**Bottom Line**
Most feasible near-term development paths are not “one drug for AD/PD/ALS/FTD.” The tractable version is mechanism-stratified development: use cross-disease biology to nominate biomarkers and patient subsets, then run disease-specific or genetically enriched trials.
My feasibility ranking:
1. **Autophagy-lysosomal dysfunction**: best broad platform, but hard CNS druggability.
2. **Mitochondrial/NAD+ quality control**: easiest early clinical path, weakest specificity.
3. **TDP-43 proteinopathy**: strong for ALS/FTD, weak as pan-NDD therapy.
4. **Retromer/endosomal trafficking**: attractive genetically, still preclinical.
5. **Microglial inflammation/NLRP3/TREM2**: biologically real, clinically high-risk after TREM2 setback.
6. **RNA/NCT defects**: compelling for C9orf72 ALS/FTD, speculative for AD/PD.
7. **Brain insulin resistance**: lowest priority as a cross-disease drug program unless reframed as metabolic biomarker stratification.
**1. Autophagy-Lysosomal Dysfunction**
Feasibility: **moderate-high as a platform; moderate as a drug program.**
Druggability is plausible but not simple. TFEB activation is attractive, but direct systemic TFEB activation risks broad effects on metabolism, immunity, lysosomal expansion, and possibly tumor biology. Better druggable entry points are **GBA1 enhancement, TMEM175 modulation, retromer stabilization, mTOR-independent lysosomal activation**, or brain-targeted gene therapy in rare genetic subsets.
Best biomarkers: CSF/plasma lysosomal enzymes, GCase activity, cathepsins, LAMP1/2, LC3-II/p62 in cells, PET if lysosomal tracers mature, plus proteinopathy-specific readouts: tau PET, α-syn SAA, NfL. iPSC neurons/macrophage-microglia co-cultures and aged knock-in models are preferable to aggressive overexpression mice.
Clinical constraint: broad sporadic AD/PD/ALS trials would be expensive and underpowered unless enriched by **GBA1, TMEM175, VPS35, lysosomal biomarker-low** patients. A first proof-of-biology trial in GBA1-PD or prodromal GBA1 carriers is more realistic.
Safety: chronic lysosomal upregulation could disturb immune function, lipid handling, and neuronal homeostasis. AAV-TFEB is not first-line clinically because dosing, reversibility, and long-term CNS safety are major barriers.
Timeline/cost: biomarker-enriched Phase 1b/2a small molecule: **4-6 years, $40-90M**. Gene therapy route: **7-10+ years, $150-300M+**.
**2. TDP-43 Proteinopathy**
Feasibility: **high for ALS/FTD precision development; low as pan-AD/PD therapy.**
The strongest investable version is not “TDP-43 across all neurodegeneration,” but **TDP-43-positive ALS/FTD and LATE/AD subgroups**. TDP-43 is hard to drug directly, but ASOs, RNA/splicing rescue, aggregation blockers, nuclear localization stabilizers, and stress-granule modulators are plausible. The tofersen precedent matters: FDA accelerated approval for SOD1-ALS was based on NfL reduction, showing that genetically defined ALS can use biomarker-driven development ([Biogen/FDA release](https://investors.biogen.com/news-releases/news-release-details/fda-grants-accelerated-approval-qalsodytm-tofersen-sod1-als)).
Best biomarkers: TDP-43 seed amplification assays are becoming more credible; a 2025 study detected CSF TDP-43 seeding in symptomatic and some presymptomatic genetic FTD/ALS carriers, with reported 67% sensitivity in TDP-43-linked symptomatic patients and 93% specificity ([PubMed](https://pubmed.ncbi.nlm.nih.gov/41399249/)). Also use cryptic-exon/splicing biomarkers, NfL, pNfH, MRI atrophy, EMG/ALSFRS-R for ALS, and CDR/FTD scales for dementia.
Model systems: C9orf72, GRN, TARDBP, UBQLN2 iPSC neurons and assembloids are useful; overexpression TDP-43 mice are often toxic artifacts. Human pathology-anchored models are essential.
Clinical constraint: AD/PD inclusion only makes sense if patients are biomarker-positive for TDP-43. Otherwise effect dilution is fatal.
Safety: lowering or altering TDP-43 is dangerous because nuclear TDP-43 has essential RNA-processing functions. Therapeutics must avoid excessive knockdown.
Timeline/cost: ALS/FTD biomarker-positive Phase 1b/2a: **3-5 years, $50-120M**. AD/LATE subgroup program: **6-9 years, $150-400M**, mostly due to diagnosis and endpoint burden.
**3. Microglial Neuroinflammation / NLRP3 / TREM2**
Feasibility: **moderate biologically, low-moderate clinically.**
This is real disease biology, but the clinical translation risk is high. The strongest caution is TREM2: AL002 showed target engagement and microglial pharmacodynamics but failed Phase 2 clinical and biomarker efficacy in early AD ([Alector 2024 results](https://investors.alector.com/news-releases/news-release-details/alector-announces-results-al002-invoke-2-phase-2-trial/)); the randomized trial has since been published in 2026 ([Nature Medicine](https://www.nature.com/articles/s41591-026-04273-1)). That does not kill microglial biology, but it weakens simple “activate TREM2” development.
Druggability: NLRP3 is druggable chemically; TREM2 is antibody-druggable; complement is druggable. But CNS exposure, cell-state specificity, and timing are the core problems. MCC950 is useful experimentally, not a clean development candidate.
Best biomarkers: TSPO PET is noisy; better panels include sTREM2, YKL-40, GFAP, IL-1β/IL-18 where measurable, complement fragments, ASC specks, snRNA-seq state signatures, amyloid/tau/α-syn/TDP-43 disease markers, and NfL.
Model systems: human iPSC microglia-neuron-astrocyte co-cultures, xenotransplanted human microglia mice, aged knock-in disease models. Standard young transgenic mice overpredict efficacy.
Clinical constraint: likely needs disease-stage selection. Early inflammation may be protective; late inflammation may be harmful. A flat inhibitor/agonist across all stages is risky.
Safety: immunosuppression, infection risk, impaired debris clearance, amyloid-related imaging changes for some immune-activating approaches, systemic inflammasome liabilities.
Timeline/cost: repurposed/known CNS-penetrant anti-inflammatory biomarker trial: **3-5 years, $30-80M**. Novel CNS microglial drug to Phase 2: **6-9 years, $120-300M**.
**4. RNA Metabolism / Nucleocytoplasmic Transport**
Feasibility: **high in C9orf72 ALS/FTD; low as cross-disease AD/PD program.**
Druggability is strongest through **ASOs, RNA-targeting small molecules, DPR-lowering strategies, nuclear import/export modifiers**, and stress granule biology. For AD/PD, the biology is currently too nonspecific.
Best biomarkers: DPR proteins in CSF for C9orf72, poly(GP), NfL, cryptic exon markers, nuclear/cytoplasmic RanGAP1/NUP localization in patient cells, RNA-seq splicing signatures. For AD/PD there is no validated clinical NCT biomarker.
Model systems: C9orf72 iPSC motor neurons/cortical neurons and organoids are appropriate. For AD/PD, use models only as secondary validation after showing patient-cell nuclear transport defects.
Clinical constraint: feasible trial population is **C9orf72 carriers**, including presymptomatic or early symptomatic cohorts. Cross-disease trials are premature.
Safety: broad nuclear transport modulation is high-risk because it touches essential cell biology. ASOs against specific toxic transcripts are safer conceptually.
Timeline/cost: C9orf72-focused ASO/small molecule program: **4-7 years, $80-200M**. Pan-NDD NCT program: not trial-ready.
**5. Mitochondrial Quality Control / NAD+**
Feasibility: **moderate-high for early trials; low as disease-modifying claim.**
This is the easiest to test clinically because NAD+ boosters and metabolic interventions have tolerability precedent. The NADPARK Phase 1 PD trial reported oral nicotinamide riboside increased brain NAD and was safe over 30 days ([Cell Metabolism](https://www.sciencedirect.com/science/article/pii/S1550413122000456)). But “mitochondrial dysfunction” is broad aging biology, so a positive biomarker effect may not translate into slowed neurodegeneration.
Druggability: NAD+ supplementation, AMPK/SIRT modulation, mitophagy enhancers, PINK1/Parkin activators, mitochondrial-targeted antioxidants. PINK1/Parkin is genetically strong in PD but not cross-disease.
Best biomarkers: 31P-MRS brain NAD, CSF/plasma NAD metabolites, lactate/pyruvate, acylcarnitines, mitochondrial DNA damage, NfL, disease-specific progression markers. Need target engagement plus neurodegeneration markers.
Model systems: mito-QC mice are useful for flux, but human iPSC neurons with stress paradigms are better for translatability. Avoid relying on toxin models alone.
Clinical constraint: good for Phase 2 biomarker trials in PD or mild AD, but pivotal disease-modification trials would be large and expensive unless enriched by mitochondrial biomarker deficits.
Safety: NR/NMN generally manageable, but long-term effects on cancer biology, methyl donor balance, liver metabolism, and immune state need monitoring. Strong mitophagy activators could harm high-energy neurons if overdosed.
Timeline/cost: nutraceutical-style biomarker Phase 2: **2-4 years, $10-40M**. Novel mitophagy drug: **5-8 years, $80-200M**.
**6. Retromer / Endosomal Trafficking**
Feasibility: **moderate; attractive but not trial-ready.**
This is one of the cleaner mechanistic intersections between AD and PD, especially SORL1/VPS35 biology. It is less convincing for ALS/FTD except through broader endolysosomal stress.
Druggability: retromer stabilization is chemically plausible, but there is no mature clinical precedent. VPS35 gene therapy or overexpression is too early and safety-sensitive. Small-molecule chaperones or cargo-specific trafficking correctors are more realistic.
Best biomarkers: endosomal morphology in patient neurons, CI-MPR trafficking, APP processing/Aβ ratios, α-syn secretion/uptake, lysosomal enzyme trafficking, SORL1/VPS35 genotype, CSF Aβ/tau, α-syn SAA in PD.
Model systems: SORL1 loss-of-function iPSC neurons for AD; VPS35 D620N knock-in dopaminergic neurons for PD; microfluidic α-syn propagation assays. NHP α-syn PFF work is expensive and should wait until a molecule has strong rodent/human-cell data.
Clinical constraint: start with genetically enriched AD/PD subsets, not ALS/FTD. A pan-NDD indication would be unjustified.
Safety: vesicle trafficking is fundamental; chronic perturbation may affect synapses, lysosomal enzymes, immune cells, and peripheral organs.
Timeline/cost: discovery-to-IND: **3-5 years, $30-80M**. First Phase 2 signal: **6-9 years, $120-250M**.
**7. Brain Insulin Resistance / Metabolic Dysregulation**
Feasibility: **low as a unifying therapeutic hypothesis; moderate as adjunctive stratification.**
Druggability is easy: intranasal insulin, GLP-1 agonists, metformin-like AMPK modulation, IGF-1 axis drugs. The problem is specificity and endpoint clarity. ALS IGF-1 history is not encouraging, and AD/PD metabolic associations are heavily confounded by age, vascular disease, and systemic diabetes.
Best biomarkers: FDG-PET, insulin/IRS1 phospho-signaling in extracellular vesicles if analytically validated, HbA1c/insulin resistance, inflammatory-metabolic panels, cognition/motor endpoints by disease.
Clinical constraint: GLP-1 and metabolic drugs should be tested in metabolically enriched subgroups, not as generic AD/PD/ALS/FTD modifiers. ALS is especially risky because weight loss and hypermetabolism complicate metabolic intervention.
Safety: hypoglycemia for insulin approaches, weight loss/GI effects for GLP-1 agonists, frailty concerns in advanced NDD.
Timeline/cost: repurposed metabolic drug Phase 2: **2-4 years, $20-60M**. New CNS metabolic drug: **5-8 years, $100M+**.
**Recommended Development Strategy**
Prioritize three parallel tracks:
1. **Biomarker-first cross-disease observational study**: AD, PD, ALS, FTD, controls; collect CSF/plasma, genetics, α-syn/TDP-43 SAA, NfL, GFAP, lysosomal enzymes, NAD metabolomics, and imaging where feasible. Cost: **$8-20M**, 2-3 years.
2. **Precision therapeutic pilots**:
- GBA1/TMEM175/lysosomal-low PD for autophagy-lysosomal drugs.
- C9orf72/GRN/TDP-43-positive ALS/FTD for TDP-43/RNA biology.
- NAD-deficient PD or early AD for mitochondrial/NAD intervention.
3. **Deprioritize pan-disease efficacy trials** until biomarkers prove that the same mechanism is active before degeneration, not merely after injury.
The investable thesis is not that all four diseases share one cause. It is that several stress-response systems recur across diseases, and the winning programs will use those systems to define **treatable molecular subtypes**.