Part 3 · Treatment & care
Riluzole, edaravone and others
The drugs approved to slow ALS act on steps that are thought to injure motor neurons, but none stops the disease. Play the synapse to see two of those steps, then lay a drug over it. Below are three decades of approvals and withdrawals, the published label details and what each trial found.
Schematic, not to scale. The gauges show direction only, not measured amounts.
Approved, refused, withdrawn: 1995 to 2026
Scrub the year. Each mark is one regulator's decision or one trial result; the panel names it and gives the source. Letters inside the marks name the drug.
The axis is stretched after 2015 so the crowded recent years stay readable. Dates are as published by each regulator; where a date could not be checked at the regulator it is left out.
Every event, in order (each button moves the timeline there):
What the labels say
Published guidance, quoted as written. This is a summary for study, not instructions for any person's treatment.
| Drug | Route and published dose | Monitoring and warnings | Common adverse effects (drug vs placebo) |
|---|---|---|---|
| Riluzole tablet; oral suspension; (oral film, now discontinued in the US) |
50 mg by mouth twice daily, at least 1 hour before or 2 hours after a meal. 177 The suspension is 5 mg/mL (50 mg = 10 mL twice daily), by mouth or PEG tube. 178 | Liver tests: the US and EU schedules differ (see the chart below). US: not recommended above 5 × ULN; stopped if there is liver dysfunction such as raised bilirubin. EU: stopped if ALT reaches 5 × ULN; contraindicated in hepatic disease or with baseline transaminases above 3 × ULN. 162,177 Warnings: severe neutropenia in the first 2 months, so febrile illness is reported; interstitial lung disease; acute pancreatitis. 177 The EU adds a chest X-ray for dry cough or breathlessness. 162 | Asthenia 19% vs 12%, nausea 16% vs 11%, decreased lung function 10% vs 9%. About 50% had at least one ALT above ULN, 8% above 3 × ULN and 2% above 5 × ULN, peaking within 3 months. 177 |
| Edaravone IV infusion or oral suspension (ORS) |
IV 60 mg over 60 minutes, or ORS 105 mg (5 mL) by mouth or feeding tube in the morning after an overnight fast, no food for 1 hour after. Cycle 1: daily for 14 days, then 14 days off. Later cycles: 10 of 14 days, then 14 days off. 176 | Hypersensitivity and anaphylaxis. Both forms contain sodium bisulfite, which can cause allergic-type reactions, more often in people with asthma. 176 | Contusion 15% vs 9%, gait disturbance 13% vs 9%, headache 10% vs 6%, dermatitis 8% vs 5%, eczema 7% vs 4%. 176 |
| Sodium phenylbutyrate-taurursodiol (AMX0035) withdrawn from the US and Canada in 2024 |
In CENTAUR: 3 g sodium phenylbutyrate + 1 g taurursodiol once daily for 3 weeks, then twice daily. 415 | No longer marketed; existing users were moved to a free-drug programme. 17 | Mainly gastrointestinal in CENTAUR. 415 |
| Tofersen intrathecal, SOD1-ALS only |
100 mg by lumbar puncture: three loading doses 14 days apart, then every 28 days. 172 Its trial, safety and approval are in Tofersen, gene-targeted. | ||
Pregnancy: the US riluzole label says it may cause fetal harm based on animal data, and the EU contraindicates it in pregnancy and breast-feeding. 162,177 Edaravone caused developmental effects in rats and rabbits, and there are no human data. 176
Riluzole liver tests: US label vs EU SmPC
Each dot is a liver (ALT) test as the label describes it. "Periodically" is the labels' own word; neither gives an interval. 162,177
What the trials found, in absolute numbers
About 2 to 3 more months of median survival
HR 0.8495% CI 0.698 to 0.997
Cochrane 2012, four trials (974 on riluzole, 503 on placebo), tracheostomy-free survival on 100 mg/day. One-year survival 49% on placebo vs 58% on riluzole; median survival 11.8 vs 14.8 months. 334
No benefit on muscle strength. The EU label says there is no evidence of effect on motor function, lung function or strength, and none in late-stage ALS. 162,334
Real-world studies claim much more
6 to 19months, in 8 of 15 studies
Andrews 2020 reviewed 15 population studies; 8 reported median survival 6 to 19 months longer in people treated with riluzole. 22
These are not randomised. People who are offered and keep taking a drug differ from those who are not (confounding by indication), and time before treatment starts can be counted in the treated group's favour (immortal-time bias). The trial figure of 2 to 3 months remains the reference. 22,334
First phase 3 (MCI186-16): no difference
0.65points, p = 0.411
ALSFRS-R change over 24 weeks: −5.70 on edaravone (n = 100) vs −6.35 on placebo (n = 99). 1
A post hoc search found a subgroup (vital capacity at least 80%, every ALSFRS-R item at least 2, onset within 2 years) where the difference was larger. That subgroup became the entry criteria for the next trial. 152
MCI186-19: positive in the post hoc–selected group
2.49points, 95% CI 0.99 to 3.98
137 people randomised (69 vs 68), chosen to match the subgroup above. ALSFRS-R change at 24 weeks −5.01 vs −7.50, about a third less decline. More than 90% were also taking riluzole. 176,325
The authors saw "no indication" of benefit outside these criteria. 325 In grade 3 disease (MCI186-18) there was no benefit. 153
Later studies found no added benefit
- Germany, 116 vs 116 matched people: −0.91 vs −0.85 points/month (P = 0.37); survival also no different. 542
- Oral FAB122 (ADORE, 313 people): −11.2 vs −10.8 at 48 weeks, primary not met. 122
- Daily oral vs on/off dosing (384 people): stopped for futility. 124
The EU application was withdrawn in 2019, and EAN 2024 recommends against IV or oral edaravone outside a clinical trial. 158,506
CENTAUR positive, PHOENIX negative
0.42points/month, 95% CI 0.03 to 0.81
CENTAUR (phase 2, 137 people): ALSFRS-R slope −1.24 vs −1.66 points/month over 24 weeks (P = 0.03); strength, breathing and survival outcomes did not differ. 415 An open-label survival follow-up reported 25.0 vs 18.5 months (HR 0.56), comparing early with later starting. 416
PHOENIX (phase 3, 664 people, 48 weeks) missed its primary endpoint (p = 0.667) and every secondary endpoint. Withdrawal began on 4 April 2024, about 19 months after US approval. 17
A narrowed population
3.4points, 95% CI 0.65 to 6.13
In 394 people on riluzole, 4.5 mg/kg/day gave this ALSFRS-R difference only in the prespecified "normal progressor" group; there was no effect in the broader cohort, and serious adverse events were 31% vs 18%. 347 EMA refused it twice, citing inspection findings and methods. 161
Ultra-high-dose vitamin B12 injections
1.97points, 95% CI 0.44 to 3.50
JETALS: 130 people within 1 year of onset, 50 mg intramuscular twice weekly for 16 weeks; ALSFRS-R −2.66 vs −4.63. 407 Backed for ALS by the Japanese committee in August 2024; no FDA or EMA approval was found. 428
What the guidelines say
EAN 2024: "Offer lifelong riluzole to all people with ALS at diagnosis" (strong), with dose reduction and then stopping considered if adverse events persist. It recommends against edaravone outside a clinical trial, did not recommend AMX0035 outside trials before phase 3 data (it was written before PHOENIX), and recommends offering tofersen first-line in progressive SOD1-ALS. 506
Earlier: AAN 2009 says riluzole should be offered to slow progression (Level A), and EFNS 2012 advises starting it as early as possible. 20,332
Phase 3 trials that were negative
| Agent (rationale) | Trial | Design and size | Result |
|---|
Ceftriaxone looked positive in stage 2 (0.51 points/month, p = 0.04) and then failed in stage 3, an example of an early result that did not replicate. 128 The lithium signal in UNC13A carriers was exploratory and led to a new trial, not to approval. 511
Some drugs treat symptoms rather than slowing ALS. Dextromethorphan-quinidine (approved by the FDA for pseudobulbar affect in 2010) improved the CNS-LS score by 3.3 points more than dextromethorphan alone over 28 days. 76,149 Symptom care is in Living with ALS and Breathing and NIV.