ALS Clinical Trials 2026: Gene Therapy, Antisense Oligonucleotides & New Treatments

ALS research has been defined by heartbreak — drugs that showed compelling signal in SOD1 mouse models failing completely in humans, a pattern repeated so many times that the field spent years debating whether any preclinical model was trustworthy. The 2022–2024 period was genuinely different. Tofersen's accelerated FDA approval for SOD1-ALS was the first precision ALS therapy in history, the ATLAS trial is treating asymptomatic gene carriers before motor neuron loss begins, and the plasma neurofilament light chain biomarker is compressing trial timelines in ways that weren't possible before it was validated. For patients and families navigating this landscape, the critical skill is distinguishing what's real versus what's promising versus what's been seen before — and all three categories exist in ALS right now.

Tofersen (Qalsody): What the VALOR Trial Actually Showed

SOD1 mutations account for approximately 2% of all ALS cases — roughly 200–300 new diagnoses per year in the US — but have been the focus of intense research because the molecular target is unambiguous. The SOD1 gene encodes superoxide dismutase 1; pathogenic mutations cause misfolding of the SOD1 protein, which is directly toxic to motor neurons. Remove the misfolded protein, and you remove the toxic insult.

Tofersen (Qalsody, Biogen) is an intrathecally administered antisense oligonucleotide that binds to SOD1 mRNA and recruits RNase H to degrade it, reducing SOD1 protein production by approximately 35–40% in the CNS. The VALOR Phase 3 trial (NCT02623699, n=108) did not achieve its primary endpoint — the 28-week change in ALSFRS-R functional score was not statistically different between tofersen and placebo in the full population (p=0.97). But the story is more nuanced than that primary endpoint result suggests, and the FDA's response to it changed how ALS drug development works.

Plasma neurofilament light chain (pNfL) — a protein released into the bloodstream when neurons are damaged — decreased by 55% in tofersen-treated patients versus a 12% increase in placebo patients (p<0.001). pNfL is a validated biomarker of neurodegeneration; it tracks disease progression in ALS more sensitively than functional scales over short time windows. Open-label extension data showed that patients who started tofersen earlier had better functional outcomes than those who crossed over from placebo at 28 weeks — a 2.6 point advantage on ALSFRS-R at 52 weeks that matched the functional impact of 6 months of disease progression. FDA approved tofersen in April 2023 under accelerated approval with pNfL reduction as the surrogate endpoint, acknowledging that confirmatory functional data would be required from ATLAS. This was the first use of a neurodegeneration biomarker as the basis for ALS drug approval.

ATLAS: Treating ALS Before It Starts

The ATLAS trial (NCT04856982, Biogen) is the most ambitious ALS intervention ever attempted. It enrolls adults who carry pathogenic SOD1 mutations and have elevated plasma NfL — evidence that neurodegeneration is already occurring — but who have no clinical symptoms of ALS. These individuals will develop ALS with near-certainty if untreated; the question ATLAS is asking is whether treating them with tofersen before motor neurons are lost can prevent or indefinitely delay symptom onset.

The biological hypothesis is straightforward: if tofersen reduces toxic SOD1 protein production before motor neuron death begins, there may be no threshold of damage to trigger clinical ALS. This is analogous to treating familial hypercholesterolemia before cardiovascular events, or treating HIV before CD4 depletion causes AIDS — targeting the pathophysiology before irreversible organ damage occurs. Plasma NfL normalization toward age-matched controls has been observed in ATLAS participants, confirming target engagement. The primary endpoint is time to clinical ALS diagnosis, which will require years of follow-up to assess definitively. Preliminary data is expected in 2026–2027.

Identification of eligible participants requires genetic testing of first-degree relatives of SOD1-ALS patients — something the ALS community has been systematically working to increase through cascade testing programs. If you have a parent or sibling with confirmed SOD1-ALS, genetic testing is the first step toward ATLAS eligibility.

Extending ASOs to Other Genetic ALS Subtypes

Tofersen's proof of concept — that ASO-mediated reduction of a toxic ALS protein can alter the disease trajectory — has directly motivated parallel programs for other genetic ALS subtypes:

C9orf72-ALS/FTD

C9orf72 hexanucleotide (GGGGCC) repeat expansions are the most common genetic cause of ALS, accounting for approximately 40% of familial and 10% of sporadic ALS cases. The pathogenic mechanism involves both a loss-of-function component (reduced normal C9orf72 protein, which regulates autophagy) and a toxic gain-of-function component (toxic RNA foci and dipeptide repeat proteins generated from the expanded repeat). This dual mechanism complicates ASO design: you want to eliminate the toxic repeat-containing transcripts without completely ablating the normal protein. Wave Life Sciences and Novartis are both advancing intrathecal ASOs targeting C9orf72 in Phase 1/2 trials, with selective targeting of the expanded repeat a key design goal. Phase 1 safety data is expected in 2026.

FUS-ALS: ION363 and the FUSION Trial

FUS mutations cause an aggressive early-onset ALS, often presenting in patients under 40. Ionis Pharmaceuticals developed ION363, an intrathecal ASO targeting FUS mRNA. The program received unusual attention when Jaci Hermstad — a 25-year-old with a de novo FUS mutation — received compassionate-use ION363 and showed disease stabilization for an extended period before eventually progressing. This single-patient experience prompted FDA to expedite a Phase 1/2 trial (FUSION, NCT04768972) that is now actively enrolling adult and pediatric patients with FUS-ALS. No comparative efficacy data exists yet, but FUS-ALS patients and families should contact the principal investigator teams if FUS mutations have been confirmed.

Ataxin-2 (ATXN2) Lowering

Ataxin-2 is a poly-Q protein that stabilizes TDP-43 — the pathological hallmark of approximately 97% of all ALS cases, including most sporadic ALS. TDP-43 aggregation and mislocalization is the final common pathway for motor neuron death in virtually all ALS regardless of genetic background. Reducing ataxin-2 expression via ASO reduces TDP-43 aggregation in mouse models, and this is currently the most promising approach to treating non-SOD1, non-C9orf72 ALS. Ionis Pharmaceuticals' Phase 1 trial of IONIS-ATXN2Rx (NCT04494256) is the most advanced program targeting the TDP-43 pathway in humans. If it demonstrates adequate safety and any reduction in pNfL at Phase 1 doses, it would represent a therapeutic candidate potentially applicable to the majority of ALS patients — a much larger population than SOD1 or C9orf72 alone.

The Relyvrio Setback and What It Means

Relyvrio (sodium phenylbutyrate/ursodoxicoltaurine, AMX0035) was approved by FDA in September 2022 based on a Phase 2 trial (CENTAUR, n=137) that showed a 2.32-point slower decline on ALSFRS-R over 24 weeks versus placebo — roughly equivalent to 25% slower progression. The drug combined an endoplasmic reticulum stress inhibitor (phenylbutyrate) with a mitochondria-protecting bile acid derivative, targeting two stress pathways hypothesized to contribute to motor neuron death.

The confirmatory Phase 3 trial (PHOENIX, n=664, NCT05021536) failed to meet its primary endpoint in September 2023 — no significant difference in ALSFRS-R decline between AMX0035 and placebo. Amylyx Pharmaceuticals voluntarily withdrew the drug from the US market in late 2023, a rapid and responsible decision that few companies would have made given the commercial investment. This was a significant setback, but it also demonstrated the system working as designed: accelerated approval requires confirmatory evidence, and when confirmatory evidence fails, the drug comes off the market. The PHOENIX failure was not a surprise to those who followed the CENTAUR data carefully — the effect size in a 137-patient Phase 2 trial was plausible but not robust, and Phase 3 trials at 5× the sample size regularly fail to confirm Phase 2 signals.

Biomarker Advances: Why pNfL Changed Everything

Plasma neurofilament light chain is now the central pharmacodynamic biomarker in ALS drug development, and its validation has changed trial economics fundamentally. Before pNfL, demonstrating that a drug slowed ALS progression required enrolling 100+ patients and following them for 12–18 months — waiting for measurable functional decline on the ALSFRS-R, which is inherently slow and variable. With pNfL, a drug that engages its target and reduces neurodegeneration can show a statistically significant pNfL reduction in 40–60 patients at 12 weeks — an order-of-magnitude reduction in the time and cost required for a proof-of-concept study.

AI-driven platforms — particularly the PRO-ACT database (5,000+ ALS patient records, Prize4Life) and the Answer ALS project (1,000+ patients with multi-omic, clinical, and wearable data) — are doing exactly what they should: identifying biological subgroups within ALS that may respond differently to different mechanisms, building predictive models of progression rate, and matching patients to trials where they're most likely to show measurable benefit. The still-unsatisfying truth is that ALS is not one disease — SOD1-ALS, C9orf72-ALS, FUS-ALS, and TDP-43-driven sporadic ALS likely have distinct enough biology that one drug will not work across all subtypes. The field is slowly accepting this and designing accordingly.

Key Takeaways

• Tofersen (Qalsody) is FDA-approved for SOD1-ALS under accelerated approval — the first precision medicine for any ALS subtype. It reduces plasma NfL by ~55% and early-start data suggests functional benefit over delayed treatment.
• ATLAS (NCT04856982) is treating asymptomatic SOD1 mutation carriers before symptom onset — possibly preventing or delaying ALS indefinitely. Eligible participants are identified by genetic testing; NfL elevation is required for enrollment.
• ASO programs for C9orf72, FUS, and ATXN2 are in Phase 1/2 trials, following tofersen's proof of concept for genetic targeting in ALS. C9orf72 represents ~40% of familial ALS; ATXN2 targeting could be relevant to virtually all ALS.
• Relyvrio's Phase 3 failure (PHOENIX, 2023) was a significant setback but also a validation of the confirmatory trial requirement — accelerated approval is provisional, and negative Phase 3 data resulted in appropriate market withdrawal.
• Plasma NfL as a surrogate biomarker has compressed ALS trial timelines dramatically — enabling proof-of-concept studies in 40–60 patients at 12 weeks versus the 100+ patients at 18 months previously required.

Frequently Asked Questions