# Compound Dive: Approved and Late-Stage Clinical Compounds Targeting TDP-43 Proteostasis, Autophagy, or Stress Granule Dynamics in ALS

**Date:** 2026-04-30
**Scope:** Approved therapeutics and compounds in late-stage clinical development that modulate TDP-43 proteostasis, autophagic flux, or stress granule dynamics, with relevance to amyotrophic lateral sclerosis (ALS).

---

## Executive Summary

TDP-43 proteinopathy is present in ~97% of ALS cases, making clearance of pathological TDP-43, restoration of autophagic flux, and dissolution of persistent stress granules (SGs) central therapeutic targets. This review identifies nine approved or late-stage clinical compounds with documented activity against these mechanisms. Of these, **edaravone** (already FDA-approved for ALS) and **nilotinib** (FDA-approved for chronic myeloid leukemia) stand out as having the most promising translational profiles, supported by mechanistic evidence in TDP-43 models and established safety databases. **Colchicine** shows encouraging phase 2 biological signals but needs confirmatory efficacy data. Several high-profile investigational agents—notably trehalose, arimoclomol, AMX0035 (Relyvrio), and the ISR inhibitors DNL343/fosigotifator—have reported negative or inconclusive phase 2/3 results in ALS as of 2024–2025, tempering enthusiasm for their immediate repurposing.

---

## 1. Compounds with Regulatory Approval (Any Indication)

### 1.1 Edaravone (Radicava / Radicava ORS)

| Attribute | Detail |
|-----------|--------|
| **Approval** | FDA-approved for ALS (2017, IV; 2022, oral) |
| **Primary indication** | Amyotrophic lateral sclerosis |
| **MW / LogP** | 174.2 Da / 1.3 (CID 4021) |
| **Class** | Free-radical scavenger (pyrazolin-5-one) |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
Edaravone was originally developed as an antioxidant/free-radical scavenger, but recent mechanistic studies demonstrate neuroprotection through TDP-43 proteostasis pathways. In TDP-43-overexpressing neuronal cells subject to oxidative stress, edaravone (≥10 µM) protects against neurotoxicity in a concentration-dependent manner. RNA sequencing revealed edaravone modulates multiple pathways directly implicated in TDP-43 proteinopathy, including:
- **Nrf2-mediated oxidative stress response**
- **Unfolded protein response (UPR)**
- **Autophagy pathways** (Soejima-Kusunoki et al., *Pharmaceuticals* 2022)

A 2025 study further showed edaravone mitigates TDP-43 mislocalization in human ALS iPSC-derived neurons, implicating the **SIRT1–XBP1 proteostasis axis** (Mikuriya et al., 2025). This suggests edaravone's neuroprotective effect in ALS extends beyond simple antioxidation and involves direct enhancement of TDP-43 handling and clearance.

**Published ALS evidence**
- Approved in Japan (2015) and the US (2017) based on phase 3 data showing slowed functional decline in a subset of early-stage ALS patients.
- Post-hoc and mechanistic studies now link its benefit to TDP-43 proteostasis normalization.

**Translational assessment**
Edaravone is the only approved ALS drug with published evidence for direct TDP-43 proteostasis enhancement. Its blood–brain barrier penetration and established safety profile in ALS are major advantages. However, clinical effect sizes are modest; identifying biomarkers of TDP-43 pathway modulation could help enrich responders.

---

### 1.2 Sodium Phenylbutyrate + Taurursodiol (AMX0035 / Relyvrio / ALBRIOZA)

| Attribute | Detail |
|-----------|--------|
| **Approval** | FDA-approved Sept 2022; **voluntarily withdrawn April 2024** |
| **Primary indication** | ALS (withdrawn) |
| **Class** | Chemical chaperone + mitochondrial protectant (fixed-dose combination) |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
Sodium phenylbutyrate (PB) acts as a chemical chaperone that promotes proteostasis by upregulating heat shock proteins and facilitating protein folding. Taurursodiol (TURSO; TUDCA) is a bile acid derivative that reduces endoplasmic reticulum stress and stabilizes mitochondrial membranes. Together, the combination targets **ER stress and protein misfolding**—processes upstream of TDP-43 aggregation. In ALS fibroblasts, AMX0035 altered transcriptional and metabolic landscapes consistent with proteostasis restoration.

**Published ALS evidence**
- **CENTAUR phase 2 (n=137):** Met primary endpoint with ALSFRS-R slope difference of 0.42 points/month (p=0.03); open-label extension suggested median survival benefit of ~4.8–6.5 months over placebo.
- **PHOENIX phase 3 (n=664):** Announced March 2024; **failed to meet primary or secondary endpoints**. Amylyx voluntarily withdrew marketing authorization.

**Translational assessment**
The CENTAUR post-hoc survival signal generated substantial hope, but the definitive phase 3 failure raises fundamental questions about whether PB/TURSO's proteostatic mechanisms are sufficiently potent or specific to TDP-43 pathology to produce clinically meaningful benefit in a broad ALS population. Its withdrawal removes it from frontline consideration for immediate repurposing.

---

### 1.3 Nilotinib

| Attribute | Detail |
|-----------|--------|
| **Approval** | FDA-approved Oct 2007 (CML) |
| **Primary indication** | Philadelphia chromosome–positive chronic myeloid leukemia (CML) |
| **MW / LogP** | 529.5 Da / 4.9 (CID 644241) |
| **Class** | c-Abl/BCR-ABL tyrosine kinase inhibitor |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
Activated c-Abl kinase negatively regulates autophagy and lysosomal function. Nilotinib inhibits c-Abl, which:
- Restores **autophagic flux** (increases LC3-II, reduces p62 accumulation)
- Promotes **lysosomal clearance** of misfolded proteins
- Reduces neuroinflammation and oxidative stress
- Crosses the **blood–brain barrier** (more effectively than first-generation imatinib)

In AD and PD models, nilotinib restored autophagy and reduced accumulation of toxic protein aggregates, including α-synuclein and tau. While direct TDP-43 aggregation studies in ALS models are limited, the conserved mechanism of c-Abl-mediated autophagy suppression and the compound's ability to penetrate the CNS provide a plausible rationale for TDP-43 proteinopathy. c-Abl activation has also been documented in ALS postmortem tissue.

**Published ALS/neurodegeneration evidence**
- AD mouse models: Nilotinib restored autophagy, prevented dopaminergic neuron degeneration, improved cognitive function (La Barbera et al., *Prog Neurobiol* 2021; Nobili et al., *Autophagy* 2021).
- PD: Phase 1/2 trials demonstrated safety; CSF drug levels reached pharmacologically relevant concentrations (Alzforum nilotinib profile).
- Phase 3 trials are ongoing in AD and other neurodegenerative indications.

**Translational assessment**
Nilotinib's **BBB penetration, established safety at low doses, and broad autophagy-restoring mechanism** make it an attractive candidate for ALS repurposing. However, cardiovascular toxicity (QT prolongation) at oncology doses requires careful dose-finding in an ALS population. Direct TDP-43 aggregation data in ALS-relevant models (iPSC-derived motor neurons, patient fibroblasts) remain sparse and would strengthen the case.

---

### 1.4 Colchicine

| Attribute | Detail |
|-----------|--------|
| **Approval** | FDA-approved for gout (Colcrys), familial Mediterranean fever |
| **Primary indication** | Gout, FMF, pericarditis (anti-inflammatory) |
| **MW / LogP** | 399.4 Da / 1.0 (CID 6167) |
| **Class** | Microtubule disruptor / anti-inflammatory alkaloid |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
Colchicine induces expression of **heat shock protein B8 (HSPB8)** and multiple autophagy players. HSPB8 recognizes misfolded mutant SOD1 and TDP-43 fragments, promoting their autophagy-mediated clearance. The **HSPB8–BAG3–HSP70 complex maintains "granulostasis"**—preventing conversion of dynamic stress granules into aggregation-prone assemblies. Preclinical data show colchicine blocks TDP-43 accumulation in motor neuron models.

**Published ALS evidence**
- **Co-ALS phase 2 trial (n=54):** Randomized, double-blind, placebo-controlled. While positive responder rates (ALSFRS-R decline <4 points) did not reach significance, **monthly ALSFRS-R decline was significantly slower in the low-dose (0.005 mg/kg/day) colchicine arm versus placebo**. Secondary biological outcomes assessed autophagy, SG responses, TDP-43 accumulation, neurofilament levels, and extracellular vesicle secretion. Results published in *Brain Communications* (2024).

**Translational assessment**
Colchicine has a **strong, specific mechanistic rationale** directly linking HSPB8 upregulation to TDP-43 and stress granule clearance. It is inexpensive and widely available. The phase 2 ALS data showed a **tantalizing dose-dependent signal on functional decline** but were underpowered. The main limitations are its narrow therapeutic index (risk of myelosuppression and GI toxicity) and the need for biomarker-confirmed target engagement in human motor neurons.

---

### 1.5 Guanabenz (Wytensin; withdrawn in US)

| Attribute | Detail |
|-----------|--------|
| **Approval** | Previously FDA-approved for hypertension (withdrawn from US market) |
| **Primary indication** | Hypertension (historical) |
| **MW / LogP** | 231.1 Da / 1.7 (CID 5702063) |
| **Class** | α-2 adrenergic agonist; PPP1R15A/PP1c phosphatase complex inhibitor |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
Guanabenz selectively inhibits endoplasmic reticulum stress-induced dephosphorylation of **eIF2α**, a key node in the **integrated stress response (ISR)**. By sustaining phospho-eIF2α, guanabenz triggers selective translational reprogramming that:
- Reduces misfolded protein accumulation and ER overload
- Modulates **stress granule assembly/dynamics**
- Protects motor neurons in yeast and mammalian ALS models

**Published ALS evidence**
- **ProMISe phase 2 trial:** Assessed guanabenz in ALS patients using a futility design. Published in *Brain* (2021). The trial did not demonstrate sufficient signal to proceed to phase 3, though the approach informed subsequent development of next-generation analogues (IFB-088 at InFlectis BioScience).

**Translational assessment**
Guanabenz's mechanism is directly on the ISR–stress granule axis, a validated ALS pathway. The phase 2 results were disappointing, and the drug's withdrawal from the US market complicates immediate repurposing. However, **structurally optimized analogues with better CNS penetration and selectivity** may warrant ALS evaluation.

---

### 1.6 Rapamycin (Sirolimus)

| Attribute | Detail |
|-----------|--------|
| **Approval** | FDA-approved (immunosuppressant, anti-proliferative agent) |
| **Primary indication** | Organ transplant rejection, lymphangioleiomyomatosis |
| **Class** | mTOR inhibitor |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
Rapamycin inhibits mTORC1, a negative regulator of autophagy. It enhances **macroautophagy**, increasing LC3-II and promoting clearance of aggregated proteins. In TDP-43 transgenic mice, genetic enhancement of autophagy rescues motor neuron survival.

**Published ALS evidence**
- **RAP-ALS phase 2 trial:** Published in *Nature Communications* (2023). Rapamycin was found to be **safe and well tolerated** in ALS patients, with no unexpected adverse events. However, the trial was not powered to demonstrate efficacy.
- **Preclinical caveat:** In SOD1^G93A mice, paradoxically, rapamycin **accelerated motor neuron degeneration and shortened lifespan** (Zhang et al., *Autophagy* 2011), suggesting mTOR inhibition may have model-dependent or mutation-specific effects that complicate broad ALS application.

**Translational assessment**
Rapamycin's autophagy-enhancing mechanism is well validated in neurodegeneration broadly, but the SOD1 mouse toxicity and lack of ALS efficacy data make it a risky candidate for TDP-43-targeted repurposing without further biomarker-driven stratification.

---

## 2. Late-Stage Investigational Compounds (Phase 2/3 in ALS or Advanced Neurodegeneration)

### 2.1 Trehalose (SLS-005)

| Attribute | Detail |
|-----------|--------|
| **Stage** | Phase 2/3 completed (HEALEY ALS Platform Trial) |
| **Primary indication tested** | ALS |
| **MW / LogP** | 342.3 Da / –4.2 (CID 7427) |
| **Class** | Non-reducing disaccharide; mTOR-independent autophagy inducer |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
Trehalose is a potent, **mTOR-independent inducer of autophagy** that activates the transcription factor **TFEB**, master regulator of the CLEAR (Coordinated Lysosomal Expression and Regulation) network. This drives lysosomal biogenesis and enhances clearance of aggregated proteins. In a cell culture model of TDP-43 proteinopathy, trehalose significantly reduced TDP-43 accumulation via TFEB activation and autophagic degradation (Wang et al., *Neurotoxicity Res* 2018).

**Published ALS evidence**
- **HEALEY ALS Platform Trial (n=161 treated; 205 placebo):** Published in *Lancet Neurology* (2025). Trehalose (0.75 g/kg IV weekly for 24 weeks) was well tolerated but **did not show a difference in ALS disease progression** compared with placebo (disease rate ratio 0.87; 95% CrI 0.665–1.102). No benefit in secondary clinical or biomarker measures.
- Several animal studies across multiple neurodegenerative models supported trehalose's autophagy activation, but heterogeneous methodologies and bioavailability questions limited translation.

**Translational assessment**
The HEALEY trial failure is a significant setback. Questions remain about whether IV dosing achieved adequate CNS exposure, whether timing of intervention was optimal, and whether autophagy induction alone is sufficient in the face of concurrent proteasome dysfunction in ALS. Repurposing is not justified without a stronger biomarker signal linking CNS exposure to target engagement.

---

### 2.2 Arimoclomol (MIPLYFFA)

| Attribute | Detail |
|-----------|--------|
| **Stage** | Phase 3 completed (ORARIALS-01) |
| **Primary indication tested** | ALS; also Niemann-Pick disease type C (approved Sept 2024 for NPC) |
| **Class** | Hydroxylamine derivative; heat shock protein (HSP) co-inducer |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
Arimoclomol is a **co-inducer of heat shock protein expression** that amplifies the stress response only under conditions of cellular stress. Beyond direct chaperoning, it activates **TFEB and TFE3**, upregulating lysosomal CLEAR genes and promoting autophagic clearance. In SOD1 ALS mice, arimoclomol rescued motor neurons, improved neuromuscular function, and extended lifespan. In human TDP-43 ALS models, however, arimoclomol showed variable effects—preserving nuclear TDP-43 in some contexts but failing to induce HSPA1A/HSPA8 in TDP-43 and FUS models (Fernández Comaduran et al., *Cell Stress & Chaperones* 2024).

**Published ALS evidence**
- **ORARIALS-01 phase 3 trial (n=231):** Published in *Lancet Neurology* (May 2024). Arimoclomol **did not improve clinical outcomes** relative to placebo in early ALS. The trial failed to meet primary and key secondary endpoints.
- Approved by FDA in Sept 2024 for Niemann-Pick disease type C, providing a post-ALS regulatory path.

**Translational assessment**
The phase 3 failure in ALS effectively removes arimoclomol from immediate ALS repurposing consideration. Its mechanism appears less effective in TDP-43–driven models than in SOD1 models, suggesting **genetic/etiological heterogeneity matters for HSP-based strategies**.

---

### 2.3 DNL343 / Fosigotifator (eIF2B activators)

| Attribute | Detail |
|-----------|--------|
| **Stage** | Phase 2/3 completed (HEALEY ALS Platform Trial); both failed |
| **Primary indication tested** | ALS |
| **Class** | Integrated stress response inhibitors (ISR) — eIF2B allosteric activators |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
Both compounds allosterically activate eIF2B to counteract phospho-eIF2α–mediated translational repression, thereby rescuing cap-dependent protein synthesis and theoretically preventing formation of persistent **stress granules**—which are hypothesized to seed TDP-43 aggregation in ALS.

**Published ALS evidence**
- Preclinical cell models: ISRIB enhanced survival of SOD1^G93A–expressing neurons (Marlin et al., 2024).
- However, in SOD1^G93A mice, two structurally distinct eIF2B activators (**2BAct and PRXS571**) **accelerated muscle denervation, anticipated disease onset, and shortened survival**—revealing ISR as a neuroprotective pathway in motor neurons and demonstrating toxicity of ISR inhibition in this model (Marlin et al., *Br J Pharmacol* 2024).
- **HEALEY trial (early 2025):** Both DNL343 (Denali) and fosigotifator (AbbVie/Calico) **failed primary endpoints** of function and survival.

**Translational assessment**
The ISR inhibitor approach for ALS suffered a dual blow: negative HEALEY trial results and alarming preclinical toxicity data in the SOD1 mouse model. Unless biomarker stratification identifies a specific ALS subtype where ISR overactivation is clearly pathological (vs. protective), repurposing trials are **not currently justified**.

---

### 2.4 NU-9 (Preclinical / Pre-IND)

| Attribute | Detail |
|-----------|--------|
| **Stage** | Preclinical (Northwestern University / pending clinical development) |
| **Primary indication tested** | ALS (preclinical) |
| **Class** | Cyclohexane-1,3-dione derivative; aggregation inhibitor |

**Mechanism relevant to TDP-43, autophagy, or stress granules**
NU-9 was identified for its ability to reduce misfolded SOD1 aggregation, but mechanistic studies reveal it also **improves mitochondrial and ER stability** and stimulates **autophagosome formation**. In both hSOD1^G93A and hTDP-43^A315T mouse models, NU-9 improved upper motor neuron health (mitochondria, ER, dendrites, axons) and preserved grip strength. Mechanistically, it targets common pathogenic mechanisms of protein aggregation shared across SOD1 and TDP-43 ALS subtypes.

**Published ALS evidence**
- Genç et al., *Clin Transl Med* 2021; follow-up PNAS 2025 study extended mechanism to amyloid-β oligomers, suggesting a unifying anti-aggregation mechanism across proteinopathies.
- Derivatives with improved BBB permeability and microsomal stability have been synthesized (Elmansy et al., *Bioorg Chem* 2025).

**Translational assessment**
NU-9 is the most mechanistically compelling preclinical compound for TDP-43 aggregation, but it is **not yet in clinical trials**. Progression to IND-enabling studies and phase 1 is the critical next step.

---

## 3. Top Three Candidates by Translational Promise

### 🥇 1. Edaravone (Radicava ORS)
**Why:** Already FDA-approved for ALS; mounting mechanistic evidence that its benefit operates in part via TDP-43 proteostasis, UPR, and autophagy modulation. The 2022 and 2025 publications specifically tie edaravone to TDP-43 handling in human and cellular models.
**What would justify expansion/repurposing:** Biomarker-driven trials correlating TDP-43 proteostasis changes (e.g., CSF or plasma TDP-43 fragments, pTDP-43) with clinical response; head-to-head or add-on studies with other proteostasis agents at earlier disease stages.

### 🥈 2. Nilotinib
**Why:** FDA-approved, BBB-penetrant, restores autophagic flux via c-Abl inhibition—a mechanism directly relevant to clearing aggregated TDP-43. Extensive safety database and ongoing neurodegeneration trials (PD, AD, DLB) provide clinical infrastructure. Preclinical rescue of autophagy is robust across multiple proteinopathy models.
**What would justify repurposing:** (a) Demonstration of c-Abl activation in ALS motor neurons and CSF; (b) TDP-43-specific efficacy in iPSC-derived motor neuron or patient fibroblast models; (c) Phase 1b dose-finding in ALS confirming target engagement (p-c-Abl reduction) at neuroprotective doses without unacceptable cardiovascular toxicity; (d) Biomarker evidence (NfL, pTDP-43) of biological activity.

### 🥉 3. Colchicine
**Why:** Inexpensive, widely approved, and built on a precise mechanism (HSPB8–autophagy–granulostasis) that directly connects to TDP-43 clearance and stress granule disassembly. The Co-ALS phase 2 low-dose signal on ALSFRS-R decline is intriguing, and secondary biological outcomes (autophagy, SG markers, TDP-43 accumulation, neurofilaments) provide a rich dataset.
**What would justify repurposing:** (a) Confirmation of HSPB8 induction and autophagy flux in human ALS biospecimens (fibroblasts, PBMCs, CSF) with low-dose colchicine; (b) A larger, adequately powered phase 2b/3 trial stratified by TDP-43 pathology burden or stress granule biomarkers; (c) Dose optimization to maximize autophagy while avoiding GI/myelosuppressive toxicity.

---

## 4. Table: Compound Mechanism-ALS Alignment

| Compound | Approved For | TDP-43 Proteostasis | Autophagy Enhancement | Stress Granule Dynamics | ALS Trial Status |
|----------|--------------|---------------------|----------------------|------------------------|------------------|
| **Edaravone** | ALS (FDA) | ✅ Direct evidence (2022, 2025) | ✅ Modulates autophagy gene expression | Indirect (UPR, oxidative stress) | Approved; mechanism expansion ongoing |
| **AMX0035** | ALS (withdrawn) | Indirect (ER stress, chaperoning) | Moderate | Moderate | Phase 3 negative |
| **Nilotinib** | CML | Plausible (autophagy-mediated clearance) | ✅ c-Abl inhibition → autophagy flux | Indirect | Phase 2/3 in PD/AD; no ALS trial yet |
| **Colchicine** | Gout, FMF | ✅ HSPB8 → TDP-43 clearance | ✅ HSPB8-BAG3-HSP70 → autophagy | ✅ Granulostasis | Phase 2 (Co-ALS) – signal at low dose |
| **Trehalose** | — | ✅ TFEB → TDP-43 clearance (cells) | ✅ mTOR-independent autophagy | Indirect | Phase 2/3 negative (HEALEY) |
| **Guanabenz** | HTN (withdrawn) | Indirect (UPR) | Moderate | ✅ ISR/eIF2α → SG dynamics | Phase 2 (ProMISe) – insufficient signal |
| **Arimoclomol** | NPC (FDA 2024) | Variable (TDP-43 models weak) | ✅ TFEB/TFE3 → CLEAR | Moderate | Phase 3 negative (ORARIALS-01) |
| **DNL343 / Fosigotifator** | — | Indirect (prevent SG seeding) | Moderate | ✅ eIF2B activation → SG dissolution | Phase 2/3 negative (HEALEY); preclinical toxicity in SOD1 mice |
| **Rapamycin** | Transplant | Indirect | ✅ mTOR inhibition | Indirect | Phase 2 safe; preclinical toxicity in SOD1 mice |
| **NU-9** | — | ✅ Aggregation inhibitor | ✅ Autophagosome stimulation | Indirect (improves ER proteostasis) | Preclinical only |

---

## 5. What Would Need to Be True for Repurposing Trials

### General Preconditions
For any approved compound to be repurposed for TDP-43–targeted ALS therapy, the following must generally hold:
1. **Target engagement in human CNS:** The compound must reach pharmacologically active concentrations in CSF or brain tissue at tolerable doses.
2. **Biomarker correlation:** A measurable change in a TDP-43–related biomarker (CSF TDP-43, pTDP-43, NfL, autophagy markers, SG components) must correlate with clinical outcomes.
3. **Stratification by pathology:** ALS is heterogeneous; compounds should be tested in populations enriched for TDP-43 pathology (e.g., sporadic ALS, C9orf72-negative, non-SOD1).
4. **Adequate preclinical validation:** Efficacy must be demonstrated in at least two independent TDP-43–relevant models (e.g., iPSC-derived motor neurons, TDP-43 transgenic mice, patient fibroblasts).

### Compound-Specific Preconditions

| Compound | Preconditions for Justified Repurposing |
|----------|----------------------------------------|
| **Edaravone** | Biomarker study linking TDP-43 proteostasis changes to functional benefit; optimal timing (earlier disease stage) in TDP-43–positive patients. |
| **Nilotinib** | ALS-specific preclinical TDP-43 data; Phase 1b showing safe CNS penetration and c-Abl inhibition; biomarker trial correlating autophagy markers to clinical trajectory. |
| **Colchicine** | Confirmation of HSPB8/autophagy target engagement in human ALS biospecimens; larger phase 3 trial at optimized low dose with TDP-43–biomarker stratification. |
| **Trehalose** | Evidence that CNS bioavailability was subtherapeutic in HEALEY; identification of a formulation/route achieving brain autophagy induction without volume overload. |
| **Guanabenz analogue** | Development of next-gen analogue (e.g., IFB-088) with improved CNS penetration and eIF2α selectivity; positive phase 2 data in stratified ALS population. |
| **ISR inhibitors (DNL343/fosigotifator)** | Resolution of paradox that ISR may be protective in motor neurons; identification of ALS subtype with pathological ISR overactivation; negative preclinical mouse data must be reconciled. |

---

## 6. Key Takeaways

1. **The TDP-43 proteostasis–autophagy–stress granule axis is pharmacologically tractable, but ALS heterogeneity and model-specific effects have repeatedly sunk promising candidates.** Trehalose, arimoclomol, AMX0035, and the ISR activators all failed phase 2/3 testing in broad ALS populations.

2. **Edaravone remains the only approved ALS therapy with direct, published evidence for TDP-43 proteostasis enhancement.** Its modest clinical benefit may reflect incomplete target engagement or late-stage intervention; earlier use and biomarker enrichment are logical next steps.

3. **Nilotinib and colchicine offer the strongest repurposing narratives among approved drugs,** combining plausible TDP-43–relevant mechanisms with existing clinical safety data. However, both require ALS-specific evidence of target engagement and refined dosing before phase 3 investment.

4. **Stress granule dynamics remain the most under-tested therapeutic target.** While guanabenz and ISR inhibitors directly modulate the SG pathway, the ISR's potential neuroprotective role in motor neurons creates mechanistic ambiguity that must be resolved before further clinical investment.

5. **NU-9 represents the frontier of TDP-43 aggregation inhibition** but requires clinical translation; its dual activity against SOD1 and TDP-43 suggests a mechanism broadly applicable across ALS genetic subtypes.

---

*References compiled from PubMed-indexed peer-reviewed publications (up to 2026), clinical trial registries (ClinicalTrials.gov), and verified company press releases (Amylyx, Denali, AbbVie/Calico, Seelos) and Wikipedia pharmaceutical records.*