Progress and prospects of gene therapy clinical trials for the muscular dystrophies

Abstract

Clinical trials represent a critical avenue for new treatment development, where early phases (I, I/II) are designed to test safety and effectiveness of new therapeutics or diagnostic indicators. A number of recent advances have spurred renewed optimism toward initiating clinical trials and developing refined therapies for the muscular dystrophies (MD's) and other myogenic disorders. MD's encompass a heterogeneous group of degenerative disorders often characterized by progressive muscle weakness and fragility. Many of these diseases result from mutations in genes encoding proteins of the dystrophin–glycoprotein complex (DGC). The most common and severe form among children is Duchenne muscular dystrophy, caused by mutations in the dystrophin gene, with an average life expectancy around 25 years of age. Another group of MD's referred to as the limb-girdle muscular dystrophies (LGMDs) can affect boys or girls, with different types caused by mutations in different genes. Mutation of the α-sarcoglycan gene, also a DGC component, causes LGMD2D and represents the most common form of LGMD. Early preclinical and clinical trial findings support the feasibility of gene therapy via recombinant adeno-associated viral vectors as a viable treatment approach for many MDs. In this mini-review, we present an overview of recent progress in clinical gene therapy trials of the MD's and touch upon promising preclinical advances.

Introduction

There are ∼7000 rare inherited diseases, examples of which include Duchenne muscular dystrophy (DMD), cystic fibrosis, the mucopolysaccharidoses and rare cancers. These disorders present a significant burden on the healthcare system where global impact at birth reaches approximately 1 per 1000 cases. Developing treatments for rare disorders is a challenge despite recent legislation aimed at providing regulatory and financial incentives to stimulate development of drugs for ‘orphan diseases’. As a measurement of the success of orphan drug products within the USA, a survey of compounds advancing from clinical development to approval from 2006 to 2010 revealed that 28% of all new drug applications fell into the rare disease category, and approval rates for rare versus common disorders have been similar (1). With the initiation of the Human Genome Project nearly 25 years ago, and recent advances in diagnostic sequencing technology (2), it was hoped that the genetic information attained would markedly change the way we understand and treat human disease.

Indeed, the concept of ‘gene therapy’, a phrase initially coined in the early 1970's by Friedmann and Roblin (3) has since generated several mood swings over the years within the field (4,5). While it is clear that progress has been made along several lines, unfortunately there are currently no FDA-approved gene therapy products, with efforts being cautiously devoted to the investigative stage to demonstrate safety and efficacy. In Europe, the first licensed gene therapy product was approved in 2012 for lipoprotein lipase deficiency (6) while in China marketing approval for a gene therapy product designed to treat cancer occurred in 2005 (7). It should be noted that with recent biotechnology advances such as the chimeric antigen receptor (CAR) technology, (8–10) newly developed biomarker evaluation, (11–15) improved gene transfer tools and expansion of disease models with improved genomics technology (16–20); there has been a return to optimism for the potential success of gene therapy. Importantly, there has been a growing list of success stories for gene therapy. For example consider the treatment of blood disorders with gene therapy covering >100 patients presenting with six different diseases, and a progressive extension to >10 diseases such as hematological malignancies (8,21–24), retinopathies (25) and others (26–29). The importance of long-term monitoring for safety and efficacy cannot be overemphasized and data continue to be obtained from patients currently undergoing treatment. However, we must understand that biologic processing and regulatory complications enter in the case of cell-based therapies; that is ex vivo modification of cells resulting in a new drug will have important consequences in terms of manufacturing and distribution of these products. Similar challenges hold true for gene replacement approaches aimed at treating the vast majority of muscle fibers affected in the MD's, where currently adeno-associated viral vectors hold the most promise as a gene delivery vehicle. In this review, we present a current overview of recent clinical trial advances and promising preclinical data toward the treatment of musculoskeletal diseases with an emphasis on DMD.

LGMD2D-alpha sarcoglycanopathy

Limb-girdle muscular dystrophies (LGMD's) comprise 30% of all progressive muscular dystrophies and have benefited from a number of recent therapeutic advances. Of note, recombinant adeno-associated viral vectors (rAAV1) mediated gene replacement of α-sarcoglycan (α-Sg) in LGMD2D showed sustained gene expression for 6 months after direct intramuscular injection in two out of three patients (30), while in an earlier trial 4- to 5-fold overexpression was observed at 6 weeks (two subjects) and 3 months (one subject) (31). All subjects received a mild course (4 h pre, 24 and 48 h post) of methylprednisolone (2 mg/kg) as an anti-inflammatory agent. The patient who failed to exhibit satisfactory gene expression demonstrated an early rise in neutralizing antibody titers and T-cell immunity to adeno-associated viral vector (AAV) capsid—highlighting the importance of cautious selection of patients enrolled in gene transfer trials. In contrast to a recent DMD trial (see below), LGMD2D patients did not demonstrate an immune response against the transgene, likely due to missense mutations predominating in LGMD2D. In follow-up studies, the Mendell group at Nationwide Children's Hospital is transitioning to an isolated limb perfusion method (32), where they have administered vector to one patient without complications and will have performed another by the submission of this review (J. Mendell, personal communication). This bodes well for LGMD gene therapies, as future trials can combine immunomodulation to overcome targeted immune response against AAV-capsid proteins. Patients can also be screened for serum neutralizing antibodies and potential T-cell immunity using ELISpot assays against various AAVs to determine the serotype best suited for successful gene transfer in individual patients.

LGMD2B-dysferlinopathy

Recent studies optimizing rAAV-mediated delivery of full-length human dysferlin in dysferlin-null mice suggest promising future therapies for LGMD2B, Miyoshi Myopathy (MM1) and distal anterior compartment myopathy (33). The latest strategy utilized two rAAV vectors, each carrying a discrete segment of the dysferlin cDNA containing a 1 kb overlap region, that greatly facilitated homologous recombination in vivo to reconstitute full-length dysferlin protein. Expression levels of dysferlin were superior to previous efforts delivering partially packaged 5′ and 3′ fragments, (34) reaching up to 27% of normal levels at high doses while fully restoring membrane repair and specific force. A similar approach had previously been used by our group to generate mini-dystrophin proteins too large to be delivered by a single rAAV vector (35). These results are in line with earlier studies demonstrating that restoration to 30% of WT dysferlin expression is required for clinically meaningful outcomes (36). Importantly, intramuscular injection of the vectors into non-human primates (NHP's) resulted in robust expression of dysferlin without inducing toxicity, inflammation, muscle fiber necrosis or an immune response to the capsid or transgene. Most recently, an investigational new drug (IND) application by Mendell and colleagues has been approved by the FDA using dual vector delivery of dysferlin for the treatment of LGMD2B (J. Mendell, personal communication).

LGMD2I and MC1CD-FKRP deficiency

AAV-mediated delivery of the gene encoding fukutin-related protein (FKRP) has been explored for treatment of LGMD2I. This disorder is most commonly caused by a C825A (Leu276Ile) mutation in the FKRP gene resulting in hypoglycosylation of α-dystroglycan. The homozygous L276I knock-in (KI) mouse model was recently generated and displayed a late onset, mild muscular dystrophy (MD) similar to that in patients (37). Systemic gene transfer using rAAV9-FKRP in neonatal and adult homozygous mice at 9 months of age effectively restored muscle morphology and function after 3 months and reduced fibrosis in various hindlimb muscles. Targeted treatment of the heart in neonatal L276I KI mice also appeared to restore some cardiac function 7 months post-treatment, although this model displays minimal cardiac dysfunction. Systemic gene transfer of rAAV9-FKRP to KI mice expressing a different mutation in FKRP (Pro488Leu; a model of congenital muscular dystrophy type IC, MDC1C) likewise restored α-dystroglycan glycosylation and ameliorated dystrophic pathology in heart and skeletal muscles (38). Because the precise localization of the endogenous protein is unclear and no reliable commercial antibodies are available for monitoring FKRP expression, we recently generated rabbit antisera against the human and mouse FKRP C-terminal region (Fig. 1). These antibodies have been used to demonstrate that functional, high level expression of human FKRP can be achieved in wild-type and FKRP mutant mice via systemic rAAV6 infusion (Fig. 1; B. Rodgers, J.S., and J.S.C., unpublished data). Together these data highlight the therapeutic potential of AAV-mediated FKRP delivery for treating both mild and severe muscle disorders caused by FKRP gene mutations.

Figure 1.

Generation of a custom rabbit anti-FKRP C-terminal antibody for detection of rAAV6-mediated muscle-specific expression of FKRP. (A) Schematic representation of the AAV6 vector genome for expression of FKRP in striated muscles. CK8: miniaturized mouse creatine kinase regulatory cassette (39–41, 85, 86); Kozak: kozak sequence, ACCATGG; FKRP: FKRP cDNA and pA: SV40 poly-adenylation signal. (B) Transverse micrograph of C57Bl/6 tibialis anterior (TA) muscle post-injection of rAAV6-CK8-FKRP (shown in A). FKRP foci (green, rabbit anti-FKRP) colocalize with golgi-specific antibody (red, GM130) in injected (white arrows) and WT control sections (inset, white arrowheads). (C) Longitudinal micrograph demonstrates FKRP-golgi co-localization throughout the myofiber. Additional FKRP-positive staining, not associated with golgi, was observed in injected muscle (asterisks) and did not colocalize with markers of nuclear membrane (D; Nucleoporin P62) or late endosomal compartment (E; EEA1).

Gene therapy for DMD using AAV

There are currently three approaches moving toward clinical trials for AAV-mediated gene therapy of DMD. These include delivery of antisense oligonucleotide (AON) cassettes to induce pre-mRNA exon-skipping, miniaturized ‘microdystrophin’ (μDys) genes, and surrogate genes to substitute or compensate for the absence of dystrophin. The first two strategies are methods to restore dystrophin expression, thus addressing the root cause of DMD: the lack of dystrophin. Using AAV vectors to deliver AON expression cassettes builds upon promising methods being developed by several groups to restore a dystrophin mRNA open reading frame by delivery of synthetic small molecule AONs (42–46). An impressive preclinical study in the canine model for DMD was recently published whereby AAV8 was used to restore moderate to high levels of dystrophin in forelimbs of treated animals (47). In contrast to small molecule AON delivery, the AAV method is more efficient for targeting of the heart and could lead to long-term gene expression, but requires high vector doses. The second method being developed for dystrophin replacement involves AAV-mediated delivery of μDys expression cassettes. Microdystrophin is used as opposed to larger or full-length dystrophins due to the limited carrying capacity of AAV vectors (48). Such μDys proteins carry the major protein interaction domains of dystrophin but lack much of the central rod domain and the carboxy-terminal domain. This method can target all muscle groups in the body, should lead to long-term dystrophin expression and could be applied to any patient with DMD or the milder Becker MD. However, as with the AON/vector approach high vector doses will be necessary to obtain systemic gene transfer. The potential for systemic delivery of AAV-μDys was originally demonstrated by our group and has been refined in many subsequent studies (49). Recent publications in the canine model of DMD support the feasibility of this approach (50–52), and several groups have initiated or are planning human clinical trials.

One clinical trial was previously performed using intramuscular injection of a μDys cassette regulated by the ubiquitously active cytomegalovirus (CMV) promoter. While little to no exogenous gene expression was obtained from the AAV2.5 vector used, evidence of a T-cell immune response against some dystrophin epitopes was found in both injected and non-injected patients (53,54). Whether these immunoreactive T-cells contributed to poor dystrophin expression is unclear; however, future studies will likely benefit from the use of muscle-specific gene regulatory cassettes (RC's) to drive expression of μDys [e.g. MCK RCs (39)]. A modified study by Mendell and colleagues using intramuscular injection of an AAV-MCK-μDys vector to the extensor digitorum brevis muscle has been initiated at Nationwide Children's Hospital (J. Mendell, personal communication; NCT02376816).

In recent preclinical studies designed to develop systemic μDys delivery methods, the groups of Dickson and Duan (G Dickson and D Duan, personal communication) have used AAV vectors to achieve robust expression of μDys in forelimbs (using AAV8, Dickson) or bodywide (using AAV9, Duan) in the canine model of DMD. Both the Mendell and Dickson groups are using vectors to produce an early generation μDys protein carrying hinge 2, although that construct has been shown to lead to ringbinden in mouse studies (55). The Duan group is delivering a recently developed μDys with an enhanced ability to generate force ((56); J.S.C., unpublished). Both the Dickson and Duan groups achieved widespread μDys expression in a range of muscle biopsy samples, and functional improvement without adverse toxicity or immune complications. Both groups plan to move into human trials in collaboration with Genethon (Dickson et al.) or Solid GT (Duan, Chamberlain et al.).

The third approach to AAV-mediated gene therapy for DMD builds upon earlier observations that some DMD patients display pre-existing dystrophin-specific cytotoxic T cells (53,54). Thus, some patients could be pre-immunized against a subset of dystrophin epitopes, and in theory this cytotoxic T-lymphocite (CTL) activity might be boosted by expression or presentation of exogenous dystrophin proteins in immune effector cells. While no direct data exist that pre-existing dystrophin immunity can attenuate expression of AAV-delivered μDys, an alternative approach to gene therapy might be to deliver a surrogate gene to substitute for dystrophin. Several promising surrogate genes are being explored for such an approach, including sarcospan (57,58), CT GalNAc transferase (59,60) and alpha7-integrin (61). Similar to dystrophin homolog utrophin, α7-integrin is upregulated in the sarcolemma in DMD, and in recent studies rAAV-mediated overexpression of human α7-integrin by isolated limb perfusion in the mdx and mdx/utrn^−/− mouse models of DMD improved a number of morphological and physiological parameters (62,63). AAV-mediated delivery of the cytotoxic T-cell GalNAc transferase (GALGT2) resulted in expression of GalNAc across the entire muscle membrane as well as having the added benefit of utrophin upregulation and resultant functional improvement (59,60). Other groups are addressing the issue of transgene immunogenicity through transient immune suppression (64–67), plasmapheresis (68), implementation of tolerization protocols (69–71) or reverse vaccine technologies (72,73). An additional approach using AAV involves delivering expression cassettes for Follistatin, which does not replace dystrophin but can increase muscle mass and strength by activating Akt/mTOR/S6K signaling (74). Like DMD, the allelic Becker muscular dystrophy (BMD) is caused by mutations in the dystrophin gene. Unlike DMD, BMD mutations often maintain the reading frame providinga partially functional protein characterized by a slowly progressive phenotype of primarily the legs and pelvis. A human trial of AAV-Follistatin delivered intramuscularly to the quadriceps muscles of BMD patients led to an increase in the 6 min walking time for four of the six patients (75). A related clinical trial has been initiated by Mendell et al. using a slightly higher dose of vector in DMD patients (J. Mendell, personal communication) (Table 1).

Table 1.

Current clinical trials in DMD/BMD registered or updated in the last year (up to 14th of July 2015)

Disease	Intervention or study-type	Sponsor	Clinical trial	Clinicaltrial.org ID	Purpose	Status
DMD	SRP-4045, -4053	Sarepta Therapeutics	Phase 3	NCT02500381	Treatment	Not yet open
DMD	CAP-1002	Capricor Inc.	Phase 1/2	NCT02485938	Treatment	Not yet open
DMD	Allogenic MSCs	Univ. of Gasiantep, Turkey	Phase 1	NCT02484560	Treatment	Ongoing
DMD/BMD	Cardiac evaluation	Univ. Childrens, Zurich	Not provided	NCT02470962	Observational	Recruiting
DMD	Pathological evaluation	Univ. Hospital, Brest	Not provided	NCT02472990	Observational	Not yet open
DMD	CAT-1004	Catabasis Pharmaceuticals	Phase 1/2	NCT02439216	Treatment	Recruiting
DMD	Upper limb evaluation	Catholic Univ. of Sacred Heart	Not provided	NCT02436720	Observational	Ongoing
BMD	Sodium nitrate	Cedars-Sinai Medical Center	Phase 1	NCT02434627	Treatment	Not yet open
DMD	Exercise dosing	University of Florida	Not provided	NCT02421523	Observational	Not yet open
DMD	2-D strain evaluation	University Hospital, Montpellier	Not provided	NCT02418338	Diagnostic	Recruiting
DMD	Eteplirsen	Sarepta Therapeutics	Phase 2	NCT02420379	Treatment	Recruiting
DMD/BMD	ACE inhibitor	InCor Heart Institute	Phase 3	NCT02432885	Treatment	Completed
DMD	rAAVrh74.MCK.μDystrophin	Nationwide Childrens Hospital	Phase 1	NCT02376816	Treatment	Recruiting
DMD	SMT C1100	Summit Therapeutics	Phase 1	NCT02383511	Treatment	Ongoing
DMD	Translarna	PTC Therapeutics	Not provided	NCT02369731	Treatment	Not yet open
DMD	Eplerenone/Spironolactone	Ohio State University	Phase 3	NCT02354352	Treatment	Recruiting
DMD	rAAV1.CMV.huFollistatin344	Nationwide Childrens Hospital	Phase 1/2	NCT02354781	Treatment	Recruiting
DMD	PRO044	Biomarin Nederland BV	Phase 2	NCT02329769	Treatment	Recruiting
DMD	SRP-4053	Sarepta Therapeutics	Phase 1/2	NCT02310906	Treatment	Recruiting
DMD	Elec. Impedance Myography	Skulpt Inc.	Not provided	NCT02340923	Observational	Recruiting
DMD	Deflazacort	Marathon Pharmaceuticals	Phase 1	NCT02295748	Treatment	Recruiting
DMD	Umbilical cord MSCs	Acibadem University	Phase 1/2	NCT02285673	Treatment	Recruiting
DMD	PF-06252616	Pfizer	Phase 2	NCT02310763	Treatment	Recruiting
DMD	Eteplirsen	Sarepta Therapeutics	Phase 2	NCT02286947	Treatment	Ongoing
DMD	Home mechanical ventilation	ResMed	Not provided	NCT02315339	Observational	Recruiting
DMD	Eteplirsen	Sarepta Therapeutics	Phase 3	NCT02255552	Treatment	Recruiting
DMD	Deflazacort	Marathon Pharmaceuticals	Phase 3	NCT02251600	Treatment	Recruiting
DMD	Stem cell therapy	Neurogen Brain & Spine Inst	Phase 1	NCT02241434	Treatment	Recruiting
DMD	TAS-205	Taiho Pharmaceutical C., Ltd.	Phase 1	NCT02246478	Treatment	Ongoing
DMD	Umbilical cord MSCs	Allergy & Asthma Consultants	Phase 1	NCT02235844	Treatment	Recruiting
DMD	Tadalafil	Cedars-Sinai Medical Center	Phase 4	NCT02207283	Treatment	Ongoing
DMD	Myoblast transplantation	Centre Hosp. Univ. de Quebec	Phase 1/2	NCT02196467	Treatment	Recruiting
BMD	l-citrulline and Metformin	Univ. Hosp., Basel, SZ	Phase 2	NCT02018731	Treatment	Recruiting
DMD	HT-100	Akashi Therapeutics	Phase 1/2	NCT01847573	Treatment	Recruiting

AAV-capsid immune considerations

In certain patients, immune responses against the AAV-capsid protein can reach significant amounts regardless of the route of delivery (76,77). A number of factors may contribute toward an ineffective outcome following AAV gene transfer in muscle tissue such as the inflammatory status (or immune experience) within the tissue (or patient) (78), potential major histocompatibility complex (MHC) I upregulation (30,31) or silencing of reactive T cells by programmed cell death (79). Such scenarios may be of particular concern if AAV readministration (80,81) is required to supplant a transgene that is lost over time. To date the half-life of an AAV vector in striated muscle is unknown, but it is encouraging that long-term transgene expression within skeletal muscle was achieved in a human patient after intramuscular AAV delivery (82). Although perhaps partially dependent on AAV serotype, immune reactivity risks could be minimized by a transient immune suppression protocol until the viral capsid is cleared from the patient (64,83). In the clinical setting, additional studies may be warranted to determine under what conditions immune suppression provides significant benefits.

Gene editing for muscular dystrophy and other recent advances

As described above, significant progress has been made in developing gene therapy approaches for DMD using AAV vectors to deliver μDys genes. An alternative and potentially longer lasting approach for therapy could be achieved by targeted correction of dystrophin gene mutations. The highly tunable bacterial clustered regularly interspersed short palindromic repeats (CRISPR)/CRISPR associated protein 9 (Cas9) system has emerged as an efficient method for gene editing in a variety of types of mammalian cells (16,84). The potential of the CRISPR/Cas9 system to restore dystrophin expression in DMD models has been demonstrated in human patient-derived induced pluripotent stem cells (iPSCs) and in murine iPSCs and isolated zygotes (85,86). Intriguingly, a dual AAV vector strategy expressing Cas9 and DMD mutation target specific guide RNA's revealed significant restoration of dystrophin expression upon direct injection into skeletal muscles of mdx^4cv mice, resulting in a majority of the myofibers displaying dystrophin expression in treated muscles within 3 weeks of vector delivery (N.B. and J.S.C., unpublished data) (Fig. 2). While the efficacy and safety of utilizing the CRISPR/Cas9 system in patients will require careful evaluation, early observations show promise and indicate that CRISPR/Cas9 might be adaptable for permanent, in vivo correction of DMD patient mutations in the future.

Figure 2.

Dystrophin restoration following rAAV6-mediated delivery of CRISPR/Cas9. (A) rAAV nuclease vector containing a muscle-specific regulatory cassette based on the mouse muscle creatine kinase promoter/enhancer element (CK8) (39–41, 88, 89), nuclear localization signals (NLS) and S. pyogenes Cas9. (B) The rAAV targeting vector containing the POL III U6 promoter driving each guide RNA (gRNA) and the intervening CMV promoter/enhancer driving a mCherry reporter gene. (C) Schematic representation of relative exon targeting in the mdx^4cv mouse and approximate gRNA positions to generate the exon Δ52–53 Dmd gene. (D) A representative 6-week-old male mdx^4cv mouse having received direct injection of rAAV-dual vectors into the tibialis anterior (TA) muscles. (E) Non-injected mdx^4cv control. TA muscles were analyzed via immunofluorescence using antibodies raised against the C-terminus of dystrophin (green) and nuclei (DAPI, blue).

While not a MD, a recent clinical trial for SMA type 1 has been initiated with important implications for a variety of neuromuscular disorders. In that trial, conducted by Brian Kaspar and Jerry Mendell at Nationwide Children's Hospital, several patients were infused with one of two doses of AAV9 vectors expressing an SMN expression cassette. The higher dose used (2 × 10¹⁴ vector genomes/kg) is in the range likely needed for systemic AAV-mediated gene delivery for a variety of different neuromuscular disorders, including DMD. To date no serious adverse effects resulted from vector infusion. However, the high vector dose led to elevated liver transaminases and a CTL response against vector. The transaminase elevations were found to be controllable with corticosteroid treatment commencing with vector delivery, and all patients treated to date appear to be doing well with increases in functional activity as measured by the Childrens Hospital of Philadelphia Infant Test of Neuromuscular Disorders (CHOP-INTEND) scoring system (Jerry Mendell, personal communication).

Closing thoughts

For DMD and numerous other MDs, impressive advances have recently been made that increase the prospects for gene-based treatment of these devastating diseases. Recent successes with AAV-mediated gene therapies for a variety of diseases have spurred renewed interest from both biotechnology companies and academic researchers to advance these methodologies in the clinic. New technologies and ingenuity continue to advance these therapeutic goals with tools such as CRISPR/Cas9, exon-skipping strategies and inducible, autologous stem cell approaches offering tremendous promise in the years to come.

Funding

This work was supported by NIH grants U54AR065139 and R01AR040864 (to J.S.C. and G.L.O.). J.T.S. is supported by the Sir Keith Murdoch Fellowship from the American Australian Association. J.K.H. is supported by NIH training grant (5T32 GM 007454-39) through the UW Department of Medical Genetics.