patients with amyloid-confirmed AD. While semaglutide was found to be safe and well-tolerated, it did not demonstrate efficacy in slowing clinical progression of AD compared with placebo, leading to discontinuation of the program
- Johnson & Johnson’s phase 2b Autonomy trial of the anti-p-tau antibody posdinemab failed to demonstrate clinical benefit in early AD across primary and secondary cognitive and functional endpoints. These results do not invalidate tau as a therapeutic target; tau PET imaging suggested that limiting propagation of tau, rather than reducing existing burden, may be critical for disease modification
How to accelerate AD clinical trials
Clinical trial execution remains a frustrating bottleneck in the development of therapies for AD.
- Patient recruitment remains slow, due to notable challenges in identifying the right patients at the right time (considering the individual trajectories of AD)
- Trials are long, large, and costly
- Cognitive measurements are laborious and can reduce clinician capacity – digitalization of cognitive measurements can accelerate testing
- There is a disconnect between clinical relevance and mechanical readouts; patients care about functional outcomes (cognition, memory, QoL) however statistically significant improvements in primary endpoints do not always translate to meaningful improvements to the patient
The growing call for shorter and more efficient phase 2 trials was evident. An earlier indication of the likelihood of success in a phase 2 trial (potentially through surrogate markers) would shorten the time to phase 3. Trials themselves need faster recruitment strategies, better patient identification, and expansion of cognitive and functional endpoints.
AI as a co-scientist
The emphasis on AI has shifted from dealing with quantity and volume to improving quality and decision-making, particularly with the incorporation of agentic AI. The clear focus at AD/PD was the use of AI as a co-scientist, augmenting human expertise rather than replacing it. A co-scientist could enhance hypothesis generation, data interpretation, and provide cross-study insights – things that even the most skilled and studious expert would struggle to provide. AI-based voice agents capable of performing cognitive tests can improve the speed of screening and enrolling patients, and were shown to perform more consistently than humans.
In neurodegenerative clinical trials, AI is being used to improve the probability of success for selected drug targets, rather than simply expanding the number of targets under evaluation. By enabling earlier and more informative readouts, AI has the potential to streamline trial timelines, and accelerate the delivery of effective therapies to patients.
As always with AI, data quality, curation, and access are the bottlenecks, particularly given its reliance on published datasets and the risks of bias and hallucination.
Parkinson’s disease: The pre-disease modification bottleneck
Approved treatments for PD (levodopa, carbidopa, deep brain stimulation, amongst others) are effective at treating the symptoms of the disease, but they do not slow neurodegeneration. At this year’s AD/PD, much discussion was held on the recent developments pointing to an encouraging trajectory towards disease-modifying therapies for Parkinson’s. However, there are significant hurdles to be overcome (and lessons learned from AD) to ensure the new wave of disease-modifying therapies has the right armamentarium to succeed.
Unmet needs with current symptomatic treatments
Considering current unmet needs in PD, attendees suggested that different non-dopaminergic and non-motor features are inadequately addressed with current therapies:
- Freezing of gait – associated with loss of independence, reduced quality of life, and falls (one of the main causes of death in Parkinson’s). Often poorly responsive to dopaminergic therapies [3]
- Constipation –laxatives can be effective, but constipation is often under-recognized, under-treated, and managed symptomatically rather than targeting the underlying pathophysiology [4]
- Psychosis – particularly in advanced disease, hallucinations and delusions are associated with increased caregiver burden, increased rates of institutionalization, and increased mortality [5]. Antipsychotics can worsen motor symptoms [6], thus safer, more effective therapies that don’t compromise motor function are needed
Expanding endpoints to capture non-motor features is essential, but it complicates trial design. Validated instruments exist for some domains, but not all, and composite endpoints that balance motor and non-motor features require careful construction to avoid diluting treatment signals.
Disease modification: What does it mean?
Disease modification means slowing, halting, or altering the underlying neurodegenerative processes in PD through durable biological and structural changes (e.g., preserving dopaminergic neurons or maintaining the structure of the putamen).
Measuring disease modification was highly debated at AD/PD. Evidence may include:
- Sustained clinical benefit beyond treating symptoms
- Reduced need for symptomatic therapies (for example, delayed initiation or reduced dose of levodopa)
- Changes in the underlying biology demonstrated through biomarkers
Trial approaches to evaluate disease modification include delayed-start designs (whether earlier treatment leads to lasting benefit vs. delayed initiation), washout designs (whether the effect persists after treatment is stopped), or using biomarker-based endpoints that capture structural or neurodegenerative changes.
However, challenges to these trial designs include unethically leaving patients untreated for long periods, symptomatic treatments confounding or confusing the interpretation of results, and the lengthy disease progression in PD – all of which mean there is no universally accepted definition for measuring disease modification in PD: this remains one of the central challenges in drug development.
Despite intense innovation, PD remains an area where scientific promise has outpaced clinical translation. That is, the lack of validated biomarkers, patient stratification strategies, and endpoints sensitive or meaningful enough to evaluate efficacy means that, in comparison with AD, PD has a lot of catching up to do.
Latest data: PD snapshots
- AC Immune’s phase 2 VacSYn trial of a first-in-class alpha-synuclein monomer [7] reported positive interim safety and tolerability data – a favorable benefit/risk ratio, impact on total and aggregated alpha-synuclein, and a trend to slowing of disease progression
- Encouraging additional data from Roche’s phase 2b PADOVA trial [8] were presented at AD/PD – including indications of the first biomarker indicative of an effect on underlying PD pathology. Despite just missing the primary endpoint, prespecified analyses indicated a trend toward reduction of motor progression, among other clinical endpoints. The follow-up PARAISO phase 3 trial has started and is the only large, global disease-modifying therapy phase 3 trial currently ongoing
Elusive breakthroughs
Trials historically tended to enroll moderate-to-advanced disease patients; however, it is challenging to demonstrate disease modification in these patients. More recent trials try to understand pre-PD and PD biology, and as such are enrolling early-stage or prodromal patients – who are harder to identify, and have slow progression (and longitudinal, challenging endpoints).
Parkinson’s disease is a heterogeneous, multi-pathway disorder, not a single, uniform disease entity. Multiple biological processes contribute to its onset and progression – and as a result, trials evaluating therapies targeting single pathways repeatedly fail to demonstrate disease-modifying effects.
This is driving a shift towards combination strategies (following lessons learned from oncology, and applied in AD). Over time, we expect the field to move towards a precision-medicine model, where multi-modal treatment regimens are required to achieve disease modification.
Sponsors must weigh the scientific rationale for targeting multiple pathways against the operational and financial burden of executing more complex programs. The data increasingly support combination approaches; the question is how to execute them efficiently.
What this means for your neurodegenerative program
Alzheimer’s disease and Parkinson’s disease are not uniform conditions, but are shaped by biological, clinical, and demographic complexity. The presence, prevalence, and impact of key biomarkers vary across race and ethnicity, and comorbidities (more common in underserved populations) further influence disease biology and outcomes. This reflects a critical challenge for sponsors: without diverse and representative populations, advances in biomarkers, stratification, and precision medicine risk being limited.
The chronic neurodegenerative disease field is moving toward a future defined by precise patient selection, multimodal biomarkers, and thoughtful combination therapies, which will complicate clinical trial design and operational execution. Faster, smarter, and more inclusive trials, integrating AI and digital tools, learning from real-world data, and addressing long-standing bottlenecks in recruitment and endpoint validation are the progress we expect in the next few years.
Two perspectives from speakers at AD/PD captured this direction: “I can’t think of a single chronic age-related illness that is cured with a single medicine” and “There will never be precision medicine for anyone without accurate representation of everyone”.
Discuss your neurodegenerative trial strategy
The breakthroughs are real. Execution is what separates programs that reach patients from those that stall in development. Contact PSI’s neurology team for expert guidance on running your Alzheimer’s and Parkinson’s disease trials on time.
Abbreviations
AD, Alzheimer’s disease;
CSF, cerebrospinal fluid;
PD, Parkinson’s disease;
PET, Positron Emission Tomography;
pTau217, phosphorylated Tau-217.
References
1. Leuzy, A., et al., Considerations in the clinical use of amyloid PET and CSF biomarkers for Alzheimer’s disease. Alzheimers Dement, 2025. 21(3): p. e14528.
2. Tsakiri, M., et al., Lipid-based nanoparticles and RNA as innovative neuro-therapeutics. Front Pharmacol, 2022. 13: p. 900610.
3. Fox, S.H. and Cardoso, F., Unmet Needs in Parkinson’s Disease. Mov Disord Clin Pract, 2023. 10(Suppl 2): p. S47–s49.
4. Al-Wardat, M., et al., Constipation and pain in Parkinson’s disease: a clinical analysis. J Neural Transm (Vienna), 2024. 131(2): p. 165–172.
5. Gryc, W., et al., Hallucinations and Development of Dementia in Parkinson’s Disease. J Parkinsons Dis, 2020. 10(4): p. 1643–1648.
6. Pirker, W., et al., Management of Parkinson’s disease psychosis-a European perspective. Lancet Reg Health Eur, 2026. 62: p. 101569.
7. National Institutes of Health. A Study to Evaluate the Effects of ACI-7104.056 in Patients With Early Stages of Parkinson’s Disease (VacSYn). Available from: https://clinicaltrials.gov/study/NCT06015841.
8. Nikolcheva, T., et al., A Phase 2b, multicenter, randomized, double-blind, placebo-controlled study to evaluate the efficacy and safety of intravenous prasinezumab in early-stage Parkinson’s disease (PADOVA): Rationale, design, and baseline data. Parkinsonism Relat Disord, 2025. 132: p. 107257.
