RESCUE-ALS
(2023)Objective
To evaluate the efficacy and safety of CNM-Au8, a catalytically-active gold nanocrystal neuroprotective agent, for treatment of amyotrophic lateral sclerosis (ALS)
Study Summary
• Significant survival benefit observed at 12-month follow-up: 60% reduction in all-cause mortality (HR 0.408, p=0.0429)
• Disease progression events significantly reduced: 78% vs 23% event-free at week 36 (p=0.0125)
Intervention
CNM-Au8 30 mg oral suspension daily vs matched placebo for 36 weeks, with open-label extension
Inclusion Criteria
Adults 30-80 years with possible/probable/definite ALS per Awaji-Shima criteria, newly symptomatic within 24 months or within 12 months of diagnosis, FVC >=60% predicted, stable riluzole for >=30 days
Study Design
Arms: CNM-Au8 30 mg daily, Placebo
Patients per Arm: 23 (CNM-Au8), 22 (placebo)
Outcome
• Exploratory: ALS disease progression events 22% vs 77% (55% absolute risk reduction, p=0.0125)
• Survival: HR 0.408 (95% CI 0.166-1.001, p=0.0429) at 12-month follow-up
Bottom Line
CNM-Au8 did not meet the primary MUNIX endpoint but showed clinically meaningful exploratory benefits including reduced disease progression events (55% absolute risk reduction) and significantly improved long-term survival (60% mortality reduction at 12-month follow-up). The novel use of neurophysiology as primary outcome advances clinical framework for ALS trials.
Major Points
- Primary endpoint not met: summated MUNIX percent change -31.8% vs -39.6% (LS mean difference 7.7%, p=0.43)
- Limb-onset subgroup showed trend favoring CNM-Au8 (LS mean difference 20.9%, p=0.074)
- Bulbar-onset patients showed minimal MUNIX decline regardless of treatment, confounding results
- ALS disease progression events (death, tracheostomy, NIV, gastrostomy) significantly reduced: 78% vs 23% event-free at week 36 (p=0.0125)
- Significant 60% reduction in all-cause mortality at 12-month follow-up (HR 0.408, p=0.0429)
- Quality of life (ALSSQOL-SF) decline significantly slowed (LS mean difference 0.9, p=0.018)
- Proportion free from >=6-point ALSFRS-R decline: 48% vs 18% (30% absolute risk reduction)
- CNM-Au8 well tolerated with no safety signals; TEAEs balanced between groups
- First ALS clinical trial to utilize neurophysiological biomarker as primary outcome
Study Design
- Study Type
- Phase 2, randomized, double-blind, placebo-controlled trial with open-label extension
- Randomization
- Yes
- Blinding
- Double-blind: computerized block randomization in blocks of four by independent statistician. Unblinded pharmacist at each site dispensed medication. All other personnel blinded. OLE participants and clinicians remained blinded to original randomization.
- Sample Size
- 45
- Follow-up
- 36 weeks double-blind + open-label extension (ongoing) + survival follow-up through at least 12 months post-LPLV
- Centers
- 2
- Countries
- Australia
Primary Outcome
Definition: Mean percent change from baseline to week 36 of summated MUNIX score (ADM, APB, BB, TA), indexed to baseline as 100%
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| -39.6% (SE 7.1) | -31.8% (SE 6.6) | - (-11.9 to 27.3 (LS mean difference)) | 0.43 |
Limitations & Criticisms
- Small sample size (n=45) vulnerable to baseline imbalances
- Primary endpoint not met - exploratory outcomes require cautious interpretation
- Bulbar-onset patients did not show expected MUNIX decline, confounding results
- Selection of MUNIX as primary endpoint limited by disease heterogeneity
- Selection of 4 spinal cord-innervated muscles may not capture bulbar phenotype decline
- Placebo group had some baseline disadvantages (older, lower FVC, higher delta-FS, less riluzole use)
- Two-site study limits generalizability and may have contributed to underpowering
- No site-level stratification in statistical analysis
- Slow CNM-Au8 absorption (3-4 weeks for detectable blood levels, >6 months for tissue equilibrium) may explain lack of 24-week efficacy
- OLE participation was optional, introducing selection bias
- Survival benefit sensitivity analysis with lost-to-follow-up imputed as deaths was not significant (p=0.126)
Citation
eClinicalMedicine 2023; 60: 102036