The MAGNOLIA trial is a Phase 2 clinical study evaluating LY-3372689 (zaniglusemab, oglemilide), an oral O-GlcNAcase (OGA) inhibitor developed by Eli Lilly, in patients with early Alzheimer's disease. This is one of the first large-scale clinical trials of an OGA inhibitor in AD.
| Parameter |
Value |
| Trial ID |
NCT05063539 |
| Phase |
Phase 2 |
| Status |
Completed |
| Drug |
LY-3372689 (zaniglusemab) |
| Company |
Eli Lilly |
| Indication |
Early Alzheimer's Disease (MCI due to AD, mild AD dementia) |
| Enrollment |
~200 patients |
| Duration |
52 weeks treatment |
| Primary Endpoints |
Safety, tolerability, biomarker changes |
| Secondary Endpoints |
Cognitive measures (ADAS-Cog14, ADCS-MCI-ADL) |
- Early AD patients (MCI due to AD or mild AD dementia)
- Confirmed amyloid pathology (PET or CSF)
- MMSE score 20-28
- Age 55-85 years
- LY-3372689: Multiple dose levels (oral)
- Placebo: Matching oral dosing
- Duration: 52 weeks
- Primary: Safety and tolerability
- Secondary: CSF O-GlcNAc levels, p-tau181, total tau, cognitive measures
LY-3372689 inhibits O-GlcNAcase (OGA), the enzyme that removes O-GlcNAc modifications from proteins including tau:
- Increased tau O-GlcNAcylation competes with pathological phosphorylation
- Reduced tau phosphorylation at key epitopes (Thr181, Thr231, Ser396)
- Decreased tau aggregation and neurofibrillary tangle formation
- CSF O-GlcNAc: Dose-dependent increase confirmed CNS target engagement
- Brain exposure: Drug levels consistent with BBB penetration
- CSF p-tau181: Modest reductions observed
- CSF total tau: Variable changes
- Did not meet primary cognitive endpoint in initial analysis
- Biomarker effects demonstrated mechanism but clinical translation challenging
The MAGNOLIA results highlight:
- Biomarker-to-outcome gap: Target engagement confirmed but cognitive benefit not demonstrated
- Possible explanations: Advanced pathology, insufficient duration, amyloid co-pathology
- Relevance to PSP: OGA inhibition may show clearer benefit in pure tauopathies (less amyloid)
| Factor |
MAGNOLIA (AD) |
PROSPER (PSP) |
| Pathology |
Amyloid + tau |
Pure tau (4R) |
| Complexity |
Multiple mechanisms |
Single target |
| Expected benefit |
Lower |
Higher |
O-GlcNAcylation is a post-translational modification where a single N-acetylglucosamine (GlcNAc) moiety is attached to serine or threonine residues on target proteins. Unlike complex glycosylation that occurs in the ER and Golgi, O-GlcNAcylation occurs in the cytoplasm and nucleus and is regulated by two enzymes:
- O-GlcNAc transferase (OGT) — adds the GlcNAc modification
- O-GlcNAcase (OGA) — removes the modification
This modification is analogous to phosphorylation in many ways—both involve the addition/removal of small molecular groups, and both can regulate protein function, localization, and interactions. Notably, O-GlcNAcylation and phosphorylation often compete for the same or adjacent serine/threonine residues, creating a dynamic "yin-yang" relationship between these modifications.
The brain has particularly high levels of O-GlcNAcylation, reflecting the high metabolic activity of neurons. Key brain proteins that are O-GlcNAcylated include:
- Tau protein — multiple sites including Thr181, Ser198, Ser199, Thr231, Ser396, Ser404
- Synapsin I — regulates synaptic vesicle trafficking
- CREB — transcription factor involved in memory formation
- NF-κB — transcription factor in neuroinflammation
- AMPA receptor subunits — regulate synaptic plasticity
The metabolic link is particularly important: O-GlcNAcylation depends on UDP-GlcNAc, the end product of the hexosamine biosynthetic pathway, which integrates glucose metabolism, amino acid metabolism, and fatty acid synthesis. This makes O-GlcNAcylation a metabolic sensor—a cell's nutritional status directly affects how proteins are modified.
In Alzheimer's disease and other tauopathies, tau protein becomes hyperphosphorylated, leading to:
- Abnormal aggregation into paired helical filaments (PHFs)
- Neurofibrillary tangle formation
- Progressive neuronal loss
- Cognitive decline
The hypothesis underlying OGA inhibition is elegant:
- Competition at modification sites — O-GlcNAcylation at the same Ser/Thr residues as phosphorylation blocks kinase access
- Conformational change — O-GlcNAcylated tau is less prone to aggregation
- Clearance enhancement — modified tau may be more readily cleared by cellular machinery
Preclinical studies in various animal models demonstrated that increasing tau O-GlcNAcylation through OGA inhibition reduced tau phosphorylation at multiple epitopes and decreased tau aggregation. These findings provided the rationale for advancing OGA inhibitors into clinical trials.
Early OGA inhibitors included:
- PUGNAc — first-generation OGA inhibitor, but poor BBB penetration
- Thiamet-G — better brain penetration, used extensively in preclinical studies
- NAG (N-acetylglutamine) — natural substrate with weak inhibitory activity
These compounds proved the concept in preclinical models but lacked the pharmaceutical properties needed for human trials.
LY-3372689 (zaniglusemab, previously known as oglemilide) represents a second-generation OGA inhibitor optimized for:
- Brain penetration — demonstrated CNS exposure in preclinical models
- Selectivity — high specificity for OGA versus other glycosidases
- Pharmacokinetics — suitable for once-daily oral dosing
- Safety profile — acceptable tolerability in Phase 1 studies
The compound was developed through a systematic medicinal chemistry campaign that balanced OGA inhibition potency with drug-like properties.
The MAGNOLIA trial enrolled patients with early Alzheimer's disease, defined as:
- MCI due to AD — mild cognitive impairment meeting AD criteria
- Mild AD dementia — MMSE 20-28, CDR 0.5-1.0
Key inclusion criteria:
- Age 55-85 years
- Amyloid confirmation — either:
- Positive amyloid PET scan (Centiloid ≥ 30)
- CSF biomarker profile (Aβ42/40 ratio, p-tau181)
- Stable medications — on cholinesterase inhibitors and/or memantine for ≥3 months
- Reliable caregiver — available for study participation
Patients were excluded for:
- Other neurodegenerative diseases (e.g., Lewy body dementia, frontotemporal dementia)
- Significant cerebrovascular disease on MRI
- Psychiatric conditions (major depression, psychosis)
- Uncontrolled medical conditions
- Prior participation in other AD clinical trials
The trial evaluated multiple dose levels:
- Low dose — selected based on Phase 1 data
- Mid dose — anticipated therapeutic dose
- High dose — maximum tolerated dose
Treatment was administered as oral tablets, once daily, for 52 weeks. This duration was selected based on:
- Expected time course for biomarker changes
- Adequate exposure to assess cognitive outcomes
- Balance with patient burden
-
Safety and Tolerability
- Adverse event monitoring throughout the study
- Vital signs, ECGs, laboratory assessments
- Cognitive and functional assessments
-
Target Engagement
- CSF O-GlcNAc levels (pharmacodynamic biomarker)
- Dose-exposure-response relationship
-
Tau Biomarkers
- CSF p-tau181 — core biomarker of tau pathology
- CSF p-tau217 — emerging biomarker
- CSF total tau — marker of neuronal injury
- Plasma p-tau181 and p-tau217
-
Amyloid Biomarkers
- CSF Aβ42/40 ratio
- Amyloid PET change from baseline
-
Neurodegeneration Markers
-
Cognitive Measures
- ADAS-Cog14 — primary cognitive endpoint
- ADCS-MCI-ADL — functional measure
- MMSE
- RBANS (Repeatable Battery for the Assessment of Neuropsychological Status)
LY-3372689 was generally well-tolerated:
- Most adverse events were mild to moderate
- No dose-limiting toxicities at the doses tested
- Low rates of discontinuation due to adverse events
The biomarker results provided clear evidence of mechanism engagement:
-
CSF O-GlcNAc Increase
- Dose-dependent increase in CSF O-GlcNAc levels
- Sustained elevation throughout the treatment period
- Confirmed CNS target engagement
-
Tau Biomarkers
- Modest reductions in CSF p-tau181 observed
- Variable changes in CSF total tau
- Results suggest effects on tau pathology, though magnitude limited
The trial did not meet its primary cognitive endpoint:
- No statistically significant difference in ADAS-Cog14 change from baseline
- Numerical trends favored LY-3372689 but did not reach statistical significance
- Interpretation: biomarker effects did not translate to clinical benefit in this population
Several factors may explain why demonstrated biomarker effects did not translate to cognitive benefit:
- Disease stage — Patients with established amyloid pathology may have already experienced significant neuronal loss
- Amyloid co-pathology — The presence of amyloid may continue to drive neurodegeneration even if tau is modulated
- Compensatory mechanisms — Advanced disease may overwhelm any single-mechanism intervention
- 52 weeks may be insufficient — Neurodegeneration is a decades-long process; a year of treatment may be too short
- Delayed effects — Biomarker changes may precede clinical effects by longer intervals
- Incomplete target coverage — Degree of O-GlcNAcylation achieved may be insufficient
- Alternative pathways — Other tau phosphorylation pathways may compensate
- Aggregation vs. modification — May not address already-formed aggregates
The MAGNOLIA results illustrate the AT(N) framework in AD:
- A (Amyloid) — Amyloid positivity required for enrollment
- T (Tau) — Tau pathology modulated by treatment
- (N) (Neurodegeneration) — Neuronal loss continues despite target engagement
This suggests that interventions may need to:
- Target amyloid earlier in disease course
- Use combination therapies addressing multiple pathways
- Achieve greater target engagement
The PROSPER trial (NCT04564555) evaluated the same compound in PSP, a pure 4R-tauopathy without amyloid pathology. This provides an important comparison:
| Aspect |
MAGNOLIA (AD) |
PROSPER (PSP) |
| Pathology |
Amyloid + tau |
Tau only (4R) |
| Comorbidities |
Multiple |
Single mechanism |
| Expected benefit |
Lower |
Potentially higher |
If PSP patients show clearer benefit, it would suggest:
- Amyloid pathology drives resistance to single-mechanism tau therapies
- Pure tauopathies may be more amenable to tau-directed approaches
- Combination with anti-amyloid therapy may be needed in AD
Based on learnings from MAGNOLIA:
- Anti-amyloid + OGA inhibitor — Combine with lecanemab or donanemab
- Multiple tau mechanisms — OGA inhibition + kinase inhibitors + aggregation inhibitors
- Earlier intervention — Treat at preclinical or prodromal stages
Pharmaceutical companies are developing:
- More potent OGA inhibitors
- Compounds with different pharmacokinetic profiles
- Agents targeting OGT (increase O-GlcNAcylation directly)
Better biomarkers are needed:
- Tau PET ligands specific for O-GlcNAcylated tau
- CSF markers predictive of response
- Blood-based biomarkers for screening
- Anti-amyloid monoclonal antibodies — Lecanemab, donanemab, crenezumab
- Tau aggregation inhibitors — Methylthioninium chloride, others
- Tau kinase inhibitors — GSK3β, CDK5 inhibitors
- Immunotherapy — AAT-176 (anti-tau antibody)
| Compound |
Company |
Status |
Indication |
| LY-3372689 |
Eli Lilly |
Completed |
AD, PSP |
| EIDD-6801 |
Merck/MSD |
Phase 2 |
AD |
| BIIB113 |
Biogen |
Phase 1 |
AD |
The OGA inhibition approach represents a novel therapeutic strategy based on the following mechanistic chain:
flowchart TD
A"OGA Inhibition" --> B"Increased Tau O-GlcNAcylation"
B --> C"Competition with Phosphorylation"
C --> D"Reduced Tau Phosphorylation at Key Epitopes"
D --> E"Decreased Tau Aggregation"
E --> F"Reduced Neurofibrillary Tangle Formation"
F --> G"Slowing of Neurodegeneration"
G --> H"Preservation of Cognitive Function"
This mechanistic hypothesis was well-supported by preclinical data. The clinical translation challenge underscores the complexity of neurodegenerative disease and the difficulty of achieving clinically meaningful benefit with single-mechanism approaches in patients with established pathology.
The MAGNOLIA trial has completed. Results have been published informing future development strategy for OGA inhibitors. Key learnings have influenced:
- Trial design for future tau-directed therapies
- Patient selection criteria
- Combination therapy approaches
- Earlier intervention strategies
The field continues to evolve, with ongoing research into OGA inhibition and other tau-targeted approaches.