Moving towards treatments for spinal muscular atrophy: hopes and limits

Abstract

Spinal muscular atrophy (SMA), one of the most frequent and devastating genetic disorders causing neuromuscular degeneration, has reached the forefront of clinical translation. The quite unique genetic situation of SMA patients, who lack functional SMN1 but carry the misspliced SMN2 copy gene, creates the possibility of correcting SMN2 splicing by antisense oligonucleotides or drugs. Both strategies showed impressive results in pre-clinical trials and are now in Phase II-III clinical trials. SMN gene therapy approaches using AAV9-SMN vectors are also highly promising and have entered a Phase I clinical trial. However, careful analysis of SMA animal models and patients has revealed some limitations that need to be taken very seriously, including: i) a limited time-window for successful therapy delivery, making neonatal screening of SMA mandatory; ii) multi-organ impairment, requiring systemic delivery of therapies; and iii) a potential need for combined therapies that both increase SMN levels and target pathways that preserve/rescue motor neuron function over the lifespan. Meeting these challenges will likely be crucial to cure SMA, instead of only ameliorating symptoms, particularly in its most severe form. This review discusses therapies currently in clinical trials, the hopes for SMA therapy, and the potential limitations of these new approaches.

Keywords:

• antisense oligonucleotide therapy
• gene modifier
• gene therapy
• neonatal screening
• neuromuscular disorder
• pharmacotherapy
• SMN1
• SMN2
• spinal muscular atrophy

1. Introduction

Spinal muscular atrophy (SMA) is a devastating neuromuscular disorder that leads to progressive muscle weakness and atrophy and that represents the most common lethal genetic disease in infants. Patients with SMA are divided into clinical sub-categories (termed SMA type I, II, III and IV) based on disease onset and severity, with SMA type I having the earliest onset and most severe phenotype [1]. Although SMA is considered to be a motor neuron disorder, additional organs can also be impaired, albeit mainly occurring in severely affected SMA mice and patients [2].

SMA is caused by functional loss of SMN1, whereas disease severity is influenced by the number of SMN2 copies and other SMA-modifying genes reviewed in [3]. As SMN2 mRNA is mainly alternatively spliced lacking exon 7 due to a single translationally silent variant, 90% of SMN protein is truncated and unstable. The remaining 10% are full-length transcripts, producing a protein identical to that encoded by SMN1 reviewed in [3]. As the SMN protein has a housekeeping function in small nuclear ribonucleoprotein biogenesis and splicing, the multi-organ impairment found in severely affected SMA mice or patients is an obvious consequence of SMN expression levels that fall under a certain critical threshold [2]. At present, there is no curative treatment available for patients with SMA, but impressive progress has recently been made towards the development of new therapies.

Here we discuss: i) current progress towards a therapy for SMA, and ii) potential limitations, based on novel biological observations in SMA animal models and SMA patients that will impact on the design and delivery of future therapies. In the expert opinion section we will discuss potential strategies to overcome these constraints.

2. Current progress towards SMA therapy

2.1 SMN-dependent therapies

The main focus of translational SMA research at present is the development of SMN-dependent therapies. These efforts include strategies directly targeting SMN protein stability, endogenous SMN2 mRNA transcription, or splicing by using small molecules (antisense oligonucleotides, AONs) or drugs, and approaches based on SMN gene replacement using self-complementary serotype 9 adeno-associated virus vectors (scAAV9) expressing SMN1.

Indeed, a high increase in central and peripheral SMN levels, leading to neuromuscular and systemic improvements, have been reported in recent preclinical trials in SMA mice using intravenous injection of scAAV9-SMN [4-6], subcutaneous delivery of SMN_Rx-AONs [7], or orally delivered small molecules [8]. A first Phase II clinical study using intrathecal delivery of SMN_Rx-AONs targeting SMN2 pre-mRNA in SMA type I patients showed some encouraging results, including increased muscle function scores [9]. Consequently, a first 2:1 randomized Phase III clinical trial in 117 SMA type II and III patients was launched by ISIS Pharmaceuticals (NCT02292537). Hoffmann La-Roche is currently recruiting 48 SMA patients (aged 2–55 years) in a double placebo-randomized Phase I study (NCT02240355) to test safety and tolerability of their orally applicable compound RO6885247 (former PTC RG7800) [8]. Novartis initiated an open label study (NCT02268552) to investigate their splice correction compound LMI070 in 22 SMA type I patients for safety, tolerability, pharmacokinetics and pharmacodynamics. Moreover, a gene therapy approach, using sc-AAV9-CB.SMN, entered the clinical phase at the Nationwide Children’s Hospital in Ohio, USA (NCT02122952), to evaluate safety and efficacy. A 3-cohort phase I study (escalade dose) has been initiated in May 2014, involving 18 SMA type I patients. Taken together, these various studies should provide a robust overview of the promise (and potential pitfalls) of targeting SMN levels in patients with SMA.

Previous drug development efforts based around histone deacetylase (HDAC) inhibitors also deserve attention as they are the only ones to date that have completed Phase III clinical trials. Valproic acid (VPA), one of the first HDAC inhibitors, was shown to increase SMN2 mRNA and SMN protein levels in vitro and in vivo. A first clinical study of VPA in 20 SMA type I, II and III patients demonstrated an increase of FL-SMN2 levels in ∼ 1/3 of patients the ‘VPA-responders’ [10], which seems to depend on CD36 expression, a fatty acid translocase [11]. Several open-label and placebo-controlled clinical trials with valproate and L-carnitine have been completed with only little significant phenotypic improvements in 2–5 year old SMA type II and III patients [12,13]. However, the placebo-controlled study lasted only 6 months, a too short time period to expect a significant outcome in this disorder. Currently, a Phase III randomized clinical trial is ongoing in India (NCT01671384), which includes 60 patients aged 2–15 years.

2.2 SMN-independent therapies

Several interesting SMN-independent pathways with the potential for therapeutic targeting in SMA have recently been identified. A clinical trial using one non-SMN targeted compound, the neuroprotective drug Olesoxime (TRO19622, Trophos), has shown modest improvements in motor function in SMA type II and III patients [14]. However, as the majority of non-SMN targeted drugs are still in pre-clinical phases of development, they will be considered in section 4 below.

3. Limitations of current SMA therapies

Although none of the above-mentioned approaches, except for those using HDAC inhibitors, have yet completed a Phase III clinical trial and showed substantial benefit for patients, there is well-warranted excitement and hope in the field, mainly based on the promising preclinical results. Importantly, however, none of the treatments currently progressing through human clinical trials detailed above are likely to offer a complete ‘cure’ for SMA. We suggest that three main confounding factors arising from the early and systemic nature of SMA will therefore need to be addressed in the next stages of therapy design and testing.

3.1 Therapeutic time window

Data from animal models of SMA provide strong evidence for the presence of a critical ‘therapeutic time-window’ for delivery of SMN-targeted therapies [15]. Indeed, increasing SMN levels after the onset of overt symptoms (postnatal day 8 in SMAΔ7 severe mouse models) provided only very little amelioration of disease symptoms [4,7] questioning the future efficacy of SMN-targeted treatments when delivered to symptomatic (particularly type I) SMA patients. This is of particular concern as there is currently no neonatal genetic screening for this disease, usually diagnosed after the appearance of the first clinical signs. Thus, SMN-targeted therapies delivered after disease onset may only have a limited capacity to ameliorate disease symptoms. In contrast, restoration of SMN levels in a milder SMA mouse model significantly improved motor abilities, underscoring a potential difference concerning the timing and nature of the therapeutic window [16] in severe versus mild forms of SMA [3].

3.2 Systemic nature of SMA

Given the growing awareness of the multi-organ nature of SMA (particularly in the most severe forms of the disease [2]), and the need to deliver therapies systemically in mouse models of SMA to achieve full benefit [7,16], standard drug-based pharmacological approaches offer a potentially attractive route to quickly develop effective, systemic disease-modifying therapies for SMA, which may be used either alone or in combination with molecular therapeutic approaches (including ASOs and AAV gene therapy).

3.3 Insufficient restoration of SMN levels and/or biodistribution using current SMN-targeted approaches

Although the first clinical trials using intrathecal injection of SMN_Rx-AON in SMA type I patients have reported some partial amelioration of disease symptoms, it is still uncertain whether the extent to which SMN levels have been increased will be sufficient to cure SMA. Measurements of SMN protein levels in cerebrospinal fluid after multiple dosing of SMN_Rx-AON showed a ∼ 120 – 160% increase from the depleted levels observed in untreated patients [9]. A range of previous in vivo and in vitro studies suggest that these modest increases most likely will not be sufficient to turn an SMA type I patient into a healthy individual, but rather may only reduce the severity of the disease (e.g., ‘convert’ a SMA type I into a type II or III patient), even if therapy could be started pre-symptomatically. Moreover, SMA mice carrying two SMN2 copies treated systemically and pre-symptomatically with SMN_Rx-AON or SMN-AAV9 remained smaller, showed reduced survival, and did not recover full muscle activity and body weight phenotype seen in wild-type control animals [4,5,7]. This means that additional functional support will be required to fully ameliorate disease pathophysiology in SMA.

4. Expert opinion

At present, there is no cure available for patients with SMA. One drug (Olesoxime, Trophos) and one AON (SMN_Rx, ISIS Pharmaceuticals) showed in Phase II – III clinical trials some encouraging results, but no major amelioration was observed. Additional drugs (RO6885247 from Roche and LMI070 from Novartis) are under clinical investigation. However, the long-term ambition of the SMA community remains to: i) substantially ameliorate SMA in patients after development of symptoms; and ii) fully counteract development of SMA in people with SMN1 deletion by delivering effective treatments at critical points in the therapeutic time-window (likely to mean pre-symptomatic delivery).

Work on animal models has clearly demonstrated that early therapeutic intervention is mandatory to achieve the best protective effect [4-7]. For this to be translated to patients, SMN1 deletion pre-symptomatic testing would need to be introduced into neonatal screening programs. However, amelioration or even stopping the disease progression in patients with SMA type II or III, in whom disease symptoms have already manifested, is also a key aim of research, where clinical translation is not dependent on neonatal screening for maximal effectiveness.

Recent breakthroughs in our understanding of pathways acting downstream from SMN that mediate disease pathogenesis in SMA have greatly expanded the range of potential therapeutic targets for SMA, opening up the possibility of delivering disease-modifying treatments outside of the ‘therapeutic time-window’ that exists for SMN. Several non-SMN-targeted pathways that contribute to disease pathogenesis in SMA have been reported to modulate the SMA phenotype, including ubiquitination pathways and beta-catenin signaling [17], PTEN signaling [18], RSK2 signaling [19], Rho-kinase pathway [20], ERK/AKT pathways [21], and miR-189/mTOR pathways [22]. Likewise, protective SMA genetic modifiers identified in asymptomatic SMN1-deleted individuals, such as PLS3 overexpression, which restores SMN-disturbed actin dynamics, also represent attractive SMN-independent therapeutic targets to stabilize and improve the neuromuscular function [23,24].

Major, achievable, ambitions for the future are therefore: i) to develop drug-based therapies that can extend the therapeutic time-window, in particular by stabilizing the neuromuscular system for a longer period, thereby facilitating a greater therapeutic benefit from parallel delivery of SMN-targeted therapies (such as AONs or gene therapy) and/or act to stabilize the neuromuscular system in their own right beyond the time-window that exists for SMN-targeted therapies; ii) to identify in SMA cells and animal models new targets acting either independently or downstream from SMN, and to develop and test new AONs/drugs acting on these new targets/pathways; iii) to validate new effective gene therapy treatments for SMA, based on the delivery of scAAV9 vectors designed to overexpress or to silence newfound targets for SMA, acting either independently or downstream from SMN and 4) to include SMA neonatal genetic screening.

Acknowledgments

The authors apologize for references omitted due to space constraints.

Declaration of interest

We gratefully acknowledge support from Deutsche Forschungsgemeinsschaft (Wi-945/14-2; Wi-945/16-1 and RTG 1960 to BW) and (SFB581, TP1 and TP4 to MS), EU-FP7 ‘NeurOmics’ (to BW), Euro-MOTOR (to MS), CMMC (to BW), the SMA Trust and Muscular Dystrophy UK (to THG) and AFM (to CM and MB). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.