Exon-skipping advances for Duchenne muscular dystrophy

Abstract

Duchenne muscular dystrophy (DMD) is a fatal genetic disorder characterized by progressive muscle wasting that has currently no cure. Exon-skipping strategy represents one of the most promising therapeutic approaches that aim to restore expression of a shorter but functional dystrophin protein. The antisense field has remarkably progress over the last years with recent accelerated approval of the first antisense oligonucleotide-based therapy for DMD, Exondys 51, though the therapeutic benefit remains to be proved in patients. Despite clinical advances, the poor effective delivery to target all muscle remains the main hurdle for antisense drug therapy. This review describes the antisense-based exon-skipping approach for DMD, from proof-of-concept to first marketed drug. We discuss the main obstacles to achieve a successful exon-skipping therapy and the latest advances of the international community to develop more powerful chemistries and more sophisticated delivery systems in order to increase potency, bioavailability and safety. Finally, we highlight the importance of collaborative efforts and early dialogue between drug developers and regulatory agencies in order to overcome difficulties, find appropriate outcome markers and collect useful data.

Introduction

Duchenne muscular dystrophy (DMD, OMIM #310200) is a rare neuromuscular X-linked disorder due to mutations in the DMD gene (OMIM 300377, Xp21.2-p21.1), one of the longest human genes spanning 2, 5 Mb and comprising 79 exons. DMD gene encodes for dystrophin (Uniprot P11532), an essential protein of the dystrophin–glycoprotein complex that links the extracellular matrix to cytoskeleton being fundamental to maintain muscle cell integrity. Among all muscular dystrophies, DMD is the most common affecting 15.9–19.5 in 100 000 newborn males (1). Patients manifest onset symptoms such as muscular weakness and walking abnormalities in their early infancy (1–3 years old). Although clinical characteristics vary among patients, it relentlessly progresses to total loss of ambulation by the age of 12–14 years. With advancing age, DMD children develop cardiomyopathies and breathing problems due to respiratory muscles weakness leading to premature death by their thirties.

DMD is caused by mutations in the dystrophin gene leading to complete loss of protein function. Most DMD patients (∼80%) carry large (one or more exons) deletions and duplications. The remaining 20% are affected by small mutations including nonsense mutations, small insertions or deletions and splice site mutations that result in a premature stop codon or a rupture of the reading frame (2,3).There is no cure for DMD and current therapies focus on alleviating symptoms and preventing inflammation and muscle necrosis/wasting. The gold standard in DMD management consists of corticosteroid therapy, rehabilitation and assisted ventilation which have improved quality of life and extended DMD patients life expectancy by 10 years (4–6). However, pharmacotherapy does not prevent disease progression and long-term steroid therapy induces undesirable side effects. In this regard, there is a major need to develop safe and efficient therapies for this devastating disease as it has been recognized by the Food and Drug Administration (FDA) that recently issued a guidance for DMD drug development (http://www.fda.gov/).

During the last 20 years new therapeutic avenues based on pharmacological, genetic or cellular approaches have been developed in animal models and some human clinical trials are already ongoing (https://clinicaltrials.gov) (7–9). All of them aim to correct or compensate dystrophin deficiency in muscles. Among pharmacological approaches, stop codon read through therapy had some success in preventing mRNA truncation due to premature stop codon mutations which affect nearly 10% of DMD patients. The administration of amynoglicoside antibiotics (gentamycin) or other nonsense suppression molecules such as arbekacyn or ataluren has demonstrated increased dystrophin levels in animal models (10,11). Ataluren (Translarna, PTC Therapeutics) recently received market approval by the European Medicines Agency (EMA) (http://www.ema.europa.eu/docs/en_GB/document_library/Summary_of_opinion_-_Initial_authorisation/human/002720/WC500167527.pdf) although Phase III clinical trial failed to show clinical benefit against placebo (12) (NCT00847379 and NCT01826487). Another positive pharmacological strategy for DMD is the upregulation of utrophin, an autosomal paralogue of dystrophin that is also involved in cytoskeleton attachment to extracellular matrix. Recent results from open-label Phase II clinical trial of Ezutromid (Summit Therapeutics), an utrophin modulator, demonstrated a decrease in muscle damage and proof of mechanism (NCT02858362) (http://g923gvc6go14zuzz1vaotb19.wpengine.netdna-cdn.com/wp-content/uploads/2018/01/2018_RNS_02-PhaseOut-DMD-Interim-Data-FINAL-1.pdf) (7).

Additional therapeutic strategies for DMD based on the recovery or complementation of dystrophin protein are gene therapy with viral vectors or cell transplantation. Gene therapy aims to introduce mini gene, named mini or micro dystrophin, using adeno-associated virus (AAV) (13). Despite huge obstacles, gene therapy for DMD has progressed positively and human clinical trials are currently ongoing (NCT03368742) (14). Cell-based therapies to promote muscle regeneration through autologous stem cells hold great promise for treating DMD but are still in early stages and need further experiments (15,16).

Among all, exon skipping is one the most promising therapeutic strategies for DMD. In this review, we will focus on this particular approach and present the exon-skipping principle, the proof-of-concept results in animal models and clinical data already available. Furthermore, we will discuss the recent advances in the development of new antisense oligonucleotide (ASO) chemistries and different available tools used to improve ASO delivery, one of the unmet challenges in exon-skipping therapy. Finally, we will discuss the limitations of each approach, regulatory challenges and promises of exon skipping.

AON Mediated Exon Skipping

From proof-of-concept to drug approval

The idea of exon skipping as a therapeutic approach for DMD came partly from the observation of the mutations found in Becker muscular dystrophy patients. Indeed, BMD is a milder form of muscular dystrophy, characterized by in-frame deletions which allow the production of a partially truncated and functional protein. Based on this observation, scientists developed the so-called ‘exon-skipping’ strategy aiming at restoring the out-of-frame mRNA of DMD patients to obtain proteins similar to those found in BMD patients. For this purpose, antisense sequences are designed to hybridize and mask the splicing signals of a specific exon, to prevent its inclusion in the final mRNA by the splicing machinery, therefore leading to a larger deletion and the synthesis of a BMD-like protein (Fig. 1). Although the exon-skipping approach appears to be applicable to a large proportion of patients [possibly up to ∼83% of all DMD patients (17)], one should keep in mind that this will not offer a definite cure but an improvement towards a BMD-like phenotype depending on the functionality of the restored dystrophin.

Figure 1.

Antisense-mediated exon skipping rationale for DMD. Most patients with Duchenne muscular dystrophy carry mutations which disrupt the open-reading frame of the dystrophin pre-mRNA. In this example, the deletion of Exon 50 creates an out-of-frame mRNA and leading to the synthesis of a truncated non-functional or unstable dystrophin (left panel). An antisense oligonucleotide (ASO) directed against Exon 51 can induce effective skipping of Exon 51 and restore the open reading frame, therefore generating an internally deleted but partly functional dystrophin (right panel).

The proof of concept of the exon-skipping therapy for DMD was first demonstrated by Pramono et al. (18) in lymphoblastoid cells and by Dunckley et al. (19) in cultured mouse cells in vitro. Since then, numerous in vivo studies have provided preclinical evidence for the therapeutic potential of the antisense strategy for DMD in several animal models. In particular, the mdx mouse model, which harbors a nonsense mutation in Exon 23, has been used extensively to evaluate the potency of various ASO chemistries such as 2′-O-methyl-phosphorothioate (PS) (2′OMePS) (20), phosphorodiamidate morpholino oligomers (PMO) (21,22), locked nucleic acid (LNA) or PNA (23,24) (Fig. 2). Following encouraging preclinical results obtained in animal models, two ASO chemistries (2′OMePS and PMO) targeting Exon 51 of the DMD gene were evaluated in clinical trials.

Figure 2.

Chemical structures of a selection of oligonucleotide analogues that have been tested as antisense agents (2′OMe, 2′OMethyl; 2′MOE, 2′OMethoxyethyl; LNA, locked nucleic acid; PMO, phosphorodiamidate morpholino oligomer; PNA, peptide nucleic acid; tc-DNA, tricyclo-DNA).

Drisapersen

The 2′OMePS ASO targeting the human DMD Exon 51, known as Drisapersen (originally PRO051) and developed by Prosensa/GSK/Biomarin Pharmaceuticals, was the first oligonucleotide tested in DMD patients. The first Phase I trial was conducted in four non-ambulant patients which received an intramuscular (i.m.) injection of 0.8mg in the tibialis anterior (TA) (PRO-051 CLIN01). Biopsies performed 28 days after the injection revealed dystrophin expression at the sarcolemma after immunofluorescence analysis (25). Subsequently, a Phase I/IIa trial evaluating the systemic delivery of drisapersen via subcutaneous injections was conducted in 12 patients with a dose escalation ranging from 0.5 to 6 mg/kg for 5 weeks (NCT01910649). In this study, detectable Exon 51 skipping levels were observed in patients dosed with 2 mg/kg or more of drisapersen and dystrophin expression was found in the TA of 10/12 patients. Following these encouraging results the study was extended to 12 weeks of treatment at 6 mg/kg and resulted in an improvement of the 6-min walk test (6MWT) in treated patients (26). The 12 initial patients were enrolled in a long term trial, consisting of 72 weeks of treatment at 6 mg/kg followed by an interruption of 8 weeks and an intermittent dosing until 188 weeks (NCT01910649). After 177 weeks, an 8-m improvement of the 6MWT was observed in the 10 patients able to complete the test at the baseline (27). A randomized controlled trial DEMAND 2 (NCT01153932) comparing two subcutaneous dosing regimen (continuous or intermittent injections) to placebo was then initiated in 53 ambulant patients for 48 weeks. That was the first time that a significant increase in the 6MWT was demonstrated using drisapersen compared to placebo with a mean difference of 35.09 m at Week 25 and 35.84 m at Week 49 (28). In parallel, a Phase III trial DEMAND 3 (NCT01254019) had started involving 186 ambulant patients worldwide with 125 receiving subcutaneous dosing at 6mg/kg per week and 61 receiving placebo (29). Unfortunately, although there was a tendency of drisapersen-treated patients to walk further than placebo-treated patients, this study failed to demonstrate a statistically or clinically significant improvement of the 6MWT. Post hoc analysis suggested that treatment may have had a therapeutic effect on younger patients (27) and a marketing approval application was filed. FDA rejected this application based on the facts that the pivotal trials did not demonstrate significant clinical benefit or clear increase in dystrophin expression, and that significant toxicities were observed (30). Different types of adverse events, caused by the subcutaneous injection of drisapersen, had indeed been observed in clinical trials, including irritation at injection site, proteinuria and thrombocytopenia in about 2% of patients. Following FDA rejection, BioMarin withdrew its marketing authorization application with the European Medicine Agency in 2016 and discontinued all 2′OMePS ASO programs for DMD.

Eteplirsen

The first clinical trial using PMO ASOs for DMD, developed by Sarepta Therapeutics (formerly known as AVI BioPharma) was launched in 2007 by the UK-lead MDEX consortium. The PMO-ASO targeting Exon 51, known as eteplirsen (previously known as AVI-4658) was injected intramuscularly in the extensor digitorum brevis muscle of 7 DMD patients receiving 0.09 mg or 0.9 mg of PMO (NCT00159250). Muscle biopsies were performed 3–4 weeks after the injection and showed an increased level of dystrophin in the treated muscle compared to the contralateral muscle injected with saline solution in the five patients receiving the highest dose (31). A dose escalation study was then performed on 19 patients receiving 0.5–20 mg/kg of eteplirsen weekly for 12 weeks (NCT00844597) and demonstrated significant dystrophin restoration in patients treated from 2 to 20 mg/kg (32). A long-term systemic dosing was subsequently conducted in 12 patients with 30 and 50 mg/kg per week for 48 weeks (NCT01396239 and NTC01540409). This extension study demonstrated an increase in dystrophin-positive fibers from 40 to 60% in patients treated with 30 and 50 mg/kg. Despite the loss of ambulation of two patients in the 30 mg/kg cohort, a significant increase in the 6MWT was measured in both cohorts compared to placebo (33).

These 12 patients were then followed for 3 years. None of the 10 remaining boys lost ambulation after 3 years of treatment and an advantage of 151 m was observed in the 6MWT compared to natural history controls. Moreover, respiratory functions of these patients had remained relatively stable during the study (34). Based on these clinical data, Sarepta applied for approval with the FDA which asked for dystrophin quantification in biopsies obtained after 180 weeks of treatment. This revealed an increase of dystrophin protein levels from 0.08 to 0.93% which was not sufficient to obtain FDA approval. In 2016, Sarepta provided western blotting data on additional patients involved in an ongoing Phase III clinical trials (4658-301 or PROMOVI, NCT02255552) to the FDA. At that time, 13 patients had been treated for at least 48 weeks and analysis of biopsies taken before and after treatment showed an increase of 0.28% of dystrophin in half of patients. In September 2016, eteplirsen received accelerated approval by the FDA based on these dystrophin data (brand name: Exondys51). The approval was shrouded in controversy and caused extensive internal debate because the clinical benefit of eteplirsen was not clearly established due to the low number of patients involved in clinical trials (the approval was based on 12 patients followed during 3 years). Sarepta therefore needs to confirm the functional effect in other trials and four clinical studies are currently ongoing using eteplirsen. These trials involve a large number of patients and take into account different stage of the disease (NCT03218995, NCT02420379, NCT02255552 and NCT02286947). Sarepta also developed PMO ASOs to treat patients amenable to Exon 45 or Exon 53 skipping for which they have 3 clinical trials ongoing (NCT02500381, NCT02310906 and NCT02530905).

Current Challenges of ASO and Development of New Compounds

To date only PMO has received regulatory approval (though the clinical benefit must still be proved as mentioned above) while other antisense molecules designed to restore dystrophin protein (i.e. 2′OMePS) have failed in DMD trials due to a marginal clinical benefit. In this sense, the main hurdle of antisense drugs to get marketed is their limited efficacy due to a poor effective delivery to target tissues. Actually, it is estimated that <1% of ASO reach the correct cellular compartment (35). In addition, oligonucleotide distribution is often restricted because of body's barriers, such as the blood–brain barrier that prevent drugs from reaching the central nervous system.

Although there is increasing knowledge in cellular uptake and trafficking [for review see (36,37)], many questions remain unanswered to fully decipher which factors or proteins direct ASOs to productive pathways. There are several important limitations/parameters to consider for an effective delivery. First of all, many antisense molecules are rapidly excreted by the kidney because of their low size. The gold standard PS-modified ASOs bind to plasma/serum proteins which decrease its renal clearance and increase its accumulation in tissues, particularly kidney and liver (37). In addition unmodified ASOs are easily degraded by serum and nucleases, a problem that has been successfully reduced with novel chemical modifications (38,39). ASOs need to cross endothelium barrier and be specifically addressed to the nucleus of the tissue of interest, minimizing its exposure in unintended tissues.

International efforts are ongoing to develop more powerful chemistries and more sophisticated delivery systems in order to increase potency, bioavailability and safety. It is important to note that any oligonucleotide with negative or uncharged backbones (PNAs or morpholinos) are not able to simply diffuse the lipidic bilayer membrane (40). Recent non-viral delivery strategies include the use of ligands (lipids or peptides) conjugated to ASO to reach specific cell types or enhance permeability like cell-penetrating peptides (41–43) or lipidic nanocarriers (44,45).

Peptide-conjugated PMO

One of the major issues with PMO AONs is their inability to restore dystrophin expression in the cardiac muscle. To solve this problem, scientists investigated the conjugation of small peptides to the uncharged PMO chemistry. These small peptides are principally arginine-rich peptides that enable enhanced cell uptake. Different types of peptide-PMO (PPMO) have been generated and tested in the DMD mouse model (46). Following encouraging preclinical data in mdx mouse, one of the PPMO, named AVI-5038 (or commonly referred to as B-PMO) and developed by Sarepta was tested in non-human primates. Unfortunately, this study revealed severe kidney toxicity at ineffective low doses (46). The toxicity may have been driven by the amount of arginine, which might not be well supported in higher species. Since these initial trials, novel promising peptides have been developed by Sarepta and clinical evaluation of PPMO SRP-5051 started in December 2017 (NCT03375255).

Additional modifications to PMO have been designed as the addition of a muscle-specific peptide (MSP) between the arginine-rich peptide and the PMO. Gao et al. demonstrated that the latter modification induced higher exon skipping levels in muscles compared to the original PPMO (41). However, no effect was observed when MSP was placed before the arginine-rich peptide, suggesting that MSP is acting more like a hydrophobic spacer than directing PMO to muscles (41).

In parallel, scientists developed novel series of transduction peptides named Pip (PMO internalization peptide) in order to increase exon skipping in the heart. After testing a large number of peptides, it appeared that Pip5e-PMO induced higher levels of dystrophin restoration in peripheral muscles after a single intravenous (i.v.) injection at 25 mg/kg, than previous B-PMO (47). Interestingly, a particularly high level of dystrophin restoration was found in the heart. The enhanced activity of Pip5-PMO may be attributed to increased nuclear PMO delivery in cardiomyocytes as it was demonstrated in vitro in a heart slice model (47).

A new generation of Pip-PMO (Pip6-PMO) has been developed recently based on Pip5e-PMO. Different modifications were tested such as alteration of the hydrophobic sequence, partial deletion of the hydrophobic core, alteration of arginine position or addition of a second-arginine (43). While modification on the arginine or partial deletion of the hydrophobic core appeared detrimental for efficacy, alteration of the hydrophobic sequence improved splicing activity in vivo compared to Pip5e-PMO. Pip6-PMO (Pip6f-PMO) was injected at 10 mg/kg for 12 weeks in mdx mice which resulted in high levels of dystrophin restoration in the heart without any apparent immune response. Moreover, injection of Pip6f-PMO to exercised mdx mice prevented onset of cardiomyopathy and reduced cardiac markers of pathology (48).

Tricyclo-DNA

Tricyclo-DNAs (tcDNA) were developed at the end of the nineties and are considered one of the most promising drugs for the treatment of DMD because of their unique pharmacological properties (49,50). TcDNAs are conformationally constrained nucleotides specially designed to limit torsional flexibility of the sugar backbone in order to reduce the entropy of homoduplex and stabilize heteroduplexes tcDNA/RNA. Their efficacy has been recently demonstrated in our laboratory where functional correction and neurobehavioral improvement were achieved in mdx mice treated with a 15-mer tcDNA (51). Compared with 2′OMePS and PMO, tcDNA chemistry revealed consistently higher exon skipping levels and evidenced for the first time dystrophin rescue in all tissues affected by the lack of dystrophin, including skeletal muscles, heart and central nervous system. Moreover, it displayed higher RNA-binding properties than 2′OMePS and PMO, permitting the use of shorter AONs (52). Most importantly, tcDNA (15- or 13-mer) treatment translated to functionally improvement in muscular, respiratory and cardiac systems of mdx mice as well as demonstrated beneficial effects on their behavior.

Preclinical toxicity testing of novel ASOs at an early stage of development is crucial to avoid subsequent drug failure in further clinical studies. We have recently evaluated the toxicological profile of tcDNA-ASOs in DMD mouse model and shown that high dose tcDNA treatment (200 mg/kg/week for 12 weeks) was well-tolerated in mice (52). Inflammatory response to PS-tcDNA is a widely known effect of fully modified PS-ASOs (53,54). Our work showed no significant differences in complement activation nor cytokine levels in mdx treated mice compared to mdx control mice. In addition, antisense molecules are known to accumulate particularly in liver and kidney causing some undesirable toxic responses, especially in repeated-dose studies. In this regard, histological findings were limited to minimal glomerular changes and few cell necrosis in proximal tubules and minimal liver inflammation. The evaluation of more sensitive renal biomarkers showed a moderate upregulation in kidney injury molecule-1 and renin, indicating an emerging renal toxicity possibly due to tcDNA accumulation, which has been previously reported elsewhere (55). However, no significant changes were shown in serum biomarkers compared to controls. Further studies including a washout period off the drug allowed us to demonstrate the disappearance of observed side toxic effects (data not published). Importantly, there were no unexpected class-related toxicological issues following tcDNA treatment.

Despite the general encouraging safety profile of tcDNAs, fully modified PS-ASO are known to inhibit coagulation and activate alternative pathway of complement as a protein binding related effect. To further understand this issue, we recently investigated the efficacy and toxicological effect of PS linkage content within tcDNA backbone. It is important to underline that tcDNA in their phophodiester (PO) version are already resistant to nuclease activity (38,56) so the PS bonds are required mainly to enhance cellular uptake and biodistribution (57). Our results confirm that the number of PS modifications correlates both with AONs efficacy and their associated side effects (Echevarría et al., manuscript in preparation). Therefore, an optimized balance between PO and PS bonds can minimize PS related toxicity without affecting the efficacy. However, the promise of tcDNA-ASO, as all novel ASO still depends on how well it will be tolerated in humans and full reglementary toxicological studies are currently being planned to envisage future clinical evaluation in DMD patients.

Stereopure ASOs

Stereoisomers are molecules that have identical chemical composition and sequence of bonded atoms but differ in their tridimensional distribution. PS modifications, in which one of the non-bridging phosphate O-atoms is replaced with an S-atom, have two possible stereoisomers, known as Sp and Rp. As a result, an oligonucleotide with n PS junctions have 2ⁿ stereoisomers, meaning half a million stereoisomers for a 20-mer ASO-PS. As explained earlier, PS linkages are one of the main chemical modifications adopted for antisense drugs in order to protect ASOs from nuclease activity and increase their stability and affinity to target RNA compared to natural phosphodiester (PO) backbones. It is now well known that chiral drugs have different biological activities as it has been recognized by the FDA (https://www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ucm122883.htm) and this has been particularly demonstrated for PS-based ASOs (58). In this study, the authors were able to produce stereopure versions of mimopersen, a PS modified ASO used to treat familial hypercholesterolemia, that showed different pharmacological properties compare to each other and to mimopersen.

Until now, the majority of PS-ASOs evaluated in clinical trials have been synthetized as racemates as it was very difficult to produced purified stereoisomers. This new method developed by Wave Life Sciences to reliably synthesize stereopure oligonucleotides offers unprecedented opportunities to increase the efficacy and specificity of ASOs. Wave Life Sciences has developed stereopure oligonucleotides targeting the human dystrophin Exon 51 and pre-clinical studies indicate an enhanced exon skipping and dystrophin protein restoration compared with other exon-skipping compounds such as eteplirsen or drisapersen (52% dystrophin protein restoration as compared with normal skeletal muscle tissue lysates, versus ∼1% when testing other exon-skipping ASOs in vitro (https://www.wavelifesciences.com/pipeline/lead-programs/)). A Phase I clinical trial has been initiated in November 2017 to assess to safety and tolerability of ascending doses of WVE-210201 administered i.v. in DMD patients.

Vectorized Exon Skipping as an Alternative for AONS

To overcome the challenges of the ASO-mediated exon skipping approach, such as delivery and potential toxicity associated with repeated administration, the delivery of antisense sequences using viral vectors has been investigated. In particular, the use of AAV vectors which are extremely efficient for muscle gene transfer offers a particularly interesting alternative. Studies using U7 or U1 small nuclear RNA (snRNA) to carry the antisense sequences have demonstrated efficient and sustained rescue of dystrophin in mdx mice (59,60). The natural antisense sequence of the snRNA is replaced with the antisense sequence of interest to target a specific exon (Fig. 3A). Linking the ASO to a snRNA allows its proper subcellular localization and thereby facilitates the inclusion of the antisense sequence into the spliceosome (61,62). Following the proof of principle in mdx mice, additional studies from our group and others have shown correction of the dystrophic phenotype and partial recovery of muscle strength after i.m. and i.v. injections of AAV-U7 snRNA in golden retriever muscular dystrophy dogs (63,64). This type of viral vector mediated exon-skipping approach can also be combined with cell-based therapies using lentiviral vectors to correct the reading frame of muscle stem cell before re-injection (65). In addition, this vectorized approach offers the possibility to efficiently deliver multiple snRNA cassettes to target multiple exons which could be extremely useful to overcome the personalized medicine issue for DMD treatment (Fig. 3B).

Figure 3.

(A) Structure of the wild-type (WT) and modified U7snRNP. WT U7snRNA (top left) is a single strand RNA molecule composed of a natural complementary sequence to the histone pre-mRNA named Histone Downstream Element (HDE), a binding site for Sm protein and a hairpin structure at the 3′ end. Altogether the U7snRNA associated with the specific U7 sm protein form the U7snRNP (ribonucleoprotein complex), playing a role in histone pre-mRNA processing. The Sm core of the U7snRNP consists of seven proteins encircled around the snRNA binding site and forming a torus structure. For splicing modulation and exon skipping approach in particular, the U7snRNA is genetically modified in two ways: (i) the U7 Sm binding site is replaced by the consensus sequence derived from the spliceosomal snRNPs called the U7 Sm OPT which allows better subcellular localization (nucleus) and avoids the cleavage of the histone pre-mRNA. The U7 Sm OPT modification leads to the replacement of two Sm proteins (represented in white in WT U7 snRNP figure), Lsm10 and Lsm 11 by D1 and D2 Sm proteins (represented in dark grey in the modified structures); and (ii) the natural HDE sequence is replaced with specific antisense sequences of the targeted exon. (B) A simplified representation of AAV-U7snRNA mediated antisense approach. Left panel: structure of the single strand DNA genomes packaged in recombinant AAV (rAAV) capsids (hexagon). A single U7snRNA molecule targeting exon A can be inserted and then transcribed from a single U7snRNA gene under the control of its natural U7 promoter. The relatively short size of the U7snRNA cassette authorizes the insertion of multiple U7 constructions in a same and unique rAAV genome. This approach allows multiexon skipping strategies through the expression of different modified U7snRNA from a single rAAV backbone (right panel).

Altogether, these data have shown the efficacy of AAV-U7 snRNA-mediated exon-skipping therapy for DMD, even if this strategy still faces the challenges associated with viral vectors and gene therapy approaches, including the maintenance of the viral genome over time (66,67).

Limitations and Regulatory Challenges of the Exon-Skipping Approach

Media has reported widely on eteplirsen, the only ASO-based treatment for Duchenne patients that has been controversially granted accelerated approval by the FDA and is currently pending for EMA review. Regardless of what the future clinical studies brings, this approval has certainly generated increasing interest in ASO based drugs and will help the development of subsequent ASOs.

Eteplirsen approval was based on restoration of dystrophin levels in biopsies from treated patients despite not being accepted as a surrogate biomarker, which raises the issue of validated biomarkers for DMD trials. The EMA regulatory position is that current available data are insufficient to establish a correlation between dystrophin levels and functional outcomes (68).

Despite being highly promising, eteplirsen represents a treatment option for only 13% of DMD patients in USA, while the other 87% do not have access to any therapeutic option. Exon skipping based drugs for additional groups of patients targeting other exons are already under development and first clinical trials for Exon 45 and Exon 53 have been launched (69) (Table 1). However, many other exons and therefore trials will have to be developed. In this regard, smaller groups of patients harboring mutations affecting <1% of DMD population (3,17) are much more challenging for drug developers as controlled clinical studies may not be feasible due to limited number of patients. One possible strategy in those cases is the partial extrapolation of clinical data from marketed ASO in order to supplement data (68). As an alternative, the Duchenne community recently proposed ASOs ‘class approval’ which aimed to get approval for a certain ASO chemistry as a group and not for a specific sequence. This could have facilitated ASO medicine approval but EU legislation currently does not allow it, as medicines must be investigated individually. One of the latest emerging approaches to overcome the limitation of such a personalized medicine is the multiple exon skipping (70), which includes exons 3–9 and 45–55 deletions using a ‘mixture’ of ASO (71,72) or vectorized approaches (73).

Table 1.

Clinical trials ongoing and listed in http://clinicaltrials.gov evaluating exon skipping in DMD patients

Compound	Sponsor	Target exon	SSO chemistry	Phase	Status	Clinicaltrial.gov Number
Eteplirsen (AVI-4658, Exondys51®)	Sarepta Therapeutics	51	Morpholino	II	Ongoing, not recruiting	NCT02286947
Eteplirsen (AVI-4658, Exondys51®)	Sarepta Therapeutics	51	Morpholino	II	Ongoing, not recruiting	NCT02420379
Eteplirsen (AVI-4658, Exondys51®)	Sarepta Therapeutics	51	Morpholino	IIb	Ongoing, not recruiting	NCT01540409
Eteplirsen (AVI-4658, Exondys51®)	Sarepta Therapeutics	51	Morpholino	II	Recruiting	NCT03218995
Eteplirsen (AVI-4658, Exondys51®)	Sarepta Therapeutics	51	Morpholino	III	Ongoing, not recruiting	NCT02255552
Drisapersen (PRO051)	Biomarin	51	2′-O-methyl PS	III	Cancelled	NCT02636686
SRP-4053	Sarepta Therapeutics	53	Morpholino	I/II	Ongoing, not recruiting	NCT02310906
SRP-4045 and SRP-4053	Sarepta Therapeutics	45 and 53	Morpholino	III	Ongoing, recruiting	NCT02500381
SRP-4053	Sarepta Therapeutics	53	Morpholino	I/II	Ongoing, not recruiting	NCT02530905
AAV-U7/Exon 53	Genethon	53	AAV/U7	Observational	Ongoing	NCT01385917
AAV-U7/Exon 53	Genethon	53	AAV/U7	I/II	In preparation	–
WVE-210201	Wave	51		I		–

In addition to the cited challenges of ASO therapy, its limited applicability and the identification and use of appropriate biomarkers, it is important to fulfill regulatory criteria in order to obtain a successful access to market. In this sense, useful information such as coordination centers and patients eligible for specific DMD trials have been greatly facilitated by TREAT-NMD, a neuromuscular disease network created to promote excellence research programs in the neuromuscular field and ensure advancing therapies for patients (3,74). Moreover, EMA propose scientific advice and regulatory guidance to sponsors through different platforms and programs (PRIME, ITF, SME support) fostering early dialogue to optimize development plans and increase success rate of clinical programs (68).

Conclusion

The development of the exon-skipping therapy for DMD has immensely progressed in the last few years with the recent approval of the first ASO-based drug. Many challenges remain and international efforts are currently tackling the delivery issue with new compounds being developed and some of them already in trials. Regardless the efficacy of eteplirsen or drisapersen, many lessons have been learned during their clinical development, in particular about trial design, outcome measures and validated biomarkers. This has undoubtedly paved the way for the next ASO generation and allows us to foresee a hopeful future for exon skipping therapeutics for DMD.

Funding

The authors have financial support from Institut national de la santé et de la recherche médicale (INSERM), Agence nationale de la recherche (ANR – Chair of Excellence HandiMedEx), the Association Monegasque contre les myopathies (AMM) and the Duchenne Parent project France (DPPF).

Conflict of Interest statement. None declared.