← Back
NeuroTrials.ai
Neurology Clinical Trial Database

VALOR-OLE

Long-Term Tofersen in SOD1 Amyotrophic Lateral Sclerosis: Integrated Analysis of the Phase 3 VALOR Trial and Open-Label Extension

Year of Publication: 2025

Authors: Miller TM, Cudkowicz ME, Shaw PJ, ..., et al.; VALOR and OLE Working Group

Journal: JAMA Neurology

Citation: JAMA Neurol. 2025 Dec 22;83(2):115-125. doi:10.1001/jamaneurol.2025.4946

Link: https://doi.org/10.1001/jamaneurol.2025.4946

Bottom Line

Over a median 4.9-year follow-up, earlier tofersen was associated with numerically less decline in ALSFRS-R, SVC, HHD strength, and quality-of-life measures and substantial sustained reductions in plasma neurofilament light chain; in the faster-progressing (high-NfL) subgroup, early start extended event-free survival by ~3.4 years, supporting use of tofersen as disease-modifying therapy for SOD1-ALS and earlier initiation in symptomatic disease.

Major Points

  • Integrated ITT analysis of phase 3 VALOR (28 weeks, 2:1 tofersen 100 mg vs placebo) and its open-label extension; 95/108 entered OLE; median opportunity for follow-up 4.9 years (range 3.6–5.4 years).
  • Plasma NfL fell ~67% (early-start) and ~64% (placebo/delayed-start) from baseline and stayed suppressed through week 148; total CSF SOD1 protein fell ~21%–25%, confirming sustained target engagement.
  • At 148 weeks, early start vs placebo/delayed start showed numerically less decline on ALSFRS-R (−9.9 vs −13.5), %-predicted SVC (−13.8 vs −18.1), HHD megascore (−0.38 vs −0.43), and ALSAQ-5 (17.0 vs 22.5) — all favoring early start but not statistically significant on ANCOVA (joint rank test for ALSFRS-R P=.05).
  • Compared with ALS natural history (expected ~22–34-point ALSFRS-R decline over 148 weeks), both groups declined substantially less, consistent with disease-progression slowing in both arms.
  • 21%–27% of early-start participants improved on function/strength measures at 148 weeks; simulation against the dexpramipexole EMPOWER natural-history model estimated the probability of a 27.3% HHD improver rate at 1 year in untreated patients to be 0% (P<.001).
  • Event-free survival in the ITT population: HR for death or permanent ventilation 0.64 (0.28–1.46); HR for death 0.52 (0.20–1.36); HR for death/PV/withdrawal-for-progression 0.61 (0.31–1.18). In the high-NfL faster-progressing subgroup, early start extended median time to death/PV by ~3.4 years (253.6 vs 76.0 weeks) and median time to death by ~2.65 years.
  • Safety: 103/104 (99%) had any AE; 55.8% had a serious AE; 21.2% had a fatal-outcome event (driven by ALS progression); 26.9% discontinued for AE. Serious neurologic AEs in 8.7% (myelitis 3.9%, papilledema/raised ICP 2.9%, aseptic/chemical meningitis 1.9%) — all reversible with standard care; only 1 myelitis and 1 chemical meningitis led to drug discontinuation.
  • Findings supported FDA acceptance of NfL as a surrogate biomarker reasonably likely to predict clinical benefit in SOD1-ALS; tofersen is approved in the US, EU, China, Japan and others, and European Academy of Neurology guidelines recommend it as first-line therapy for progressive SOD1-ALS.

Design

Study Type: Phase 3 randomized, double-blind, placebo-controlled trial (VALOR) with prospectively integrated open-label extension (OLE)

Randomization: 1

Blinding: Double-blind in VALOR (28 weeks); OLE personnel, participants, and caregivers remained blinded to VALOR randomization assignment

Enrollment Period: VALOR: March 2019 – July 2021; OLE completed August 2024

Follow-up Duration: Median opportunity for follow-up 4.9 years (range 3.6–5.4 years); main efficacy analysis at week 148; survival analyses through week 276

Centers: 32

Countries: USA, UK, Canada, Japan, Italy, Belgium, Germany, France, Spain, Netherlands

Sample Size: 108

Analysis: Intention-to-treat with multiple imputation; ANCOVA for 148-week change scores; joint rank test for ALSFRS-R; Cox regression for time-to-event (adjusted for baseline plasma NfL and riluzole/edaravone use); NfL-based prognostic subgrouping (median 75.6 pg/mL)


Inclusion Criteria

  • Adults ≥18 years of age
  • Weakness attributable to ALS
  • Confirmed pathogenic or likely pathogenic SOD1 variant
  • Pre-randomization ALSFRS-R slope and SOD1 variant type used to enrich for faster progression in VALOR primary analysis population
  • Able to complete the 28-week double-blind VALOR study and consent to OLE participation

Exclusion Criteria

  • Contraindications to intrathecal administration / lumbar puncture
  • Clinically significant medical or neurological comorbidity that would interfere with study participation
  • Concomitant investigational ALS therapies (riluzole and edaravone permitted)
  • Inability to comply with study procedures (functional or cognitive)

Baseline Characteristics

CharacteristicPlacebo/Delayed Start (n=36)Early Start (n=72)
Mean Age (SD), y51.2 (11.6)48.1 (12.6)
Sex - Male53% (19/36)60% (43/72)
BMI, mean (SD)27.4 (6.5)26.4 (5.6)
Riluzole use61%62%
Edaravone use8%8%
Time from symptom onset, median (range), mo14.6 (2.4–103.2)11.4 (1.7–145.7)
Plasma NfL, mean (SD), pg/mL89.7 (86.5)100.4 (82.8)
Plasma NfL, geometric mean, pg/mL56.666.6
ALSFRS-R total score, mean (SD)37.3 (5.8)36.9 (5.9)
%-Predicted SVC, mean (SD)85.1 (16.5)82.1 (16.6)

Arms

FieldEarly-Start TofersenControl
InterventionIntrathecal tofersen 100 mg from VALOR week 1 (with continuation in OLE); dosing per VALOR protocol every 4 weeks after loadingIntrathecal placebo for 28 weeks in VALOR, then crossover to tofersen 100 mg at OLE entry (~6 months later)
Duration≥3 years on-drug (up to 5.4 years total follow-up)~0.5 years placebo + ≥2.5 years tofersen (up to 5.4 years total follow-up)

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Integrated 148-week change from baseline in ALSFRS-R total score (early-start vs placebo/delayed-start)Primary−13.5 points−9.9 points
%-Predicted SVC change to week 148Secondary−18.1−13.8.44
HHD megascore change to week 148Secondary−0.43−0.38.55
Plasma NfL reduction (week 148, 1 − geometric mean ratio)Secondary64.0%67.0%
CSF total SOD1 protein reduction (week 148, 1 − geometric mean ratio)Secondary25.0%21.0%
ALSAQ-5 change to week 148 (lower = better QoL)Secondary22.517.0
EQ-5D-5L utility change to week 148Secondary−0.2−0.1
Time to death or permanent ventilation (ITT)SecondaryMedian not estimableMedian not estimable
Time to death (ITT)SecondaryMedian not estimableMedian not estimable
Time to death, PV, or withdrawal for disease progression (ITT)SecondaryMedian NEMedian 260.7 weeks
Faster-progressing (high-NfL) subgroup: time to death or PVSecondaryMedian 76.0 weeksMedian 253.6 weeks
Faster-progressing subgroup: time to deathSecondaryMedian 115.4 weeksMedian 253.6 weeks
Proportion improving (ITT) at 148 weeks on ALSFRS-RSecondary17.3% (95% CI 4.7–29.8)21.0% (11.6–30.5)
Proportion improving (ITT) on HHD megascore at 148 weeksSecondary10.7% (−0.5 to 21.8)27.3% (16.3–38.3)
Simulated probability of observed 27.3% HHD improver rate at 1 year in untreated patients (dexpramipexole/EMPOWER natural-history model)Secondary0%<.001
Any AEAdverse99.0% (103/104)
Any related to trial agentAdverse63.5%
Any related to lumbar punctureAdverse83.7%
Serious AEAdverse55.8%
Serious AE related to trial agentAdverse9.6%
Fatal-outcome eventAdverse21.2%
AE leading to discontinuationAdverse26.9%
HeadacheAdverse60.6%
Procedural painAdverse59.6%
FallAdverse47.1%
Back painAdverse47.1%
Pain in extremityAdverse39.4%
ArthralgiaAdverse36.5%
COVID-19Adverse34.6%
FatigueAdverse30.8%
CSF protein increasedAdverse26.9%
Post-lumbar-puncture syndromeAdverse25.0%
NauseaAdverse25.0%
ConstipationAdverse25.0%
Serious respiratory failureAdverse12.5%
Serious aspiration pneumoniaAdverse11.5%
Serious dysphagiaAdverse7.7%
Serious pulmonary embolismAdverse5.8%
Serious neurologic AE (any)Adverse8.7% (9/104) — all reversible
Myelitis (incl. transverse myelitis)Adverse3.9%
Papilledema / raised ICPAdverse2.9%
Aseptic / chemical meningitisAdverse1.9%
RadiculitisAdverse1.0%
Postbaseline CSF leukocytes >10×10^6/LAdverse79.8% (83/104)
Shift to elevated CSF protein during treatmentAdverse88.2% (60/68)

Subgroup Analysis

Numerical favoring of early start was consistent across the ITT and NfL-based faster- and slower-progressing subgroups. Largest treatment effects were observed in the faster-progressing (high-NfL) subgroup, where median time-to-event for both treatment arms could be estimated (early start delayed median time to death/PV by ~3.4 years). In carriers of the p.Ala5Val (A5V) variant (the most common North-American/faster variant), median disease duration was 23.2 mo (early start, n=11) vs 15.4 mo (delayed start, n=6).


Criticisms

  • Small ITT sample (N=108) — analyses were not powered to detect statistically significant treatment differences on clinical outcomes; reported between-group comparisons are numerical/exploratory.
  • All participants had the opportunity to receive tofersen from week 28 onward, so the long-term comparison reflects only a ~6-month timing offset rather than treatment vs no treatment.
  • Substantial dropout/death by OLE end (only 46/108 completed), and time-to-event medians not estimable in the ITT population — extension-of-survival estimates rely on the faster-progressing NfL subgroup.
  • SOD1 variant heterogeneity (42 unique variants) limits subgroup precision and generalization across SOD1 genotypes.
  • Funded and largely analyzed by Biogen; multiple authors are Biogen employees/shareholders, with the sponsor involved in design, monitoring, statistical analysis, interpretation, and drafting.
  • Comparisons to natural history rely on external cohorts (PRO-ACT, dexpramipexole/EMPOWER) and modeling assumptions; no contemporaneous untreated control beyond 6 months.
  • Multiple imputation and joint-rank test results were positive only as alternative analyses (ALSFRS-R P=.05 for JRT vs P=.14 for ANCOVA), highlighting sensitivity to analytic choice.

Funding

Biogen (study funder, drug supplier, sponsor of analysis and medical writing)

Based on: VALOR-OLE (JAMA Neurology, 2025)

Authors: Miller TM, Cudkowicz ME, Shaw PJ, ..., et al.; VALOR and OLE Working Group

Citation: JAMA Neurol. 2025 Dec 22;83(2):115-125. doi:10.1001/jamaneurol.2025.4946

Content summarized and formatted by NeuroTrials.ai.