Disease Modification in Parkinson's Disease: α-Synuclein, GLP-1, and Emerging Targets

Despite decades of effort, no therapy has been proven to slow, halt, or reverse the neurodegeneration of Parkinson's disease. Every medication used in PD today is purely symptomatic. Yet the field has never been closer to a disease-modifying breakthrough: α-synuclein immunotherapy (prasinezumab) is advancing to Phase III after consistent signals across two trials, GLP-1 receptor agonists have produced the most provocative positive Phase 2 result in years (LIXIPARK), and genetically targeted therapies for LRRK2 and GBA1 carriers are entering pivotal testing. This article reviews the history of disease-modification attempts, the current landscape of active programs, and the key trials that define each therapeutic strategy.

The Challenge of Proving Disease Modification

Demonstrating that a drug slows neurodegeneration — rather than simply masking symptoms — is the central methodological challenge of PD disease-modification trials. Every approach has limitations:

• Simple parallel-group designs cannot distinguish between symptomatic and neuroprotective effects. A drug that improves UPDRS scores could be doing so by treating symptoms, not slowing disease
• Washout designs (stop the drug, see if benefit persists) are confounded by long-duration symptomatic effects and carry-over pharmacology. ELLDOPA showed persistent clinical benefit after washout, but DaT-SPECT paradoxically suggested greater dopamine transporter loss with levodopa — raising more questions than answers
• Delayed-start designs (randomize to early vs late treatment start) are the most rigorous approach. Both ADAGIO and LEAP used this design, with conflicting results
• Imaging biomarkers (DaT-SPECT, neuromelanin MRI) offer objective measures but may not perfectly correlate with clinical progression and can be confounded by pharmacological effects on the transporter itself
• α-Synuclein SAA now enables biological confirmation of synucleinopathy at enrollment, ensuring that disease-modification trials enroll patients who actually have PD pathology (~7% of clinically diagnosed PD are SAA-negative)

Historical Disease-Modification Attempts

Understanding the history of failed and ambiguous trials is essential context for interpreting the current landscape.

Selegiline — DATATOP (1989)

• DATATOP (N=800): Selegiline (deprenyl) 10 mg/day delayed the need for levodopa by ~9 months vs placebo. Tocopherol (vitamin E) showed no benefit
• Initially interpreted as potential neuroprotection (MAO-B inhibition reducing oxidative stress). However, selegiline has mild symptomatic effects that confound interpretation
• Later consensus: the benefit was primarily symptomatic, not disease-modifying

Levodopa — ELLDOPA (2004)

• ELLDOPA (N=361): Levodopa 150/300/600 mg/day vs placebo for 40 weeks, then 2-week washout. Higher levodopa doses produced greater clinical benefit even after washout (600 mg: net improvement vs baseline; placebo: worsened by 7.8 UPDRS points)
• The paradox: DaT-SPECT imaging showed greater dopamine transporter loss with levodopa (−7.2% vs −1.4%, P=0.036) — the opposite of what would be expected if levodopa were neuroprotective
• Interpretations: (1) levodopa is neurotoxic (contradicted by long clinical experience), (2) levodopa has a long-duration symptomatic effect masking progression, or (3) levodopa pharmacologically downregulates the dopamine transporter (confounding the imaging biomarker)
• The debate continues, but the LEAP trial subsequently provided reassurance

Rasagiline — ADAGIO (2009)

• ADAGIO (N=1,176): Delayed-start design. Early rasagiline 1 mg/day met all three hierarchical endpoints for disease modification (UPDRS worsening: 2.82 vs 4.50, P=0.02)
• However: The 2 mg/day dose failed to meet criteria (P=0.60) — a biologically implausible dose-response that fundamentally undermined confidence in the result. If the drug truly modified disease, 2 mg should have worked at least as well as 1 mg
• Post hoc explanations (2 mg had more symptomatic effect that "washed in" faster, narrowing the early/delayed-start gap) were not convincing to regulators or the broader field
• Rasagiline is not marketed as disease-modifying

Levodopa — LEAP (2019)

• LEAP (N=445): Definitive delayed-start trial. Levodopa for 80 weeks (early-start) vs placebo × 40 weeks then levodopa × 40 weeks (delayed-start). No difference in UPDRS at 80 weeks (−1.0 vs −2.0, P=0.44)
• Conclusions: Levodopa does not modify disease progression — but it does not worsen it either, effectively putting the "levodopa toxicity" hypothesis to rest. There is no reason to delay levodopa initiation

α-Synuclein Immunotherapy

α-Synuclein is the pathological hallmark of PD — misfolded, aggregated α-synuclein forms Lewy bodies and Lewy neurites that propagate in a prion-like manner through connected brain regions. Immunotherapy aims to clear pathological α-synuclein aggregates or prevent their cell-to-cell spread.

Prasinezumab (Roche/Prothena)

Prasinezumab is a humanized monoclonal antibody that selectively binds aggregated α-synuclein, designed to reduce neuronal toxicity and intercellular spread. It is the most advanced anti-α-synuclein therapy in development.

• PASADENA (Phase 2, 2022): 316 early-stage PD patients (treatment-naïve or on MAO-Bi), randomized to placebo vs prasinezumab 1500 mg or 4500 mg IV Q4W for 52 weeks
  • Primary endpoint MISSED: MDS-UPDRS I+II+III change at 52 weeks — 1500 mg: −2.0 points vs placebo (P=0.24); 4500 mg: −0.6 (P=0.72)
  • Exploratory motor subanalyses showed numerical trends favoring prasinezumab on Part III (motor) and especially on bradykinesia subscores
  • 4-year OLE data (Nature Medicine 2024): prasinezumab-treated patients showed 51–65% slower motor decline vs PPMI external comparator cohort on MDS-UPDRS III — a sustained, growing signal
• PADOVA (Phase 2b, 2024–2025): 586 early-stage PD patients on stable symptomatic therapy (74% levodopa, 26% MAO-Bi), randomized to prasinezumab 1500 mg IV Q4W vs placebo for ≥76 weeks
  • Primary endpoint (time to confirmed ≥5-point MDS-UPDRS III worsening): HR 0.84 (95% CI 0.69–1.01; P=0.0657) — missed statistical significance but a clinically meaningful trend
  • Pre-specified levodopa-treated subgroup (75%): HR 0.79 (0.63–0.99; P=0.04 nominal)
  • Covariate-adjusted analysis: HR 0.81 (0.67–0.98) — nominally significant
  • Consistent positive trends across all secondary/exploratory endpoints
  • First biomarker evidence: Prasinezumab-treated participants showed less neuromelanin MRI signal loss in substantia nigra and significantly less iron accumulation — suggesting biological impact on the underlying pathology
  • Well tolerated; no new safety signals. Infusion reactions were the main AE
• Phase III announced June 2025 by Roche/Genentech — based on totality of PASADENA + PADOVA + OLE data. Over 900 patients treated, >500 for 1.5–5 years

Cinpanemab (Biogen) — SPARK (2022)

• SPARK (Phase 2): Anti-α-synuclein antibody targeting monomeric and aggregated forms. Terminated early for futility
• No clinical benefit at any dose at 52 or 72 weeks. No DaT-SPECT signal. No delayed-start effect
• The failure may reflect differences in epitope (cinpanemab binds a different α-synuclein domain than prasinezumab), binding characteristics, or dosing. It does not necessarily invalidate the α-synuclein target

Active α-Synuclein Immunization

• UB-312 (Vaxxinity): Synthetic peptide vaccine designed to induce endogenous antibodies against pathological α-synuclein. Phase 1 completed with acceptable safety; Phase 2 in development. Advantage: less frequent dosing than IV infusions
• Concept: shift from passive (exogenous antibody) to active immunization, enabling the patient's own immune system to continuously generate anti-α-synuclein antibodies

GLP-1 Receptor Agonists

Glucagon-like peptide-1 (GLP-1) receptor agonists — developed for type 2 diabetes — have shown neuroprotective effects in preclinical PD models, including anti-inflammatory, anti-apoptotic, and neurotrophic actions. GLP-1 receptors are expressed on neurons in the substantia nigra and striatum. Epidemiologic data have consistently shown lower PD incidence in diabetic patients treated with GLP-1 RAs.

Lixisenatide — LIXIPARK (2024)

• LIXIPARK (Phase 2, N=156): Early PD (diagnosed <3 yr, on stable dopaminergic therapy, no motor complications). Lixisenatide 20 μg SC daily vs placebo for 12 months, then 2-month washout
  • Positive primary endpoint: MDS-UPDRS III at 12 months: lixisenatide −0.04 (essentially stable) vs placebo +3.04 (worsened). Difference 3.08 points (95% CI 0.86–5.30; P=0.007)
  • After 2-month washout (off-medication): 3-point benefit persisted (17.7 vs 20.6) — though not adjusted for multiplicity
  • No difference in non-motor symptoms (MDS-UPDRS I+II), cognition, or quality of life
• Limitations: Small (N=156), short (12 months), single-country (France). The 3-point UPDRS difference, while statistically significant, is at the border of clinical significance. GI side effects were substantial (nausea 46%, vomiting 13%) — raising concern about unblinding. Washout period (2 months) may be insufficient to fully clear pharmacologic effects
• Larger, longer confirmatory trials needed. LIXIPARK-2 (Phase 3) is being planned

Exenatide — Exenatide-PD3 (2025)

• Exenatide-PD3 (Phase 3, N=194): Exenatide 2 mg SC weekly for 96 weeks in moderate PD (more advanced than LIXIPARK population)
  • Definitively negative: MDS-UPDRS III OFF: +5.7 (exenatide) vs +4.5 (placebo). Adjusted coefficient 0.92 (95% CI −1.56 to 3.39; P=0.47)
  • No benefit on any secondary outcome: motor ON, cognition (MoCA), quality of life, DaT-SPECT
  • CSF penetration was only ~1% of plasma levels — raising the critical question of whether exenatide adequately reached the CNS
• This follows a smaller, positive Phase 2 trial (Athauda et al., Lancet 2017) that had generated substantial excitement
• How to reconcile LIXIPARK (positive) vs Exenatide-PD3 (negative)? Potential explanations include different drugs (lixisenatide vs exenatide), different populations (early vs moderate PD), different CNS penetration, different study designs, or that LIXIPARK was a false positive driven by symptomatic effects/unblinding

Other GLP-1 RA Trials in Progress

• Semaglutide (multiple trials): The most widely prescribed GLP-1 RA globally. Higher potency and potentially better CNS penetration than first-generation agents. Phase 2 trials in PD are ongoing (including Novo Nordisk-sponsored studies)
• NLY01 (pegylated exenatide, Neuraly): Designed for improved BBB penetration. Phase 2 completed; results pending
• The field is keenly watching whether next-generation GLP-1 RAs with better CNS penetration will clarify the class's potential

Genetically Targeted Therapies

The identification of high-penetrance PD genes (LRRK2, GBA1, SNCA, PRKN/Parkin, PINK1) has opened the door to precision medicine approaches — therapies designed for specific genetic subtypes of PD.

LRRK2 Kinase Inhibitors

Gain-of-function mutations in LRRK2 are the most common genetic cause of PD (~1–2% of all PD, up to 40% in certain populations such as North African Arabs and Ashkenazi Jews). Increased LRRK2 kinase activity impairs lysosomal function, disrupts vesicular trafficking, and promotes neuroinflammation. Importantly, elevated LRRK2 kinase activity has also been observed in idiopathic (non-genetic) PD, broadening the potential therapeutic application.

• BIIB122/DNL151 (Biogen/Denali): Oral, CNS-penetrant, selective LRRK2 inhibitor. Phase 1/1b showed dose-dependent LRRK2 kinase inhibition (pS935-LRRK2 ≤98% reduction), good CNS penetration (CSF/plasma ratio ~1), and acceptable safety
  • LUMA trial (Phase 2, NCT05348785): ~650 patients with early-stage idiopathic PD (now also including LRRK2 mutation carriers after LIGHTHOUSE was folded in). BIIB122 225 mg daily vs placebo for 48–144 weeks. Primary endpoint: time to UPDRS worsening. Results expected 2025–2026
  • BEACON trial (Phase 2, 2024): 50 LRRK2-PD patients, 225 mg daily, 3-month primary safety endpoint + 2-year OLE. Completion expected 2028
• NEU-411 (Neuron23): Another oral LRRK2 inhibitor entering Phase 2 (NEULARK trial). Uses digital biomarkers as primary endpoint and genetic stratification to identify LRRK2-driven PD patients (including those without classical LRRK2 mutations but with elevated pathway activity)
• Safety considerations: LRRK2 is expressed in kidney and lung. Preclinical LRRK2 knockout models showed renal and pulmonary changes. Clinical-stage inhibitors at therapeutic doses have not shown significant organ toxicity so far, but long-term monitoring is essential

GBA1-Targeted Therapies

Heterozygous mutations in GBA1 (encoding glucocerebrosidase, GCase) are the most common genetic risk factor for PD (~5–10% of all PD patients carry at least one GBA1 variant). GBA1-PD tends to be more aggressive, with earlier cognitive decline, faster motor progression, and higher risk of dementia. Reduced GCase activity leads to lysosomal dysfunction, α-synuclein accumulation, and impaired autophagy.

• Ambroxol: An over-the-counter mucolytic that acts as a GCase chaperone, increasing GCase activity and facilitating its transport to lysosomes. The AIM-PD Phase 2 trial (N=17, open-label) showed ambroxol crossed the BBB, increased CSF GCase activity, and reduced CSF α-synuclein — proof of concept. Larger RCTs are planned/ongoing
• Venglustat (Sanofi): Substrate reduction therapy — a glucosylceramide synthase inhibitor that reduces the substrate accumulation downstream of GCase deficiency. Phase 2 MOVES-PD trial (2023) did not meet primary endpoint (MDS-UPDRS II+III: −0.84 vs placebo, P=0.67). Development paused
• PR001/LY3884961 (Prevail/Lilly): AAV9-based gene therapy delivering functional GBA1 gene directly to the CNS via intracisternal injection. Phase 1/2 (PROPEL study) in GBA1-PD is ongoing. First-in-class gene therapy approach for GBA-PD
• Small-molecule GCase activators: Multiple companies developing allosteric GCase activators that could benefit both GBA-PD and potentially idiopathic PD (where GCase activity is also reduced)

SNCA-Targeted Approaches

• Antisense oligonucleotides (ASOs): ION-464 (Ionis/Biogen) targets SNCA mRNA to reduce α-synuclein production at the source. Intrathecal delivery. Phase 1 in development
• Rationale: Rather than clearing already-misfolded α-synuclein (immunotherapy), reduce its production altogether — potentially more effective but carries risk of reducing physiological α-synuclein function (role in synaptic vesicle dynamics)

Other Emerging Therapeutic Strategies

Cell-Based Therapies

• Transplantation of dopaminergic progenitor cells derived from human embryonic stem cells (hESCs) or induced pluripotent stem cells (iPSCs) into the putamen to replace lost dopaminergic neurons
• Multiple Phase 1 trials ongoing globally (including Bayer's bemdaneprocel/BRT-DA01). Early safety data are encouraging but efficacy remains unproven
• Not truly "disease-modifying" (does not halt neurodegeneration) — rather, a restorative/replacement strategy

Neuroinflammation Targets

• NLRP3 inflammasome inhibitors: NLRP3 activation by α-synuclein aggregates drives neuroinflammation in PD. Dapansutrile (Phase 2 planned for 2025) and other inflammasome inhibitors entering clinical testing
• Rationale: Neuroinflammation is not just a bystander — microglial activation and cytokine-mediated toxicity actively contribute to dopaminergic neuron death. Targeting the inflammatory cascade could complement α-synuclein-directed therapy

Iron Chelation

• Iron accumulates in the substantia nigra in PD and promotes oxidative stress and α-synuclein aggregation
• Deferiprone: Phase 2 trials (FAIR PARK II) showed mixed results — no significant benefit on primary endpoint but some imaging signal of reduced nigral iron. Development continues at lower doses

Landscape Summary: Active Disease-Modification Programs

Trial Comparison Table: Completed Disease-Modification Trials