Spinal muscular atrophy (SMA; 5q-SMA) is an autosomal recessive neuromuscular disorder caused by deficiency of survival motor neuron (SMN) protein resulting from homozygous deletions or mutations in the SMN1 gene, with an incidence of approximately 1/10,000 births [1, 2]. Although the SMN gene is ubiquitously expressed and the SMN protein is required for fundamental cellular processes such as RNA splicing and ribonucleoprotein assembly, motor neurons of the anterior horn are particularly sensitive to its depletion [3, 4]. The resulting degeneration of motor neurons leads to progressive, symmetric muscle weakness and atrophy. The natural history and clinical presentation of SMA span a spectrum from type 0 to 4 based on age of onset and achievement of motor milestones, with severity inversely correlated with SMN2 copy number (Fig. 1) [1, 5, 6]. However, in the era of disease-modifying therapies for SMA, clinical classification is increasingly based on SMN2 copy number, as it is the strongest predictor of disease severity and treatment responses, while the historical SMA types are primarily relevant to the natural history of untreated disease [7].

Relationship between SMN2 Copy number, SMA Severity and clinical phenotype

Across all copy numbers and clinical phenotypes (Fig. 1), untreated patients with SMA develop progressive muscle weakness, respiratory failure, and contractures, most prominently in the severe infantile-onset forms (type 0 and type 1) [8]. As, historically, the leading genetic cause of infant mortality, SMA prognosis has changed dramatically with recognition of SMN2 as a disease modifier and the widespread approval of disease modifying therapies since 2016, beginning with the approval of nusinersen [9,10,11,12]. Early treatment, particularly in presymptomatic infants, enables near-normal motor development, which underscores the importance of newborn screening and rapid initiation of therapy [7, 13].

SMA results primarily from homozygous deletion or mutation of SMN1, leading to absent SMN protein production and degeneration of anterior horn motor neurons [14]. Nearly all affected individuals have SMN1 exon 7 deletion or compound heterozygous pathogenic variants. Disease severity, however, is determined mainly by the backup SMN2 copy number (Fig. 1) and its ability to produce small amounts of full-length SMN protein [15]. Because SMN2 undergoes inefficient exon 7 inclusion, only ~ 10–25% of transcripts generate functional SMN [3, 4, 16]. Therapies that promote exon 7 inclusion in SMN2 enhance full-length SMN production and significantly improve motor and survival outcomes [17, 18].

Beyond its motor neuron function, SMN is ubiquitously expressed throughout the body and is essential for normal development and maintenance of peripheral tissues, including skeletal muscle, the heart, and the autonomic nervous system. In skeletal muscle, SMN deficiency causes intrinsic myopathic changes independent of denervation, contributing directly to weakness and atrophy [19, 20]. In cardiac tissue, it has been linked to structural abnormalities (e.g., septal and outflow-tract defects) and arrhythmias, especially in severe SMA [21, 22]. These findings indicate a direct requirement for SMN in cardiac development and cardiomyocyte RNA regulation and signaling. In the autonomic nervous system, SMN loss can produce cardiovascular and gastrointestinal autonomic dysfunction in both patients and animal models [21, 23].

Early recognition of SMA has become a cornerstone of disease management. Because irreversible motor neuron loss begins before symptom onset, identification during the presymptomatic phase is essential to achieve optimal outcomes with disease-modifying therapies [7, 13]. Newborn screening bridges molecular genetics and clinical care, enabling rapid diagnosis and treatment in infants who would otherwise present only after clinical symptom onset [24, 25]. Newborn screening for SMA typically uses a tiered molecular approach centered on detecting SMN1 exon 7 homozygous deletion, which accounts for > 95% of cases [26]. The recommended workflow integrates high-throughput deletion testing, reflex SMN2 copy-number analysis, and confirmatory testing [27]. Because standard assays detect only homozygous deletions, < 5% of patients with a deletion on one allele and point mutation on the other (compound heterozygote) may be missed. Infants with persistent clinical suspicion after a negative screen require SMN1 sequencing or broader genomic testing [28]. Since its addition to the U.S. Recommended Uniform Screening Panel (RUSP) in 2018, SMA newborn screening has expanded rapidly. By 2025, more than 40 countries, including the United States, Germany, Austria, Japan, Taiwan, and Australia, have adopted or piloted national screening programs [29, 30]. Despite this progress, significant global disparities remain due to limited resources, infrastructure gaps, and policy barriers in many low- and middle-income countries.

Extensive clinical trial and population-based evidence shows that presymptomatic treatment fundamentally transforms outcomes in SMA. Infants treated before symptom onset routinely achieve age-appropriate motor milestones, such as sitting and walking, and rarely require ventilatory or nutritional support, outcomes not seen in untreated or postsymptomatic patients [7, 31,32,33].

Therapeutic development has transformed the natural history of the disease, shifting SMA from a progressive and often fatal disorder to one in which survival and motor function can be substantially improved. All currently approved disease-modifying therapies increase levels of functional survival motor neuron (SMN) protein, either by modifying splicing of the paralogous SMN2 gene in the case of nusinersen and risdiplam or by replacing the defective SMN1 gene in the case of onasemnogene abeparvovec. While these approaches share a common biological target, they differ markedly in mechanism, route of administration, durability of effect, and safety profile. Treatment efficacy is strongly influenced by disease severity and timing of initiation, with the greatest benefits observed in presymptomatic or early symptomatic patients identified through newborn screening. Nevertheless, meaningful stabilization or improvement can be achieved in later-onset disease, particularly in patients with less advanced disability.

Nusinersen is an antisense oligonucleotide that modifies SMN2 pre-messenger RNA splicing to promote exon 7 inclusion, thereby increasing production of functional SMN protein. Approved by the FDA in 2016, it was the first therapy to demonstrate that targeted RNA splicing modulation could alter the course of a previously fatal neurogenetic disorder [34]. It is administered intrathecally and requires lifelong maintenance dosing every 4 months.

Clinical trials and real-world studies consistently show improved survival, motor function, and motor milestone attainment across SMA subtypes. In infantile-onset SMA (type 1), the ENDEAR trial demonstrated a 47% reduction in the risk of death or permanent ventilation, with motor milestone responses observed in 51% of treated infants compared with none in controls [35]. In later-onset SMA (types 2 and 3), the CHERISH trial showed clinically meaningful motor improvements in 57% of treated patients at 15 months [36]. Long-term extension and real-world data support sustained benefit, with continued gains in several key motor outcomes over extended treatment periods [37,38,39].

Treatment benefit is greatest with early initiation. In the NURTURE study of presymptomatic infants identified through newborn screening, all participants achieved independent sitting and most attained ambulation, outcomes markedly superior to the natural history of untreated SMA [32, 40]. Among symptomatic patients, earlier treatment, particularly within two years of symptom onset, is associated with greater motor gains, while those with advanced disability, contractures, or skeletal deformities demonstrate more limited functional improvement [36, 41].

Despite clear motor benefits, improvement in respiratory and bulbar function is less predictable, and some patients continue to require ventilatory or nutritional support [42]. In adolescents and adults, objective functional outcomes are variable, though patient-reported benefits are commonly reported, especially in those with milder disease [37,38,39].

Nusinersen is administered as four loading doses over two months followed by maintenance dosing every four months [43]. Adverse events are generally manageable and include headache, vomiting, back pain, pyrexia, and post–lumbar puncture syndrome. Less common but serious risks include thrombocytopenia, proteinuria, and rare communicating hydrocephalus, necessitating routine laboratory monitoring [41, 44, 45]. Intrathecal administration may be challenging in patients with scoliosis or prior spinal fusion, often requiring image-guided techniques or a transition to risdiplam and becomes more challenging as patients get older and weaker [46,47,48].

Risdiplam, approved by the FDA in 2020, is the first oral therapy for SMA and offers a non-invasive alternative to intrathecal treatment [49]. It is a systemically distributed small molecule that acts similarly to nusinersen, modifying SMN2 pre-mRNA splicing to increase functional SMN protein in central and peripheral tissues. Risdiplam is taken orally once daily and requires lifelong treatment.

Clinical studies demonstrate that risdiplam improves or stabilizes motor function across SMA subtypes. In infantile-onset SMA (type 1), the FIREFISH trial showed that treated infants achieved motor milestones not seen in natural history, with significantly improved function, and 85% surviving without permanent ventilation [35]. In later-onset SMA (types 2 and non-ambulant 3), the SUNFISH trial demonstrated modest but significant gains at 12 months, and long-term data indicate sustained improvements over 24 months [50, 51]. Benefits are greatest with early initiation; the RAINBOWFISH study of presymptomatic infants showed outcomes comparable to other presymptomatic approaches, while symptomatic patients with advanced disability or contractures experience more limited functional gains [50, 52].

Long-term data and real-world experience indicate that risdiplam provides sustained benefit across age groups. Comparative effectiveness versus nusinersen is not well established, as no head-to-head trials exist; indirect comparisons suggest broadly similar efficacy in motor outcomes, with differences largely related to delivery method and patient characteristics [53, 54].

Risdiplam is generally well tolerated. Common adverse events include pyrexia, diarrhea, rash, and upper respiratory infections, while serious events are uncommon [50]. Oral administration avoids procedural risks, though systemic exposure necessitates ongoing safety monitoring [49].

Nonclinical animal studies demonstrate potential reproductive risks that warrant counseling for both female and male patients of reproductive potential [55]. In rats and rabbits, risdiplam administration during pregnancy resulted in embryofetal mortality, fetal malformations (including hydrocephaly), decreased fetal body weights, and reproductive impairment in offspring at clinically relevant exposures [56]. In particular, male fertility may be theoretically compromised during treatment, as animal studies demonstrated adverse effects on reproductive organs and germ cells [55]. To date, there is no clinical evidence suggesting that risdiplam causes male fertility issues in humans [51] and there are cases of risdiplam-treated males fathering healthy children [57].

Onasemnogene abeparvovec is an adeno-associated virus serotype 9 (AAV9)-mediated gene replacement therapy that delivers a functional copy of SMN1 to motor neurons, directly addressing the genetic cause of SMA. Approved by the FDA in 2019, it was the first one-time, systemically delivered gene replacement therapy for a genetic disorder [58, 59]. The therapy is available as an intravenous formulation for young children under 2 years and very recently as of late 2025, an intrathecal formulation for older patients (2 to 18 years old) [60].

Clinical studies show substantial improvements in survival and motor milestone attainment across SMA subtypes, particularly when treatment is initiated early [11, 59]. Presymptomatic treatment can result in motor development approaching age-appropriate expectations. In infantile-onset SMA, the STR1VE trial demonstrated that 59% of treated infants achieved independent sitting at 18 months compared with 0% of untreated controls and 91% survived free from permanent ventilation at 14 months compared with 26% of controls [11]. Long-term follow-up data demonstrate durability of response for years, with maintained motor milestones and no patients requiring new mechanical respiratory support during follow-up [61]. In the STEER trial of treatment-naive patients aged 2 to less than 18 years, intrathecal onasemnogene abeparvovec demonstrated significant albeit more modest benefit [60]. The STRENGTH trial of treatment-experienced patients who had previously discontinued nusinersen or risdiplam showed that intrathecal onasemnogene abeparvovec also demonstrated an acceptable safety and tolerability profile [62].

Treatment benefit is greatest with early initiation. In the SPR1NT trial of presymptomatic infants treated within 6 weeks of life, all 14 infants with two SMN2 copies (expected to develop SMA type 1) achieved independent sitting. Additionally all survived without permanent ventilation, and none required nutritional or respiratory support—outcomes never observed in the natural history of untreated SMA type 1 [63]. Among 15 presymptomatic infants with three SMN2 copies (expected to develop SMA type 2), all stood independently and 14 walked independently, with most achieving these milestones within normal developmental windows [