← Back

AMX0035 CENTAUR

Share Share Copied! Compare

Year of Publication: 2020

Journal: New England Journal of Medicine

Link: https://doi.org/10.1056/NEJMoa1916945

PDF: https://www.nejm.org/doi/pdf/10.1056/NEJMoa1916945

Clinical Question

Does sodium phenylbutyrate-taurursodiol slow the rate of decline on the ALSFRS-R in persons with definite ALS?

Bottom Line

In 137 patients with definite ALS within 18 months of onset, sodium phenylbutyrate-taurursodiol (PB-TURSO) slowed ALSFRS-R decline by 0.42 points/month vs placebo over 24 weeks (-1.24 vs -1.66; difference +0.42; 95% CI 0.03-0.81; p=0.03). Secondary outcomes (strength, SVC, pNF-H, time-to-event) did not differ significantly. Trial supported FDA approval of AMX0035 in 2022; approval later withdrawn in 2024 after confirmatory PHOENIX failed.

        Major Points
        Phase 2 multicenter randomized double-blind placebo-controlled trial at 25 NEALS sites in US (CENTAUR, Paganoni NEJM 2020)
N=137 adults with definite ALS by El Escorial criteria within 18 months of symptom onset
2:1 randomization to AMX0035 (sodium phenylbutyrate 3 g + taurursodiol 1 g) vs placebo × 24 weeks
Primary endpoint: rate of ALSFRS-R decline over 24 weeks
ALSFRS-R decline: -1.24 (active) vs -1.66 (placebo) points/month; difference 0.42 (95% CI 0.03-0.81); p=0.03
Secondary outcomes (ATLIS strength, SVC, pNF-H, time-to-event) trended favorably but did not reach significance
Composite death/tracheostomy/hospitalization HR 0.53 (95% CI 0.27-1.05)
Baseline edaravone imbalance (50% placebo vs 25% active) may have blunted placebo decline
GI AEs (diarrhea, nausea, abdominal pain) drove 19% active vs 8% placebo discontinuation
Supported FDA approval of Relyvrio in 2022 for ALS
Confirmatory PHOENIX phase 3 (N=664) failed in 2024 and Amylyx voluntarily withdrew Relyvrio from US/Canada market

      

Design

Study Type: Phase 2 multicenter randomized double-blind placebo-controlled trial

Randomization: 1

Blinding: Double-blind

Follow-up Duration: 24 weeks

Sample Size: 137

Analyzed: 135

Analysis: Mixed-effects regression with prebaseline ALSFRS-R slope and age as covariates; mITT

Baseline Characteristics

Characteristic	Control	Active
N	48	89
Age mean	~58	~58
Symptom duration	<18 mo	<18 mo
Riluzole	~77%	~77%
Edaravone	50% (higher than active)	25%
Bulbar onset	21%	30%

Arms

Field	Control	PB-TURSO (AMX0035)
N	48	89
Intervention	Matching placebo powder sachet, once daily for 3 weeks then twice daily	Sodium phenylbutyrate 3 g + taurursodiol 1 g per sachet, once daily for 3 weeks then twice daily
Duration	24 weeks	24 weeks

Outcomes

Outcome	Type	Control	Intervention	HR / OR / RR	P-value
Rate of decline in ALSFRS-R total score (0-48, higher = better function) over 24 weeks	Primary	-1.66 points/month	-1.24 points/month		p=0.03
ATLIS total muscle strength	Secondary	-3.54%/month	-3.03%/month	Difference 0.51 (95% CI -0.12 to 1.14)	NS (hierarchical stop)
Plasma pNF-H rate of change	Secondary	-2.34 pg/mL/month	+3.58 pg/mL/month	Difference 5.92 (95% CI -4.41 to 16.26)	NS
SVC (% predicted)	Secondary	-4.03%/month	-3.10%/month	Difference 0.93 (95% CI -0.10 to 1.95)	NS
Time to death/tracheostomy/permanent ventilation	Secondary	Reference	Favorable trend	Not primary (cumulative)	NS individually
Composite: death/tracheostomy/hospitalization	Secondary	Reference	Reduced hazard	HR 0.53 (95% CI 0.27-1.05)	NS
Edaravone-corrected ALSFRS-R diff at week 24	Secondary	Reference	2.15 points (95% CI -0.05 to 4.35)		Trend (did not reach significance)
Riluzole-corrected ALSFRS-R diff at week 24	Secondary	Reference	2.34 points (95% CI 0.19-4.48)		Significant in sensitivity
Any AE	Adverse	96%	97%		Comparable
Diarrhea	Adverse	Less common	More common		Drug-specific
Nausea	Adverse	Less common	More common		Drug-specific (early)
Abdominal discomfort	Adverse	Less common	More common		Drug-specific (early)
Salivary hypersecretion	Adverse	Uncommon	More common		Drug-specific
Serious AE	Adverse	19%	12%		Lower with PB-TURSO
Respiratory SAE	Adverse	8%	3%		Lower with PB-TURSO
Discontinuation for AE	Adverse	8%	19%		Higher with PB-TURSO (driven by GI)
Death	Adverse	2 (4%)	5 (6%)		Similar; respiratory failure most common

Subgroup Analysis

Sensitivity analyses correcting for edaravone (which was imbalanced at baseline, 50% placebo vs 25% active) gave a nonsignificant 2.15-point week-24 difference; riluzole-corrected analysis reached significance (2.34 points; 95% CI 0.19-4.48). Fine-motor subscore drove most of the ALSFRS-R benefit. Trial was not powered for subgroup analyses; benefit appeared concentrated in earlier, more rapidly progressing disease.

Criticisms

Modest N (137) and 24-week duration — short relative to ALS natural history; PHOENIX (N=664, 48 weeks) was needed for confirmation and later failed
Baseline edaravone imbalance (50% placebo vs 25% active) may have blunted the placebo arm's decline — edaravone-corrected sensitivity analysis did not reach significance
Pre-specified alpha was 0.1 for sample size; final alpha 0.05 with barely-significant p=0.03 — marginal on classic criteria
Hierarchical secondary outcomes all failed after primary; biomarker (pNF-H) and strength/SVC did not support a biological effect
Open-label extension suggested longer-term survival benefit via historical control comparison — not a randomized finding
FDA approval in 2022 was controversial (9-2 advisory vote after initial 6-4 against); 2024 PHOENIX failure led to market withdrawal

Funding

Amylyx Pharmaceuticals, ALS Finding a Cure Foundation, ALS Association

Based on: AMX0035 CENTAUR (New England Journal of Medicine, 2020)

Content summarized and formatted by NeuroTrials.ai.

AMX0035 CENTAUR

(2020)

Objective

Sodium phenylbutyrate-taurursodiol (AMX0035) — to evaluate whether an oral coformulation that targets endoplasmic reticulum stress and mitochondrial dysfunction slows functional decline in amyotrophic lateral sclerosis.

Study Summary

• PB-TURSO slowed ALSFRS-R decline by 0.42 points/month vs placebo over 24 weeks (p=0.03).
• Mean monthly rate: -1.24 (active) vs -1.66 (placebo); absolute 2.32-point difference at week 24.
• Secondary outcomes (ATLIS strength, SVC, pNF-H, time to death/tracheostomy) all trended favorably but did not reach significance.
• Hazard ratio for death/tracheostomy/hospitalization: 0.53 (95% CI 0.27-1.05).
• Gastrointestinal AEs (diarrhea, nausea, abdominal pain) drove 19% active vs 8% placebo discontinuation.
• Supported FDA approval (2022) as Relyvrio; withdrawn 2024 after negative PHOENIX phase 3 confirmatory trial.

Intervention

Sodium phenylbutyrate 3 g + taurursodiol 1 g per sachet, orally or via feeding tube, once daily for 3 weeks then twice daily, vs matching placebo. 24-week planned treatment, 2:1 randomization.

Inclusion Criteria

Adults with definite ALS by revised El Escorial criteria within 18 months of symptom onset; SVC >60% predicted; no riluzole or stable riluzole ≥30 days; edaravone permitted after 2017 protocol amendment.

Study Design

Arms: Sodium phenylbutyrate-taurursodiol vs Placebo (2:1)

Patients per Arm: PB-TURSO 89; Placebo 48 (randomized N=137; mITT 87 active + 48 placebo)

Outcome

• Primary: ALSFRS-R rate of decline — -1.24 points/month (PB-TURSO) vs -1.66 (placebo); difference 0.42 (95% CI 0.03-0.81); p=0.03
• ATLIS total strength: -3.03 vs -3.54%/month; difference 0.51 (95% CI -0.12 to 1.14); NS (hierarchical stop)
• SVC: -3.10 vs -4.03%/month; difference 0.93 (95% CI -0.10 to 1.95); NS
• Plasma pNF-H: +3.58 vs -2.34 pg/mL/month; difference 5.92 (95% CI -4.41 to 16.26); NS
• Composite death/tracheostomy/hospitalization HR 0.53 (95% CI 0.27-1.05); 7 deaths total (5 active, 2 placebo)

Bottom Line

Major Points

Phase 2 multicenter randomized double-blind placebo-controlled trial at 25 NEALS sites in US (CENTAUR, Paganoni NEJM 2020)
N=137 adults with definite ALS by El Escorial criteria within 18 months of symptom onset
2:1 randomization to AMX0035 (sodium phenylbutyrate 3 g + taurursodiol 1 g) vs placebo × 24 weeks
Primary endpoint: rate of ALSFRS-R decline over 24 weeks
ALSFRS-R decline: -1.24 (active) vs -1.66 (placebo) points/month; difference 0.42 (95% CI 0.03-0.81); p=0.03
Secondary outcomes (ATLIS strength, SVC, pNF-H, time-to-event) trended favorably but did not reach significance
Composite death/tracheostomy/hospitalization HR 0.53 (95% CI 0.27-1.05)
Baseline edaravone imbalance (50% placebo vs 25% active) may have blunted placebo decline
GI AEs (diarrhea, nausea, abdominal pain) drove 19% active vs 8% placebo discontinuation
Supported FDA approval of Relyvrio in 2022 for ALS
Confirmatory PHOENIX phase 3 (N=664) failed in 2024 and Amylyx voluntarily withdrew Relyvrio from US/Canada market

Study Design

Study Type: Phase 2 multicenter randomized double-blind placebo-controlled trial
Randomization: Yes
Blinding: Double-blind
Sample Size: 137
Follow-up: 24 weeks

Primary Outcome

Definition: Rate of decline in ALSFRS-R total score (0-48, higher = better function) over 24 weeks

Control	Intervention	HR/OR	P-value
-1.66 points/month	-1.24 points/month	- (0.03 to 0.81)	p=0.03

Limitations & Criticisms

Modest N (137) and 24-week duration — short relative to ALS natural history; PHOENIX (N=664, 48 weeks) was needed for confirmation and later failed
Baseline edaravone imbalance (50% placebo vs 25% active) may have blunted the placebo arm's decline — edaravone-corrected sensitivity analysis did not reach significance
Pre-specified alpha was 0.1 for sample size; final alpha 0.05 with barely-significant p=0.03 — marginal on classic criteria
Hierarchical secondary outcomes all failed after primary; biomarker (pNF-H) and strength/SVC did not support a biological effect
Open-label extension suggested longer-term survival benefit via historical control comparison — not a randomized finding
FDA approval in 2022 was controversial (9-2 advisory vote after initial 6-4 against); 2024 PHOENIX failure led to market withdrawal

Citation

View Original Publication →

AMX0035 CENTAUR

Clinical Question

Bottom Line

Major Points

Design

Baseline Characteristics

Arms

Outcomes

Subgroup Analysis

Criticisms

Funding

Ask about AMX0035 CENTAUR