Bridging the Gap: Gene Therapy in a Patient With Spinal Muscular Atrophy Type 1

SMA as a Neurologic Genetic Disease

SMA, a leading genetic cause of infant mortality,³ is archetypal in that knowledge of the underlying genetic mechanisms has allowed for use of gene-specific approaches to alter the disease course. Most SMA cases (95%–98%) are caused by homozygous deletions in SMN1, which lies in the telomeric region of chromosome 5q13(Figure 1A).⁴ The remaining cases involve small intragenic SMN1 alterations.⁵ These alterations lead to progressive limb muscle weakness, hypotonia, and atrophy with motor function decline and the development of bulbar difficulties and respiratory insufficiency.⁶ A centromeric SMN2 variant⁷ may be present in variable copy numbers on 5q13 and is an almost identical homolog of SMN1 except for a few nucleotides.³ A crucial C-to-T substitution in exon 7 at position 6 alters SMN2 splicing so that ∼90% of transcripts lack exon 7, and the resulting truncated SMN protein is rapidly degraded. The remaining 10% of protein produced is insufficient for motor neuron survival,³ and SMN2 copy number is inversely correlated with phenotypic severity. Current advanced molecular treatments for SMA target repletion of SMN in motor neurons by replacing SMN1 or acting on SMN2 genes transcription.

Figure 1 Molecular Mechanisms of Spinal Muscular Atrophy (SMA), Basic Design of Adeno-Associated Virus (AAV)–Based Gene Therapy and Molecular Therapies for SMA

(A) Alterations in the telomeric SMN1 gene (red crosses) lead to reduced production of the functional survival motor neuron protein (SMN-FL). In SMN2, a transition switch from C to T alters the splicing of exon 7, and the resulting transcript (SMNΔ7) lacks the exon in 90% of cases, leading to an unstable, rapidly degraded product. The remaining 10% of SMN-FL from SMN2 cannot sustain motor neuron survival, with consequent motor neuron death and muscle atrophy. (B) The general design of AAV-based gene therapy. Modular structure consisting of specific promoter (red), intron, open reading frames (ORFs, green), the 3′ untranslated (UTR) region, and polyadenylation (poly(A), light blue) region enable targeted and cell-specific transgene (orange) expression. AAV vectors can carry genomes up to 5 kb in length. (C) Mechanisms of action of the three approved molecular therapies for SMA. The precise mechanism of action of risdiplam is unknown, but it supposedly displaces the heterogeneous nuclear ribonucleoproteins (hnRNP G) from exon 7 and enhances the interaction of U1 small nucleolar RNA (snRNA) with exon 7 splicing enhancer (ESE2). These molecular changes increase translation of SMN2 into functional SMN. Nusinersen is a single-stranded antisense oligonucleotide taken up into cells by endocytosis. It binds nuclear SMN2 pre-mRNA and displaces the hnRNP protein that usually prevents retention of exon 7 in the final transcript, enabling production of complete SMN. Gene replacement with onasemnogene abeparvovec consists of SMN1 cDNA carried by a nonreplicating, nonintegrating AAV9. When the cDNA is released in the episomal form in the motor neuron cytoplasm, it begins to produce copies of SMN1 transcripts encoding functional SMN.

Molecular Therapies in Neurology

Molecular therapies target the underlying disease mechanism with the aim of reconstituting gene function or manipulating RNA expression. Three techniques are currently available in clinical neurology¹ (additional data listed in eTable 1 in the Supplement, links.lww.com/WNL/C379). Their mechanism of action depends on the underlying gene alterations, which may cause partial or complete loss of function, gain of toxic function, or both.

When loss of function is the prevalent mechanism, the aim of treatments is restoring expression of the affected protein. If the prevailing mechanism is gain of toxic function, the therapy goal is to suppress expression of the disease-related protein. Gene replacement with viral vectors (Figure 1B) or ASOs/miRNAs can be tailored to downregulate or upregulate target gene/mRNA expression. For instance, when protein expression is reduced or absent, as in SMA or DMD, gene replacement therapy and ASOs increase the expression of the missing protein. The viral vector delivers cDNA encoding a functional SMN protein or dystrophin, and ASOs modulate splicing of SMN2 or the mutant dystrophin gene to increase overall levels of functional protein. Conversely, with hereditary transthyretin amyloidosis or SOD1-related amyotrophic lateral sclerosis,⁸ the underlying pathomechanism is associated with a toxic gain of function alteration, and viral vectors, siRNAs, or ASOs are used to reduce levels of the mutated protein.

Molecular Therapies in SMA

EU and US regulatory bodies have approved 3 therapies for SMA: onasemnogene abeparvovec (Zolgensma), nusinersen (Spinraza), and risdiplam (Evrysdi) (Figure 1C and eTable 1, links.lww.com/WNL/C379). All 3 increase levels of functional SMN by re-expressing a healthy copy of SMN1 cDNA or by modifying SMN2 splicing. Onasemnogene relies on nonreplicating, recombinant AAV9 that can cross the blood-brain barrier, at least in infants, to deliver a functional SMN cDNA transgene to motor neurons⁹ (eTable 1, mechanism 1, links.lww.com/WNL/C379). It is given once intravenously and indicated for children younger than age 2 years with SMA type 1.⁹ Nusinersen is a short single-stranded oligonucleotide that is delivered intrathecally and alters SMN2 pre-mRNA splicing to ensure exon 7 inclusion, boosting production of the functional protein¹⁰ (eTable 1, links.lww.com/WNL/C379, mechanism 2). The small molecule risdiplam, an oral therapy for SMA, binds SMN2 pre-mRNA in 2 regions, increasing the quantity of functional SMN⁹ (eTable 1, links.lww.com/WNL/C379, mechanism 3). The most suitable treatment for each patient, best initial approach, and possible advantages of combination therapies remain unclear.^9,11

Early diagnosis is paramount for access to and benefit from molecular therapies in neurology, yet diagnostic delays persist.¹² Implementation of prenatal genetic testing or neonatal screening for SMA could lead to earlier diagnosis and improve access to personalized therapies.¹³ Our patient showed a remarkable clinical amelioration despite being symptomatic with a low CHOP-INTEND score at treatment initiation. Although the diagnosis was made at a tertiary care center when she was age 4.5 months, symptoms may have manifested 3 months earlier. Our case highlights the need for neonatal screening to treat patients as soon as possible to yield maximum benefit from molecular treatments.

Other gene replacement therapies approaches are under investigation for a wide range of neurologic diseases. In most cases, these experimental therapies target loss-of-function alteration diseases such as DMD, lysosomal disorders (e.g., Pompe, Gaucher, and Fabry diseases),¹⁴ idiopathic Parkinson disease, Leber hereditary optic neuropathy, and Charcot-Marie-Tooth 1A.¹ Less frequently, viral-mediated delivery of microRNAs or short hairpin RNAs to silence disease genes is used for disorders involving gain-of-function mechanisms, such as amyotrophic lateral sclerosis (SOD1, C9orf72, and FUS alterations) and Huntington disease.⁸ In addition, genome editing with mRNA CRISPR-Cas9 is being explored for Huntington disease¹³ and hereditary transthyretin amyloidosis.¹⁵

The efficacy:safety ratio must be carefully evaluated for classical gene transfer approaches with AAV or other viruses and with genome editing. Overall, genetic molecular strategies offer new perspectives and hope for patients affected by genetic and acquired neurologic diseases with known molecular mechanisms.