Therapeutic Challenges and Future Breakthroughs in Amyotrophic Lateral Sclerosis: From Precision Medicine to Innovative Trial Design Authors: Yutong Wei, Jie Wang, Yanan Ji, Tongxin Shang, Yuntian Shen, Hualin Sun, Yong Cai Published: 2026 ABSTRACT AbstractAmyotrophic lateral sclerosis (ALS) is a highly heterogeneous and fatal neurodegenerative disorder, for which clinical management and drug development have long faced formidable challenges. Since the approval of riluzole and edaravone, dozens of promising drug candidates that showed efficacy in preclinical models have failed in Phase III trials, highlighting an urgent need for systematic re-evaluation of the field. This review provides a comprehensive summary of the major limitations of current clinical therapies for ALS. These include the modest survival benefit of riluzole, the narrow eligible population for edaravone, and the complex trajectory of sodium phenylbutyrate-taurursodiol, which received accelerated approval but was subsequently voluntarily withdrawn after its confirmatory Phase III trial failed to meet its primary endpoints. On this basis, we discuss four major challenges that contribute to clinical trial failures: disease heterogeneity, paucity of reliable biomarkers, insufficient translational validity of preclinical models, and inherent flaws in conventional trial designs. Subsequently, we discuss emerging therapeutic strategies, encompassing precision medicine and gene therapy (exemplified by the development of the antisense oligonucleotide tofersen for SOD1-ALS), targeting protein homeostasis, modulation of neuroinflammation, metabolic and energetic support, neuroprotection and regeneration, as well as multi-target combination approaches. Innovative trial designs, including adaptive platform trials (exemplified by the HEALEY ALS Platform Trial), enrichment designs, sequential designs, N-of-1 trials, and virtual clinical trials are fundamentally reshaping the drug development paradigm in ALS. In conclusion, ALS treatment is at a historic turning point from a “one-size-fits-all” approach toward “precisely stratified” medicine. Future success depends on establishing multimodal biomarker panels, implementing genetic testing-guided individualized therapy, developing combination regimens, and integrating patient-reported outcomes with palliative care. Although substantial challenges remain, the clinical success of Tofersen provides evidence that precision therapeutic strategies may gradually transform ALS management toward a more individualized and disease-modifying approach.Keywords: amyotrophic lateral sclerosis, clinical treatment, challenges, biomarkers, gene therapy, clinical trial design Introduction Amyotrophic lateral sclerosis (ALS), a progressive neurodegenerative disorder primarily affecting motor neurons in the brain and spinal cord, is the medical term for “Lou Gehrig’s disease”. Clinically, it is characterized by progressive muscle weakness and atrophy, ultimately involving the respiratory muscles and leading to death from respiratory failure.1,2 ALS exhibits profound heterogeneity across multiple dimensions, including age of onset (young adult to elderly), site of onset (bulbar or limb), rate of progression (months to years), and genetic background (approximately 10% familial, 90% sporadic).3 This remarkable heterogeneity represents the primary challenge in the clinical management and drug development of ALS, rendering traditional “one-size-fits-all” therapeutic strategies largely ineffective. Since the first Phase III clinical trial in 1994 demonstrated that riluzole (a glutamate release inhibitor) could extend patient survival by approximately 2–3 months,4 drug development for ALS entered the “riluzole era”.5 More than two decades later, edaravone (a free radical scavenger) showed efficacy in delaying functional decline in a specific subset of early-stage, rapidly progressing patients in Japan and the United States,6 becoming the second disease-modifying therapy approved for ALS. Although edaravone has been approved by the US. Food and Drug Administration and in several other countries, the European Medicines Agency has not granted marketing authorization, citing insufficient evidence of clinical benefit. Nevertheless, the clinical benefits of both drugs remain very limited and far from meeting patient needs. Furthermore, real-world studies have raised concerns about the long-term efficacy of edaravone: a propensity score-matched cohort study showed that although long-term intravenous edaravone was feasible and well tolerated, it did not confer any disease-modifying benefit compared with standard care.7 Consequently, the development of more effective therapeutic strategies remains an urgent task in the ALS field. Since then, dozens of drug candidates that showed strong promise in cellular and animal models, including anti-glutamate agents (e.g, memantine, topiramate), neurotrophic factors (e.g, IGF-1, CNTF), anti-inflammatory drugs (e.g, celecoxib, minocycline), antioxidants (e.g, coenzyme Q10), and modulators of autophagy/proteostasis (e.g, lithium, valproic acid), have failed in large clinical trials. For example, despite encouraging preclinical data, minocycline not only failed to slow disease progression but also showed a potential trend toward accelerating functional decline in a Phase III trial.8 In a dose-finding study of lithium, a high proportion of participants withdrew due to adverse events or perceived inefficacy, highlighting tolerability issues.9 Similarly, lithium, which possesses multi-target potential, failed to replicate initial small-sample positive results in a series of rigorously designed trials.10,11 These sobering lessons have compelled the field to critically reflect on key questions: Was the target choice incorrect? Do drugs fail to reach the central nervous system effectively? Or are clinical trial designs themselves unable to capture modest treatment effects? Answering these questions is essential for guiding future drug development (Table 1). Table 1Clinical Trials Shaping Therapeutic Development in ALSTherapyTarget/MechanismKey Trial(s)PhaseOutcomeKey LessonsRiluzoleGlutamate release inhibitionBensimon et al (1994)IIIModest survival benefit (~2–3 months)First approved therapy; effect largely confined to late-stage diseaseEdaravoneFree radical scavengerMCI186–19 (2017)IIIBenefit in selected early-stage subgroupValidates enrichment strategies for trial designPB-TURSOMitochondrial/proteostasisCENTAUR (2020) / PHOENIXII / IIIPositive Phase II, negative Phase IIIHighlights critical need for confirmatory Phase III trialsMinocyclineAnti-inflammatoryGordon et al (2007)IIINo benefit; potential acceleration of declineDemonstrates risk of extrapolating preclinical dataTofersenSOD1 ASOVALOR (open-label extension)IIIBiomarker reduction and long-term clinical benefitLandmark success of genetically targeted therapyBIIB078C9orf72 ASOVan den Berg et al (2024)ISafe, no clinical benefit (discontinued)Target engagement alone insufficient for efficacyApilimodPIKfyve inhibitorBabu et al (2024)IIaReduced CSF poly(GP) protein levelsDemonstrates CNS target engagement for C9orf72-ALSZilucoplanComplement C5 inhibitorHEALEY ALS Platform TrialIIINo benefit (discontinued)Terminal complement pathway may not be effective target Clinical Trials Shaping Therapeutic Development in ALS In recent years, a deeper understanding of the genetic and pathological mechanisms underlying ALS, particularly the discovery of TAR DNA-binding protein 43 (TDP-43) pathology, the C9orf72 hexanucleotide repeat expansion, and mutations in superoxide dismutase 1 (SOD1), has driven a paradigm shift in therapeutic strategies from a “one-size-fits-all” approach toward precision medicine-based subtyping.12 The antisense oligonucleotide (ASO) drug Tofersen, which targets SOD1 mRNA, has significantly reduced levels of SOD1 protein and neurofilament light chain (NfL) in its Phase III trial and open-label extension study. Moreover, long-term follow-up data have revealed trends toward decelerated functional decline and extended survival.13 These findings represent an important advance in ALS treatment and provide clinical validation of gene-targeted therapy in a genetically defined ALS subgroup. Although the development of the ASO BIIB078 targeting C9orf72-associated ALS was discontinued due to a lack of clinical benefit,12 subsequent studies have highlighted the complexity of targeting this genetic subtype. Similarly, the Wave Life Sciences C9orf72-targeting antisense oligonucleotide program demonstrated target engagement but failed to provide clinical benefit, further emphasizing that molecular target engagement alone does not necessarily translate into functional improvement in genetically distinct ALS populations.14 In a Phase IIa trial, the PIKfyve inhibitor Apilimod reduced cerebrospinal fluid levels of poly(GP) proteins by 73% in patients with C9orf72-associated ALS, providing important evidence of central nervous system target engagement in this patient population.15 These advances indicate that genetic subtype-guided precision therapeutic strategies are offering new hope for ALS patients. In parallel, exploration of mechanism-driven therapies beyond genetic targeting has continued. For example, a Phase IIa clinical trial demonstrated the immunomodulatory effects of low-dose interleukin-2 in ALS patients, providing proof-of-concept for therapeutic strategies targeting neuroinflammation.16 Building on these developments, this review synthesizes the key barriers to clinical translation in ALS and the strategies now emerging to overcome them. Current Limitations of Clinical Management in ALS Despite decades of research, ALS pharmacotherapy remains dominated by agents with limited efficacy and uncertain clinical utility. The following three sections critically examine the current landscape by asking (Figure 1): First, how modest is the survival benefit of riluzole, and does it truly modify early disease progression? Second, does edaravone offer genuine value given its narrow therapeutic window and conflicting real-world evidence? Third, what lessons can be learned from the recent failure of PB‑TURSO, a drug that once promised breakthrough efficacy? Answering these questions is essential for guiding future therapeutic development and clinical decision‑making. Key clinical limitations of the three major approved or previously approved therapies for ALS: riluzole, edaravone, and sodium phenylbutyrate-taurursodiol (PB-TURSO). Riluzole offers only modest survival benefit (approximately 2–3 months) without substantially slowing functional decline. Edaravone is approved only for a narrow subgroup of early-stage patients, and its real-world efficacy remains controversial. PB-TURSO followed a trajectory from accelerated approval to voluntary withdrawal after its confirmatory Phase III trial failed to meet primary endpoints. Riluzole: Modest Survival Benefit Riluzole remains the only widely validated pharmacotherapy shown to extend survival in patients with ALS. The therapeutic foundation of riluzole is built upon two landmark Phase III clinical trials. The first study demonstrated that daily administration of 100 mg riluzole significantly improved 12-month survival from 58% to 74%.4 Subsequently, a dose-ranging trial further confirmed that the 100 mg daily dose offered the most favorable risk-benefit profile, reducing the risk of death by 35%.5 Nevertheless, the magnitude of this survival benefit is notably modest, corresponding to a median extension of only approximately 2–3 months. More critically, riluzole does not substantially affect the progressive decline in muscle strength,4 suggesting that it is unlikely to fundamentally reverse the disease trajectory. Mechanistically, riluzole primarily alleviates excitotoxicity by inhibiting glutamate release; although subsequent studies have indicated broader neuroprotective actions, such as modulation of voltage-gated sodium channels and enhancement of neurotrophic signaling pathways,17 its therapeutic “ceiling effect” remains evident, falling far short of achieving disease modification. In recent years, a retrospective analysis of the original dose-ranging trial data using the King’s clinical staging system has revealed a critical insight: the survival benefit of riluzole is largely confined to the final clinical stage (Stage 4), with no significant deceleration observed during the earlier stages (Stages 2 and 3) of the disease.18 This finding strongly suggests that riluzole may act primarily by slowing disease progression in the terminal phase rather than by genuinely modifying the early disease course. Furthermore, although riluzole is generally well tolerated in elderly or advanced-stage ALS patients, its survival advantage in these subgroups has not been robustly substantiated.19 In summary, while riluzole remains an indispensable cornerstone of ALS pharmacotherapy, its limited efficacy and benefit pattern restricted to late-stage disease underscore the urgent need for more effective disease-modifying therapies. Edaravone: A Narrow Therapeutic Window and Persistent Controversy The regulatory history of edaravone illustrates the challenges of identifying treatment effects in a clinically heterogeneous ALS population. An initial Phase III trial (MCI186-16) failed to meet its primary endpoint in the overall ALS population.20 However, a post-hoc analysis of this negative study unexpectedly revealed a potential treatment effect in a specific subgroup characterized by early-stage disease (onset ≤2 years), a moderate rate of functional decline (ALSFRS-R score decline of 1–4 points over a 12-week observation period), and relatively preserved pulmonary function (FVC ≥80%). Subsequently, a confirmatory trial (MCI186-19) was specifically designed to enroll this “expected-to-respond” subgroup, demonstrating that edaravone significantly slowed the decline of ALSFRS-R scores.6 Notably, further analyses have indicated that among four non-linear latent classes defined by disease progression trajectories, edaravone showed efficacy in three, suggesting differential effects depending on the patient’s progression pattern.21 Moreover, a retrospective analysis based on the King’s staging system revealed a lower proportion of patients experiencing stage progression in the edaravone group compared with the placebo group (42.0% vs 55.9%), with the most pronounced effect observed in patients at Stage 1 at baseline.22 These findings suggest that the clinical benefit of edaravone is likely contingent upon specific disease stages, progression rates, and functional status, rather than being universally effective across the entire ALS population. Despite its regulatory approval, the clinical application of edaravone remains highly contentious. First, its stringent inclusion criteria inherently exclude the majority of ALS patients, particularly those with advanced or rapidly progressive disease. Second, real-world studies have raised serious questions regarding its long-term efficacy. For instance, a German propensity score-matched cohort study found that edaravone treatment for up to 13.9 months did not confer disease-modifying benefits superior to standard care.7 Similarly, a retrospective study from Northeast Italy observed no significant advantage of edaravone compared with historical controls.23 These discordant findings suggest that the efficacy of edaravone may be restricted to a specific, narrow patient subset, thereby warranting a cautious appraisal of its broad clinical utility. On the other hand, the development of an oral suspension of edaravone offers a more convenient administration route. Studies have demonstrated that its pharmacokinetic profile is comparable to that of the intravenous formulation, with favorable safety and tolerability over 48 weeks of treatment.24,25 Nevertheless, a Phase IIIb study (MT-1186-A02) comparing once-daily dosing against the approved intermittent regimen showed that the once-daily schedule did not demonstrate statistical superiority on the primary endpoint (a combined assessment of function and survival at 48 weeks), thereby further supporting the rationale for the current intermittent regimen.26 In summary, the precise position of edaravone in ALS therapy remains to be definitively established through more rigorously designed, high-quality clinical studies. Other Approved/Exploratory Drugs The development of sodium phenylbutyrate-taurursodiol (PB-TURSO, brand name Relyvrio) illustrates the uncertainty associated with accelerated approval based on early-phase clinical evidence. Following positive Phase II results and regulatory approval, the drug was voluntarily withdrawn after the confirmatory Phase III trial failed to meet its primary and key secondary endpoints. The efficacy of this combination therapy was initially demonstrated in a Phase II clinical trial named CENTAUR. The study showed that over a 24-week treatment period, patients receiving PB-TURSO experienced a significantly slower rate of decline in the Amyotrophic Lateral Sclerosis Functional Rating Scale-Revised (ALSFRS-R) score compared with placebo recipients (−1.24 vs −1.66 points/month27). Long-term survival analyses further revealed that the original PB-TURSO group had a median survival extended by 6.9 months relative to the original placebo group (25.0 vs 18.5 months28). To correct for potential confounding of survival analyses due to crossover of placebo-group patients to PB-TURSO during the open-label extension period, the investigators applied a rank-preserving structural failure time (RPSFTM) model. The results further substantiated the survival benefit: after correction, the median survival of the placebo group was only 15.2 months, whereas PB-TURSO extended survival by 10.6 months.29 On the basis of these positive Phase II data, PB-TURSO received accelerated approval from the US. Food and Drug Administration (FDA). Moreover, biomarker analyses indicated that PB-TURSO treatment significantly reduced plasma levels of YKL-40 (by approximately 20%) and C-reactive protein (CRP; by approximately 30%), and changes in these markers correlated with ALSFRS-R scores,30 suggesting a potential anti-inflammatory mechanism of action. However, in the subsequent larger and more rigorously designed Phase III confirmatory trial (PHOENIX), PB-TURSO failed to meet its primary and key secondary endpoints, ultimately leading to the manufacturer’s voluntary withdrawal of the drug from the market. The drug was subsequently withdrawn from the Canadian market as well. This episode has once again highlighted a critical concern for the field: positive results derived from small-sample, short-term surrogate endpoints (such as ALSFRS-R slope) may not be replicable in broader populations or over longer observation periods. It profoundly underscores the critical importance of conducting large-scale, prospective confirmatory studies in ALS clinical trials, while also highlighting the necessity of adequately controlling for disease heterogeneity, natural history variability, and placebo effects. Although subsequent analyses of the CENTAUR data suggested that early randomization to PB-TURSO reduced the risk of a composite endpoint of death, tracheostomy, permanent assisted ventilation, or hospitalization by 53% compared with the placebo-to-drug group (HR = 0.47, P = 0.02), suggesting that early intervention may be particularly critical,31 it must be recognized that these post-hoc exploratory analyses are insufficient to replace confirmatory evidence from prospective validation trials. The complex trajectory of PB-TURSO demonstrates that, in a highly heterogeneous disease such as ALS, decisions regarding accelerated approval based on Phase II results warrant caution. Conducting well-designed, adequately powered, and sufficiently prolonged Phase III confirmatory trials remains an essential prerequisite for ensuring that novel therapies truly benefit patients. Core Challenges in Clinical Trial Failures The numerous disappointing negative results compel a profound re-evaluation of the current drug development paradigm for ALS. Despite considerable breakthroughs in basic research, the path to clinical translation remains remarkably arduous. The following four core challenges are widely recognized as major bottlenecks in the field (Figure 2). A deep understanding of these challenges, along with the exploration of corresponding counterstrategies, will be critical to improving the success rate of future clinical trials. Four interconnected barriers contributing to repeated Phase III failures in ALS drug development. (1) Disease heterogeneity, characterized by marked variability in genetic background, pathology, and clinical progression. (2) Lack of validated biomarkers; neurofilament light chain (NfL) is widely used but not yet accepted as a surrogate endpoint. (3) Poor translational validity of preclinical models; the classic SOD1-G93A mouse does not recapitulate sporadic ALS or TDP-43 pathology. (4) Limitations of traditional trial designs, including fixed sample sizes, insensitive endpoints (e.g, ALSFRS-R nonlinearity), and lack of adaptive elements. Disease Heterogeneity ALS is not a single disease entity but rather a clinical syndrome characterized by diverse presentations and varied underlying pathophysiological mechanisms. Specifically, this heterogeneity manifests at multiple levels. First, from a genetic perspective, more than 40 causative or risk-associated genes have been identified, of which C9orf72 hexanucleotide repeat expansions and SOD1 mutations are the most common genetic causes. Second, from a pathological standpoint, although the majority of patients exhibit TDP-43 protein pathology, the upstream triggers and downstream signaling abnormalities differ substantially across individuals. Finally, regarding clinical phenotype, disease trajectories vary enormously, ranging from slowly progressive, “benign” ALS with survival exceeding ten years, to rapidly progressive “fulminant” ALS that leads to death within months. Consequently, applying a “one-size-fits-all” therapeutic approach to such a heterogeneous disease inevitably results in diluted or even absent efficacy. For example, drugs targeting glutamate excitotoxicity may be completely ineffective in patients with rapid progression and metabolic dysregulation. Conversely, agents that modulate autophagic pathways (e.g, rapamycin) may be more effective in patients with specific genetic backgrounds, such as those carrying C9orf72 mutations.32 Similarly, a post-hoc analysis of a clinical trial on rasagiline suggested that the drug might confer survival benefits only in patients with intermediate-to-rapid disease progression, while showing no effect in those with very slow progression.33 Furthermore, the antisense oligonucleotide tofersen has demonstrated clear clinical benefits, including slowing functional decline and prolonging survival, in patients with SOD1-linked ALS, further supporting the feasibility of gene-specific therapies.34 However, in C9orf72-associated ALS, although multiple targeted therapeutic attempts (e.g, the antisense oligonucleotide BIIB078) have failed to show clinical benefit, drugs such as apilimod have exhibited target engagement activity at the biomarker level, underscoring the need for differentiated trial designs tailored to distinct genetic subtypes.12,15 In summary, future clinical trials should adopt refined patient stratification strategies based on genetic background, biomarker profiles, or clinical phenotypes to achieve “precision enrollment”, thereby validating the efficacy of targeted therapies in relatively homogeneous populations. The Shortage of Biomarkers Reliable biomarkers are of paramount importance for drug development in ALS, offering value in three core areas: (1) target engagement validation, that is, directly demonstrating that a drug acts on its intended molecular target in vivo; (2) patient stratification and enrichment, which involves identifying subgroups most likely to benefit from treatment prior to a trial, thereby enhancing statistical power and efficiency; and (3) serving as surrogate endpoints to provide accelerated pathways for regulatory approval, a particularly critical need for rapidly progressive fatal diseases. For a long time, the ALS field has been severely lacking such tools. Consequently, numerous candidate drugs have advanced prematurely into large-scale, costly Phase III trials without robust pharmacodynamic evidence, only to ultimately fail due to a lack of efficacy. This has not only consumed substantial resources but has also considerably delayed the development of effective therapies. Therefore, the development of biomarkers has become an urgent priority to overcome current R&D bottlenecks. Among the various biomarkers under investigation, neurofilament light chain (NfL) stands out as one of the most successful examples in recent years. A wealth of studies has shown that NfL levels are significantly elevated in both the cerebrospinal fluid (CSF) and blood of ALS patients, and that these levels are closely correlated with disease progression rate and patient survival. Currently, NfL is widely used as a pharmacodynamic and prognostic biomarker in clinical trials. For instance, in the Tofersen trial for patients with SOD1-ALS, a marked reduction in CSF NfL levels was observed, which served as a key piece of supportive evidence for the drug’s regulatory approval.13 Similarly, in the clinical trial of Apilimod, researchers not only observed changes in CSF poly(GP) protein (a C9orf72-associated marker) but also detected alterations in NfL levels, providing strong evidence for target engagement at the level of the central nervous system.15 Furthermore, a post-hoc analysis of the CENTAUR trial revealed that PB-TURSO treatment reduced plasma levels of YKL-40 and CRP; changes in these neuroinflammatory markers correlated with the ALSFRS-R score, suggesting their potential as treatment-sensitive inflammatory biomarkers.30 However, despite the promising prospects of NfL, it has not yet received formal regulatory endorsement as a standalone surrogate endpoint. This is largely attributable to its limitations: NfL fails to directly reflect upstream pathological mechanisms of the disease (such as aberrant protein aggregation or mitochondrial dysfunction), and its dynamic changes vary significantly across different genetic subtypes and disease stages.35 Hence, at present, NfL is better suited as an adjunctive marker for assessing disease prognosis and drug pharmacodynamics rather than as a standalone surrogate endpoint independent of clinical outcomes. Beyond fluid-based biomarkers, neurophysiological markers represent another important avenue of exploration. Such markers enable dynamic, non-invasive assessment of the functional state of the nervous system. For example, in a Phase II trial of Ezogabine (a potassium channel opener), short-interval intracortical inhibition (SICI) measured via transcranial magnetic stimulation (TMS) was successfully employed as a pharmacodynamic marker to assess the drug’s effect on reducing motor cortex hyperexcitability, thereby directly demonstrating that the drug could target and modulate the intrinsic excitability of cortical motor neurons.36 Although the development of Ezogabine was discontinued due to tolerability issues, this study paradigm clearly illustrates that neurophysiological biomarkers capable of directly reflecting target engagement are particularly valuable for early proof-of-concept trials, providing crucial mechanistic rationale for subsequent drug screening and dose selection. In summary, developing a diversified panel of biomarkers is imperative. Such a panel should comprehensively capture information from multiple dimensions, including neurodegeneration (e.g, NfL), neuroinflammation (e.g, YKL-40, CRP), target engagement (e.g, SICI, poly(GP)), and specific pathway abnormalities (e.g, metabolites). Undoubtedly, this approach represents an essential pathway to overcoming the current bottlenecks in ALS drug development. Insufficient Translational Validity of Preclinical Models The field of ALS drug development has long faced a major challenge: the vast majority of compounds that demonstrate significant survival benefits in transgenic mouse models, particularly the classic SOD1-G93A model, ultimately fail in human clinical trials. This disappointing reality underscores the severe limitations of this model in predicting clinical efficacy. First, the SOD1-G93A model recapitulates only the SOD1 mutation subtype, which accounts for approximately 2% of familial ALS cases. Its core pathogenic mechanism (primarily gain-of-toxicity of the SOD1 protein) is fundamentally distinct from that of sporadic ALS, which constitutes over 90% of all ALS cases and is pathologically characterized by TDP-43 proteinopathy. Second, the disease course in this model is extremely rapid, progressing from onset to death within just a few months, a stark contrast to the human disease course, which typically unfolds over years. Consequently, both the “therapeutic window” for drug intervention and the drug sensitivity observed in this model differ considerably from those in human patients. More critically, the model fails to fully recapitulate the core pathological features of human ALS, including TDP-43 protein pathology, axonal transport deficits, and neuroinflammation, factors that play pivotal roles in the pathogenesis and progression of sporadic ALS.37 Therefore, drug screening that relies solely on the SOD1-G93A model is inherently limited in its translational value. Beyond model limitations, inadequate central nervous system exposure represents another important translational barrier. Many compounds showing efficacy in cellular systems fail clinically because sufficient drug concentrations cannot be achieved within the CNS. This issue highlights the importance of incorporating pharmacokinetic and brain-penetration assessments into early-stage drug development to improve the predictive value of preclinical studies. In recent years, patient-derived induced pluripotent stem cell (iPSC)-derived motor neuron-based drug screening models have emerged as a promising alternative. Because these models preserve the patient’s genetic background and recapitulate key disease-relevant cellular phenotypes (e.g, neuronal hyperexcitability, TDP-43 aggregation), they exhibit superior translational potential. For example, the potassium channel opener retigabine and the dopamine agonist ropinirole, both identified through iPSC-based screening, have successfully advanced into clinical trials and demonstrated the potential to modulate neuronal excitability. Although further development of retigabine was hampered by dose-dependent tolerability issues (e.g, dizziness, fatigue), this “bedside-to-bench-to-bedside” research paradigm has illuminated a path toward developing more clinically predictive translational models.38 Furthermore, the generation of iPSC lines carrying specific ALS-associated mutations (e.g, C9orf72, TARDBP) using CRISPR gene-editing technology, along with the development of more physiologically relevant multi-cell co-culture systems incorporating motor neurons, astrocytes, and microglia, enables a more faithful recapitulation of the complex cell-cell interactions that characterize ALS pathogenesis.39,40 These advanced models hold promise for improving the predictive validity of drug screening and thereby bridging the long-standing gap between preclinical research and clinical translation. Limitations of Clinical Trial Design Traditional parallel-group, fixed-sample-size, long-term (6–12 months) follow-up randomised controlled trials (RCTs) face severe challenges in ALS, including low efficiency, high costs, and poor adaptability to disease heterogeneity. These limitations are primarily reflected in the following aspects. First, the sensitivity of current outcome measures is insufficient. The most widely used core measure, the ALS Functional Rating Scale–Revised (ALSFRS-R), exhibits non-linear changes (slow progression in early and late stages, faster progression in the middle stage), ceiling/floor effects (insensitive to mildly and severely affected patients), and suboptimal assessment of certain key functional domains (e.g, respiratory and cognitive functions). Furthermore, interpretation of ALSFRS-R changes remains complicated by ongoing debate regarding what constitutes a clinically meaningful difference. Statistical significance does not always correspond to a meaningful improvement perceived by patients, creating additional challenges for regulatory decision-making. As alternatives, clinical staging systems such as King’s stage and MiToS stage, which assess the order and number of functional losses (e.g, walking, swallowing, breathing), may provide better composite endpoints. For instance, in a post-hoc analysis of the edaravone trial, King’s stage successfully demonstrated the ability to distinguish differences in progression between treatment and placebo groups.22 Furthermore, composite measures of function and survival, such as the Combined Assessment of Function and Survival (CAFS), are increasingly being adopted to provide a more comprehensive evaluation of clinical benefit.12,41 Therefore, the development and application of more sensitive endpoints that capture multidimensional disease changes represent a critical step toward improving clinical trial efficiency. Second, conventional trial designs often overlook inter-individual variability in disease natural history. Their inclusion criteria are generally overly broad, frequently enrolling patients based on time from onset rather than biological progression rate, thereby resulting in substantial within-group variability in progression rates. This not only demands larger sample sizes to detect potential treatment effects, but also increases the risk of false-negative findings if a given therapy is only effective in a specific progression-rate subgroup. To address this limitation, enrichment designs leveraging historical controls or real-world external controls have emerged. Through rigorous matching (e.g, based on progression rate and genetic subtype), such approaches can more effectively identify patient populations likely to respond to treatment, thereby enhancing statistical power without unnecessarily expanding sample size.42 Consequently, integrating patient stratification strategies into trial design is essential for improving drug development efficiency. Third, traditional RCTs lack mechanisms for flexible adjustment. Such designs do not allow modifications to sample size, early termination of futile treatment arms, or discontinuation of doses with poor safety profiles based on interim analyses during the trial. This leads not only to substantial waste of time and resources but also to prolonged exposure of participants to ineffective or potentially harmful therapies. To address this challenge, adaptive platform trial designs have emerged and been implemented, exemplified by the HEALEY ALS Platform Trial. This design allows multiple drug arms to share a single placebo group, permits early discontinuation of futile drug arms based on pre-specified futility rules, and enables seamless addition of new drug arms into the ongoing trial, thereby greatly improving research efficiency and resource utilisation.43–45 Such innovative designs represent the future direction of ALS clinical trials and hold promise for screening promising drug candidates more rapidly and at lower cost. In summary, reforming clinical trial design by adopting more sensitive endpoints, biology-driven patient stratification strategies, and flexible adaptive platform designs is an urgent priority to accelerate the ALS drug development pipeline. Future Therapeutic Strategies Despite the considerable clinical challenges associated with ALS, recent successes exemplified by Tofersen, alongside clinical explorations of multiple innovative targets, have brought genuine promise to the field. From gene therapy to targeting protein homeostasis, and from modulating neuroinflammation to metabolic support, a diverse array of therapeutic strategies is progressively transitioning from the laboratory to the clinic, shaping a new paradigm for ALS treatment (Figure 3). Evolving therapeutic landscape for ALS, organized into six interconnected domains. Precision medicine and gene therapy, exemplified by the antisense oligonucleotide Tofersen for SOD1-ALS and the PIKfyve inhibitor Apilimod for C9orf72-ALS. Targeting protein homeostasis, including strategies such as autophagy induction (e.g, trehalose, rapamycin) and chaperone co-induction (e.g, arimoclomol). Modulation of neuroinflammation, encompassing regulatory T-cell expansion via low-dose IL-2, complement C5 inhibitors (ravulizumab, zilucoplan), and other immunomodulators. Metabolic and energetic support, including high-calorie nutrition and metabolic modulators such as trimetazidine. Neuroprotection and regeneration, involving ion channel modulation (e.g, ezogabine) and stem cell-based therapies [e.g, MSC-NTF (mesenchymal stem cells secreting neurotrophic factors)]. Combination therapies represent multi-target approaches, exemplified by the PB-TURSO concept, which requires further validation in adequately powered clinical trials. Precision Medicine and Gene Therapy At the core of precision medicine lies the development of individualized therapeutic regimens based on a patient’s specific genetic background. This concept has achieved landmark breakthroughs in ALS, marking a strategic shift from traditional symptom-directed supportive care toward mechanism-driven interventions targeting the root causes of the disease. The advent of ASO technology has made it possible to intervene in disease progression at the genetic level. The success of Tofersen, an ASO drug targeting SOD1 gene mutations, stands as a landmark event in ALS precision medicine. Delivered directly to the central nervous system via intrathecal injection, Tofersen specifically binds to SOD1 mRNA and induces its degradation, thereby reducing the production of toxic SOD1 protein at its source. In the pivotal Phase III VALOR trial, although the primary clinical endpoint was not met at 28 weeks, breakthrough progress was achieved in both biomarker responses and long-term clinical benefits. Specifically, patients who initiated Tofersen treatment early showed slower declines in clinical functional scores (ALSFRS-R change: −9.9 vs −13.5 points), better preservation of respiratory function (slow vital capacity change: −13.8% vs −18.1%), and longer survival during up to 3.5 years of follow-up, accompanied by a marked reduction in plasma neurofilament light chain (NfL) levels. These data establish Tofersen as the standard of care for SOD1-ALS.13 However, the development of ASO drugs targeting other genetic subtypes has not been straightforward. For example, BIIB078, an ASO targeting C9orf72-related ALS, showed favourable safety in a Phase I clinical trial but was discontinued due to a lack of observed NfL reduction and clinical benefit.12 These results suggest that significant differences in pathological mechanisms exist across genetic subtypes, and optimizing target selection and drug design remains a critical focus for future research. Small-molecule targeted therapy has also shown potential in C9orf72 ALS. Apilimod, an oral PIKfyve inhibitor, demonstrated not only favourable safety but also a significant 73% reduction in cerebrospinal fluid poly(GP) protein levels in C9orf72 ALS patients following 12 weeks of treatment in a Phase IIa randomized, double-blind, placebo-controlled trial, providing robust evidence of target engagement and central nervous system mechanism validation.15 This finding brings new hope to patients with C9orf72 ALS and lays the groundwork for larger-scale efficacy trials. In addition, other agents targeting different pathways are under investigation. For example, trehalose, a disaccharide capable of activating the autophagic pathway, was evaluated in the HEALEY ALS platform trial for its potential to address abnormal protein aggregation and autophagic dysfunction in ALS; however, results showed that at the dose studied, it failed to effectively slow disease progression.45 Similarly, fasudil, which aims to exert neuroprotective and axon-regenerative effects by inhibiting Rho-associated kinase (ROCK), showed favourable safety and tolerability in the Phase II ROCK-ALS trial, but its clinical effect on disease progression requires validation in larger, longer-duration trials.46 The successes and failures of these studies underscore that success in precision medicine depends on a deep understanding of the pathogenic mechanisms underlying specific genetic mutations, and the design of highly targeted intervention strategies accordingly. Targeting Protein Homeostasis The abnormal aggregation of proteins such as TDP-43 and SOD1 is not only a core pathological hallmark of ALS but also a key driver of disease progression. Consequently, strategies aimed at enhancing intracellular protein clearance mechanisms, such as the autophagy-lysosome pathway and the ubiquitin-proteasome system, have long represented an attractive direction in ALS drug development. However, despite the promise shown by these approaches in preclinical models, clinical trial outcomes have generally been disappointing, underscoring the substantial complexities involved in translating fundamental research into clinical practice. Among autophagy inducers, trehalose and rapamycin are representative candidates. Rapamycin enhances autophagy via inhibition of the mTOR pathway and also exerts immunomodulatory effects. Nevertheless, in a randomized, double-blind, placebo-controlled Phase II trial (RAP-ALS) involving 63 ALS patients, although rapamycin failed to meet the primary endpoint of modulating regulatory T cell (Treg) counts, it demonstrated a favorable safety profile and suppressed pro-inflammatory factors such as IL-18, suggesting potential value in specific patient subgroups.32 Meanwhile, another study explored the efficacy of trehalose; however, in an exploratory analysis of the HEALEY ALS platform trial, weekly intravenous administration of trehalose (0.75 g/kg) did not significantly slow disease progression.45 Although the treatment was well tolerated, this outcome again indicates that inducing autophagy alone may be insufficient to counteract the complex protein aggregation pathology in ALS, particularly in later disease stages. Beyond these strategies, promoting chaperone activity to clear misfolded proteins represents another important direction. Arimoclomol, a co-inducer of heat shock protein 70 (HSP70), aims to facilitate the clearance of misfolded proteins by amplifying the cellular stress response. However, in the Phase III ORARIALS-01 trial encompassing 239 patients with early-stage ALS, arimoclomol failed to demonstrate any clinical benefit.47 Furthermore, exploration of other protein homeostasis modulators has also yielded disappointing results. For example, in a Phase II clinical trial, curcumin, an antioxidant with neuroprotective properties, showed only a trend toward modestly slowing disease progression without achieving significant efficacy.48 In another open-label trial involving patients with SOD1 mutations, although the antimalarial drug pyrimethamine significantly reduced SOD1 protein levels in the cerebrospinal fluid, the study did not evaluate its long-term impact on disease progression.49 These failures highlight the complexity of therapeutic strategies targeting protein homeostasis. It is worth noting that the pathogenic mechanisms of ALS-associated proteins extend far beyond simple aggregate formation. Taking TDP-43 as an example, in addition to forming toxic aggregates, it also undergoes pathological changes involving both gain- and loss-of-function, including impaired nucleocytoplasmic shuttling and loss of splicing function. Therefore, strategies solely targeting aggregates are likely to prove insufficient. Moreover, the failure of clinical trials also suggests that intervention in the mid-to-late disease stages may be too late, as irreversible neuronal damage may have already occurred. In summary, although targeting protein homeostasis remains theoretically highly promising, the negative outcomes of multiple clinical trials indicate that future research should focus on exploring more effective combination regimens and consider earlier intervention to overcome the current barriers to clinical translation. Modulation of Neuroinflammation Chronic neuroinflammation mediated by microglia and astrocytes has been established as a key pathogenic driver in both the onset and progression of ALS. Nevertheless, the developmental trajectory from early broad-spectrum anti-inflammatory agents (e.g, celecoxib, minocycline) to more selective targeted therapies, despite being fraught with difficulties, continues to reveal considerable therapeutic promise. In the realm of targeting specific immune pathways, the modulation of regulatory T cells (Tregs) has emerged as a core strategy for controlling neuroinflammation in ALS. A representative approach involves the expansion of Tregs using low-dose interleukin-2 (ld-IL-2) to suppress excessive inflammatory responses. The IMODALS Phase IIa trial demonstrated that ld-IL-2 effectively increases the number of Tregs in the peripheral blood of ALS patients and reduces the levels of the chemokine CCL2 in a dose-dependent manner.16 Subsequent pharmacokinetic studies further revealed the modulatory effects of ld-IL-2 on key metabolic pathways, such as the kynurenine pathway, providing a crucial theoretical basis for future personalized medicine.50 Moreover, a combination strategy of Treg infusion plus ld-IL-2 has shown favourable safety and bioactivity in a Phase I study, with six slowly progressing patients maintaining relatively stable functional status over one year of treatment.51 Beyond the Treg pathway, the anti-CD40L monoclonal antibody Tegoprubart, evaluated in a Phase IIa open-label dose-escalation study involving 54 ALS patients, also demonstrated a favourable safety profile and a dose-dependent ability to downregulate anti-inflammatory biomarkers, including 18 pro-inflammatory chemokines and cytokines.52 Concurrently, progress has been made in exploring other immunological targets. For instance, a randomized double-blind placebo-controlled trial of rapamycin, while not meeting its primary endpoint of increasing Treg proportions, observed a reduction in the pro-inflammatory cytokine IL-18 and demonstrated favourable safety and tolerability, thereby laying a foundation for future research.32 These studies indicate that precise modulation of specific immune signaling nodes represents a highly promising interventional direction; however, larger-scale trials are required to validate clinical benefits. In stark contrast to the above progress, clinical translation efforts targeting the complement system have encountered major setbacks. Although complement system activation has been observed in both ALS animal models and patients, implicating it in neuroinflammatory processes, interventions targeting the terminal pathway have so far failed to translate into clinical benefit. Specifically, two large-scale Phase III trials targeting complement C5 were prematurely terminated due to lack of efficacy: the Phase III trial of the anti-C5 monoclonal antibody Ravulizumab, which enrolled 382 patients, observed no functional improvement after 50 weeks of treatment;53 similarly, another Phase III trial evaluating the C5 inhibitor Zilucoplan failed to alter disease progression.44 These disappointing negative results suggest that although complement activation is an important component of neuroinflammation in ALS, simply targeting the terminal C5 component may not represent an effective therapeutic strategy. The reason may be that C5 acts relatively downstream in the complement cascade, and inhibiting its function may be insufficient to block more upstream and broader inflammatory signalling pathways. Therefore, future research efforts should perhaps shift towards upstream regulators of the complement system, such as C1q, which plays a critical role in initiating the classical complement activation pathway and may represent a more promising therapeutic target.54 In summary, the therapeutic prospects of complement inhibition in ALS may need to be re-evaluated through more precise targeting of upstream components or modulation of the broader inflammatory network. Regarding targeted innate immune modulation, a post-hoc analysis of NP001 (an immunomodulator that converts to chlorite-based oxidant in vivo) revealed that it may have a positive effect on stabilising pulmonary function in patients with elevated baseline C-reactive protein (CRP) levels, i.e, a subgroup with evidence of systemic inflammation.55 This finding is consistent with previous studies showing that elevated serum CRP levels are strongly associated with faster functional decline and poorer prognosis in ALS patients, suggesting that CRP could serve as a valid biomarker of neuroinflammatory status and disease progression.56 Consequently, targeted innate immune pathway interventions show potential clinical value in the inflammatory subgroup characterized by elevated CRP, despite the lack of significant efficacy in broader, unselected patient populations.55 This once again underscores the critical importance of biomarker-based subtype stratification in ALS clinical trials, precisely identifying patient subgroups with specific inflammatory signatures may be key to the future clinical success of immunomodulatory therapies. Metabolic and Energy Support Metabolic disturbances, particularly hypermetabolism characterized by weight loss and elevated resting energy expenditure, have been established as independent risk factors for poor prognosis in ALS. Consequently, targeted metabolic support strategies are increasingly being recognized as an important avenue of adjunctive therapy. However, clinical studies on high-calorie nutritional interventions have yielded complex results, suggesting that the benefits of such approaches may be highly contingent on patient selection. A randomized controlled trial (RCT) involving patients with pre-existing respiratory insufficiency who underwent percutaneous endoscopic gastrostomy (PEG) demonstrated that a high-carbohydrate, high-calorie diet reduced adverse events and potentially extended survival compared with an isocaloric control group.57 In contrast, the LIPCAL-ALS trial, conducted in 201 patients at an earlier disease stage, failed to confirm a survival benefit of a high-fat, high-calorie diet in the overall population, although potential benefits were observed in the fast-progressing subgroup.58 Therefore, these seemingly contradictory findings underscore that the precise identification of patient subgroups with significantly increased energy requirements, along with careful consideration of disease stage, is a critical prerequisite for the success of future metabolic intervention strategies. Beyond simply increasing caloric intake, regulatory strategies targeting specific metabolic pathways have shown emerging promise. For instance, trimetazidine, a partial inhibitor of fatty acid oxidation, not only significantly reduced levels of the oxidative stress marker 8-hydroxy-2’-deoxyguanosine (8-OHdG) and lowered resting energy expenditure but also demonstrated a favorable safety and tolerability profile in an open-label Phase IIa study involving 21 ALS patients.59 This finding offers a novel approach for ameliorating metabolic abnormalities in ALS, namely, optimizing the body’s energy metabolic status by modulating substrate utilization efficiency rather than merely increasing caloric supply. Furthermore, other interventions targeting mitochondrial dysfunction and specific metabolic pathways, such as elevating uric acid levels60 or modulating particular signaling pathways,61 are under investigation; however, their clinical benefits await validation in larger-scale studies. In summary, future metabolic support strategies should evolve toward greater individualization and precision, integrating caloric supplementation with pathway modulation based on the distinct metabolic profiles and disease stages of different patient populations, thereby maximizing therapeutic efficacy. Neuroprotection and Regeneration Modulating neuronal excitability has historically been one of the conventional therapeutic strategies for ALS. The concept of motor neuron “exhaustion” due to overexcitation, i.e, excitotoxicity, remains a core hypothesis in ALS, grounded in the notion that glutamate-mediated hyperexcitability contributes to neuronal damage. Nevertheless, most early clinical trials targeting NMDA receptors (e.g, memantine, dextromethorphan) or AMPA receptors (e.g, perampanel, talampanel) either failed or demonstrated only marginal benefits.4,62,63 For instance, although perampanel at the 8 mg dose resulted in a significant decline in ALSFRS-R scores, largely attributable to bulbar function deterioration, poor overall tolerability and frequent adverse events ultimately led to premature trial termination.62,63 However, efforts in this area have not ceased. Ezogabine, a Kv7 potassium channel opener, successfully reduced cortical and spinal motor neuron hyperexcitability in a Phase II randomized clinical trial, thereby validating the feasibility of “calming” overactive neurons.36 Similarly, mexiletine, a sodium channel blocker, has shown modulatory effects on axonal hyperexcitability in Phase II studies,64 although its impact on clinical functional improvement awaits further validation. These findings underscore that precisely targeting specific ion channels to restore neuronal electrical homeostasis remains a highly promising interventional direction. Concurrently, despite considerable controversy and setbacks, stem cell-based therapies continue to be actively explored. Early proof-of-concept studies confirmed the technical feasibility of intrathecal or intraspinal injection of mesenchymal stem cells (MSCs) or neural stem cells (NSCs).65,66 However, a Phase III study involving 196 patients reported that MSC-NTF (MSCs secreting high levels of neurotrophic factors) did not meet the primary endpoint in the overall population. Notably, a promising trend of clinical response was observed in patients with milder baseline disease severity (ALSFRS-R ≥ 35).67 More importantly, subsequent biomarker analyses confirmed the multimodal anti-inflammatory and neuroprotective effects of MSC-NTF, including a significant reduction in neurofilament light chain (NfL) levels and modulation of multiple inflammatory factors.68 Furthermore, a Phase II clinical trial using Muse cells, a stress-tolerant pluripotent stem cell population, reported favorable safety profiles and preliminary signals of efficacy.69 Together, these findings suggest that deploying stem cell therapy primarily as a “neurotrophic factor delivery system” rather than a direct cell replacement strategy may be more clinically feasible. Accordingly, future research should focus on optimizing cell types, routes of administration, and patient selection criteria to fully harness their potential in modulating the disease microenvironment. In summary, although interventions targeting single pathways have repeatedly encountered obstacles, multifaceted strategies aimed at repairing compromised neural circuitry, through ion channel modulation and cell-mediated neuroprotection, remain a source of hope for delivering clinical benefits to ALS patients. Combination Therapies Given the extreme complexity of the pathophysiology of ALS, which involves interconnected and potentially synergistic mechanisms such as genetic defects, protein aggregation, neuroinflammation, metabolic dysregulation, and excitotoxicity, traditional single-target therapeutic development approaches have encountered repeated setbacks. Numerous single-target agents that showed promise in preclinical models, such as arimoclomol targeting protein aggregation,47 zilucoplan44 and ravulizumab53 targeting inflammatory responses, and reldesemtiv targeting specific ion channels,41 have failed to demonstrate expected clinical benefits in large-scale trials. The concept of combination therapy itself is not new and has been proposed in ALS for decades; however, implementing biologically rational combinations remains challenging because optimal target selection, timing, and patient stratification are still unresolved.70 Given these repeated setbacks, future therapeutic strategies may need to shift toward multi-target, multi-mechanism synergistic approaches rather than relying solely on single-pathway interventions. An ideal combination regimen should be conceived as a systematic approach aimed at comprehensively arresting disease progression. Its core components may include: (1) anti-inflammatory/immunomodulatory agents, such as low-dose interleukin-216,50 or cell therapies like MSC-NTF,67 to quell the chronic neuroinflammation that drives disease progression; (2) anti-excitotoxicity/antioxidant agents, with riluzole4 and edaravone6 serving as foundational therapies to protect motor neurons from hyperexcitation and oxidative damage; (3) metabolic support, via high-caloric nutrition or metabolic modulators (e.g, trimetazidine59) to correct the common state of hypermetabolism and energy depletion in patients;58 and (4) targeted correction of genetic defects, using ASO drugs (e.g, tofersen for SOD1 mutations13) to block pathogenic processes at their source in patients carrying specific genetic mutations. This “multi-pronged” strategy is expected to produce synergistic or additive therapeutic effects by simultaneously acting on multiple nodes of the disease network. The clinical feasibility of combination therapy has received preliminary validation, with the most notable example being the sodium phenylbutyrate-tauroursodeoxycholic acid (PB-TURSO) combination used in the CENTAUR trial. This agent is believed to possess anti-apoptotic, anti-inflammatory, and chemical chaperone functions. Although the subsequent confirmatory Phase III trial did not meet its primary endpoint, the CENTAUR trial itself preliminarily confirmed the potential of multi-target strategies. During the 24-week treatment period of that trial, the PB-TURSO group showed a 25% slower decline in ALSFRS-R scores compared with the placebo group; more importantly, long-term follow-up analysis revealed that patients receiving active treatment had a median survival extended by 6.5 months compared with placebo recipients (25 months vs 18.5 months; HR = 0.5628). Further biomarker analyses demonstrated significant reductions in YKL-40 and C-reactive protein (CRP) levels in the PB-TURSO group, directly supporting its anti-inflammatory mechanism of action.30 These data suggest that rational multi-target combinations may provide biological and clinical signals of benefit, although their efficacy requires confirmation in adequately powered clinical trials. Looking ahead, the success of combination therapies will be highly dependent on precise pathophysiological subtyping of patients and, based on such subtyping, the design of individualized “cocktail” regimens. Given the profound heterogeneity of ALS, different patients may exhibit markedly different sensitivities to distinct therapeutic mechanisms. For instance, high-caloric diets that benefit only patients in rapidly progressing phases58 or the anti-CD14 monoclonal antibody IC14 that is effective only in specific subgroups71 underscore the importance of precise subtyping. Consequently, future clinical trial designs may need to shift from a “one-size-fits-all” model toward biomarker- and genetic-background-driven adaptive designs, aiming to deliver the appropriate combination to the appropriate patient population at the optimal time. In summary, although the treatment of ALS remains enormously challenging, diversified explorations ranging from gene-targeted therapies to multi-target combination strategies are gradually unveiling new prospects for overcoming this refractory disease. Future success will no longer rely on a single “wonder drug”, but rather on a comprehensive, individualized therapeutic system built upon a deep understanding of disease heterogeneity and precise intervention. Innovative Clinical Trial Designs The development of therapeutics for ALS has long been constrained by the inherent limitations of traditional randomized controlled trials (RCTs), including substantial disease heterogeneity, difficulties in recruiting patients with a rare disease, poor sensitivity of outcome measures to intervention effects, and prohibitive time and financial costs (Figure 4). In response to these challenges, the field is actively pivoting toward a series of innovative trial designs aimed at improving developmental efficiency, reducing costs, and maximizing the information gleaned from every patient’s data. The core philosophy underpinning these approaches is a shift away from a “one-size-fits-all”, static assessment model toward more adaptive, data-driven, and patient-centric frameworks. Comparison of traditional randomized controlled trials (RCTs) with four innovative design paradigms that are reshaping ALS clinical research. Adaptive platform trials, exemplified by the HEALEY ALS Platform Trial, which allows simultaneous evaluation of multiple regimens sharing a common placebo arm, with pre-planned futility analyses. Enrichment designs and external controls, using genetic or biomarker-based patient selection and real-world data to reduce placebo group size. Sequential designs and N-of-1 trials, enabling interim analyses and individualized crossover assessments for symptom management. Virtual and decentralized trials, incorporating telemedicine, wearable devices, and digital biomarkers to reduce patient burden and enhance data objectivity. Platform Trials Conventional RCTs typically test only one drug at a time, rendering the process inefficient. In stark contrast, adaptive platform trials represent a paradigm shift by employing a single master protocol to simultaneously evaluate multiple investigational agents (or “regimens”) against a shared placebo arm within the same infrastructure. The HEALEY ALS Platform Trial stands as a landmark exemplar in the field.43–45 This platform permits new protocols to be added at any time, whereas completed regimens can be terminated independently. Each regimen incorporates pre-planned futility interim analyses; consequently, should a drug demonstrate inefficacy or an unfavorable benefit-risk profile, enrollment and treatment for that specific arm can be halted immediately without disrupting the ongoing platform. Such a design not only substantially shortens trial timelines and reduces patient exposure to ineffective or potentially harmful placebo treatment but also significantly lowers overall developmental costs through resource sharing. Accordingly, platform trials offer a highly efficient, flexible, and sustainable drug evaluation ecosystem for a heterogeneous disease with few effective therapies, such as ALS. Enrichment Designs and External Controls In ALS, patients with distinct genetic subtypes, sites of disease onset, and rates of progression may exhibit markedly divergent responses to the same therapeutic intervention. Patient enrichment based on specific biomarkers or clinical characteristics is, therefore, a pivotal strategy to enhance trial sensitivity and facilitate the detection of efficacious drugs. For example, the clinical trial of the antisense oligonucleotide Tofersen, which targets SOD1 gene mutations, successfully validated its efficacy precisely through the targeted enrollment of patients with SOD1-ALS, a specific subpopulation.13 Similarly, the regulatory approval of Edaravone underscores the value of an enrichment strategy, as its success hinged on a post-hoc analysis of a “clinically expected-to-benefit” subgroup (i.e, individuals with early-stage disease and well-defined functional decline) from prior trials, followed by prospective validation.6 Beyond prospective enrichment, leveraging high-quality real-world data (e.g, the PRO-ACT database) to construct external controls represents another powerful acceleration strategy. Through rigorous statistical matching, comparing a trial’s treatment arm with an external control can reduce the required sample size for the placebo group, thereby expediting the trial process. Sequential Designs and N-of-1 Trials For rare diseases such as ALS, patient resources are particularly precious. Sequential designs allow for multiple interim analyses during the course of a trial, with pre-specified stopping rules that enable early termination once sufficient evidence of efficacy or futility has accumulated. Compared with conventional fixed-sample designs, group-sequential designs allow prespecified interim analyses and may permit early termination for efficacy or futility, thereby improving trial efficiency and reducing unnecessary use of limited patient resources.72 Furthermore, the N-of-1 crossover trial represents an extreme form of individualized assessment, particularly suitable for evaluating the efficacy of symptom-management drugs. In this design, individual patients receive multiple cycles of active treatment and placebo in a randomized sequence, effectively serving as their own controls. For instance, a study evaluating the efficacy of the traditional Chinese medicine TJ-68 (Shakuyaku-Kanzo-To) on muscle cramps in ALS successfully employed an N-of-1 design, detecting significant improvements across multiple symptom dimensions in a small cohort of only 11 patients.73 This approach provides a powerful tool for exploring individualized treatment responses in the highly heterogeneous ALS patient population. Virtual Clinical Trials and Remote Visits The COVID-19 pandemic has greatly accelerated the adoption of decentralised clinical trials (DCTs). The COURAGE-ALS trial, for example, employed a hybrid model in which approximately half of the study visits were conducted via telemedicine, substantially reducing the participation burden on ALS patients with limited mobility and their families while maintaining data quality.41 Going a step further, the integration of wearable devices (e.g, smartwatches, activity trackers) for continuous, passive data collection (such as step counts, activity levels, sleep quality, and heart rate variability) can provide more objective, high-frequency, real-world functional outcome measures compared with conventional periodic in-clinic assessments.60 A trial investigating inosine-elevated urate for the treatment of ALS (SURE-ALS2) has successfully explored the feasibility of remote functional assessment using a smartphone application, demonstrating good correlations with in-person evaluations.60 The incorporation of such “digital biomarkers” not only significantly reduces the burden of travel to clinical sites but also enhances data objectivity and statistical power, thereby transforming the operational paradigm of future ALS clinical trials. These innovative designs are reshaping the operational framework of ALS clinical trials by improving accessibility, efficiency, and patient-centered data collection. Patient advocacy groups have also increasingly shaped the ALS research landscape by facilitating research funding, supporting patient registries, and promoting patient-centered approaches in clinical trial development. However, balancing the urgent need for new therapies with rigorous evaluation of efficacy and safety remains an important challenge. Conclusion and Future Perspectives The clinical management of ALS stands at a historic inflection point. Over the past three decades, the field has transitioned from the “riluzole era” into an extended “era of negative results”, during which we have gained profound insights into the core challenges of this disease, including its remarkable heterogeneity, biomarker scarcity, preclinical model limitations, and trial design inefficiency. These challenges have acted like icebergs, sinking a multitude of promising drug candidates. From the early riluzole trial targeting glutamatergic excitotoxicity,4 to the subsequent series of disappointing Phase III studies, such as those of topiramate,74 celecoxib,75 dexpramipexole,76 and more recently, verdiperstat43 and reldesemtiv,41 these negative results have not only failed to translate into clinical benefit but have also laid bare the superficiality of our understanding of disease mechanisms and the inherent limitations of existing clinical trial designs. Nevertheless, failure has been a fertile ground for progress. Each negative outcome has provided invaluable lessons and catalyzed paradigm shifts in research. The success of tofersen in SOD1-ALS represents not only a triumph of precision medicine but also a powerful endorsement of the “target-specific pathogenic genes” strategy across the entire field. Indeed, the latest long-term follow-up data reveal that patients who initiated tofersen treatment early experienced slower clinical functional decline over 148 weeks and showed significantly prolonged survival.13 Similarly, although the antisense oligonucleotide BIIB078, which targets C9orf72-associated ALS, did not demonstrate clinical benefit in a Phase I study, its safety data have provided critical insights for the design of subsequent therapies targeting C9orf72.12 These advances demonstrate identifying the right target and deploying appropriate tools such as ASOs can alter the natural history of the disease in specific genetic subtypes of ALS. Looking ahead, the therapeutic landscape for ALS is poised to evolve along five major interrelated trends, which collectively drive the field away from conventional paradigms and toward a more efficient, precise, and patient-centered future. First, clinical trial design will undergo a fundamental transformation, embracing platform trials, adaptive designs, enrichment strategies, and pragmatic clinical trials. By incorporating shared placebo arms, dynamically adjustable protocols, and external control arms, these innovative approaches will enable faster, lower-cost evaluation of multiple candidate drugs. N-of-1 trials will provide individualized evidence for patients with highly heterogeneous disease manifestations. Second, the integration and validation of biomarkers have become urgently needed. A rigorously validated multimodal biomarker panel is essential for patient stratification, target engagement monitoring, and early efficacy signal detection. Neurofilament light chain (NfL) has approached this goal, yet standardization and prospective validation remain required. Imaging and multi-omics biomarkers will further complement this picture. Third, therapeutic strategies are shifting from single-target to combination or multi-target paradigms. Given the complex pathological network of ALS, combination therapies or multi-target drugs that simultaneously address key nodes such as excitotoxicity, protein aggregation, and neuroinflammation are becoming imperative. The case of PB-TURSO (despite subsequent controversies) provides important lessons in this regard. Future clinical trials should add novel mechanism-based agents on top of standard care (riluzole/edaravone). Fourth, the deepening and dissemination of precision medicine are imperative. Every newly diagnosed ALS patient should undergo comprehensive genetic testing to enable early identification of hereditary ALS cases suitable for ASO-based therapies (e.g, Tofersen) or enrollment in relevant clinical trials. At the same time, molecular subtyping of sporadic ALS based on non-genetic factors, such as metabolic phenotype, inflammatory status, and protein aggregation type, will be necessary to achieve individualized treatment. However, realizing this model at scale will not be straightforward: genetic testing and ASO therapies remain costly, intrathecal administration requires specialized clinical infrastructure, and access to these treatments may not be equitable across patient populations. Fifth, patient-reported outcomes and palliative care are gaining prominence. While pursuing disease-modifying therapies, equal emphasis must be placed on improving patients’ daily quality of life. Thus, integrating palliative care, psychological support, advanced rehabilitation technologies, and augmentative and alternative communication devices into holistic, longitudinal, multidisciplinary management is as important as drug development. Although the road ahead remains long, the ice has begun to break. With the continuous emergence of gene editing, cell therapy, next-generation biomarkers, and intelligent clinical trial platforms, there is good reason to believe that in the next decade, ALS will transition from an “incurable disease” to a “treatable and manageable” chronic condition. Conquering ALS is a long and difficult journey, but as long as we keep moving forward, success will eventually come. Acknowledgments All figures in this study were generated using BioRender. Funding Statement This work was funded by the National Natural Science Foundation of China (Nos. 82401633, 82072160), the Natural Science Research Projects in Universities of Jiangsu Province (No. 24KJA310007, No. 24KJB310013), and Nantong Science and Technology Program (No. 20241118084). Author Contributions All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure The authors report no competing interests. References