Gene Therapy for DMD

Delivery of the gene to muscle
We have a lot of muscle – about 30-40% of our body weight is muscle, and we have more than 750 different muscles, each consisting of billions and billions of cells. The healthy gene has to be delivered to a significant portion of the cell nuclei of all muscles.

Fortunately, there is an organism that is quite good at injecting genes into cells: the virus. Thus, the gene therapy field has developed viral vectors, where the viral genes are removed, so there is room for the new gene and the modified viruses cannot replicate themselves.

Finding the right viral vector: AAV

Most viruses like to infect dividing cells. Muscle tissue hardly divides and thus is a poor target. In addition, muscle fibers are enveloped by layers of connective tissue, which trap viral particles, so the virus cannot reach the muscle fiber to inject its dystrophin gene.

There is a virus that is relatively good at infecting muscle cells, the so called adeno-associated virus (AAV). This virus can infect human cells but is not pathogenic (it does not cause a disease, about 50% of people have been infected by an AAV in the past without realizing this). There are many different types of AAV (called serotypes). For Duchenne serotypes that preferentially go to skeletal muscle and heart are used.

Note that these AAVs do not exclusively go to heart and muscle, a large amount also goes to the liver. Also while AAV does not cause a disease during a natural infection, with gene therapy a huge amount of viruses is injected, which can cause (severe) side effects.

Making the dystrophin gene code fit in AAV
Unfortunately, AAV is so small that the dystrophin genetic code does not fit (the entire gene is ~500 times to big, the genetic code ~4 times too big).

Scientists have attempted to create the smallest possible dystrophin, containing only the bare essential domains (micro-dystrophin). The genetic code of this micro-dystrophin is small enough to fit into the AAV vector. This micro-dystrophin contains only ~32% of the full-length dystrophin protein and is smaller than the dystrophins that are found in Becker patients. In the Duchenne mouse model (mdx mouse) treatment with microdystrophin containing AAV viruses resulted in an improved muscle quality and function.

Immune response to AAV

Since the body does not know the AAV-micro-dystrophin has good intentions, there will be an immune response to the viruses.

To temper the immune response, high doses of corticosteroids are used when AAV-microdystrophin is delivered. Despite this, there will still be an immune response to AAV due to the extremely large amount of virus particles used for gene therapy (100,000 billion viral particles per kg bodyweight).

Many individuals have been infected with a subtype of AAV (serotype). These individuals have antibodies against AAV that would preclude them from receiving viral vectors of that specific subtype. Since huge amounts of viruses are used for AAV-microdystrophin gene therapy, the immune response would be very dangerous and potentially lethal. The percentage of individuals with AAV antibodies varies per region (e.g. 20-50% in the Western world but higher in Asia) and per serotype (e.g. more for AAV2 and AAV9 than for AAV74). Ways to avoid a response of the antibodies to the AAV particles are under investigation, e.g. one could filter out the antibodies (a process called plasmapheresis). For now, Duchenne patients with antibodies against AAV are excluded from clinical trials with AAV-micro-dystrophin or treatment with micro-dystrophin.

In addition to an immune response to the AAV, some patients develop an immune response to the microdystrophin. See below for more information.

Components of a gene therapy viral vector

The gene therapy product consists of the viral vector (AAV) and the gene that is delivered. For Duchenne this is a micro-dystrophin gene with a volume switch (promotor), that makes sure the micro-dystrophin is only produced in muscle tissues.

There are currently trials ongoing with 5 variations of AAV-micro-dystrophin, which differ in the type of AAV used, the parts of dystrophin selected for the micro-dystrophin versions, and the gene switch used to ensure proper expression of the micro-dystrophin in skeletal muscle and heart.

Clinical development summary

So far >600 Duchenne patients have been treated with AAV micro-dystrophin in clinical trials or commercially. Detailed findings of each trial and results is given below per company. However, general findings are as follows:

Micro-dystrophin is restored after treatment, in 5-80% of muscle fibers at levels of 1-90% of normal. Higher doses of virus result in higher levels of micro-dystrophin.

AAV gene therapy can result in severe side effects, such as extreme nausea and vomiting, liver damage, inflammation of the heart (myocarditis), transient kidney failure and death. One type of side effect (thrombotic micro-angiopathy, TMA) seems to occur primarily for AAV9, because this type of AAV can trigger a specific part of the immune system (complement activation). TMA is associated with renal failure and reduced platelets and problems with red blood cells. Sadly, so far 4 Duchenne patients have passed away after receiving AAV therapy. All patients received AAV9.

The likelihood of side effects increase with the amount of virus used. As older patients generally are heavier and thus receive more viral particles, this means older patients are more likely to receive side effects. However, older patients also have a reduced capacity to cope with severe side effects due to respiratory and cardiac pathology.

Immune response to micro-dystrophin

For a subset of patients an immune response to the micro-dystrophin occurred resulting in rhabdomyolysis (breakdown of muscle) and myocarditis (inflammation of the heart). This appears to be associated with deletions at the beginning of the gene.

Duchenne patients do not produce dystrophin and the immune system recognizes and attacks foreign proteins. So in that sense, it is not strange that an immune response occurs. However, why does it only occur in the patients with a deletion at the start of the protein and not for all patients?

The phenomenon is still being studied but the current consensus on how patients with deletions at the start of the gene can have an anti-microdystrophin response while other do not, is as follows. Most patients have deletions and most of these deletions occur in the middle of the gene, which encodes the repeat domain. The micro-dystrophin has the two crucial domains of the protein: the actin-binding part and the part that binds to the extracellular matrix. For patients with a deletion in the middle of the gene, the micro-dystrophin components will not be foreign. Their cells will have produced the beginning of the protein (up till the deletion). This protein is not functional and not stable, but it is produced so the immune system ‘knows’ it. These patients will also know the second component of the micro-dystrophin binding to the extracellular matrix. In the brain different dystrophins are formed including a very short one that only contains the part that binds to the extracellular matrix. For patients with deletions before exon 62, this smaller dystrophin can be produced. As such, the micro-dystrophin as such is never seen, but the individual components have been encountered.

For patients with a deletion at the beginning of the gene, however, the first component will be foreign as they will not have produced this. Thus is it recognized by the immune system as foreign and attacked. Note that this recognition does not occur for all patients with deletions at the beginning of the gene but only a subset. It is not yet known why the immune system is triggered in some patients and not in others.

Furthermore, it is not known whether patients who cannot produce the last part of dystrophin will have an immune response to micro-dystrophin. These mutations are very rare.

Based on these developments companies have adapted their inclusion criteria. Pfizer excluded patients of a deletion involving exon 9-13 or 29-30, Sarepta/Roche excludes patients with a deletion involving exon 1-17 or exon 45. Solid likely will exclude patients with deletions involving exon 8-13 and 42-45. For the approved Sarepta product (delandistrogene moxeparvovec), the FDA approval specifies that patients with a deletion involving exon 8 or 9 are excluded for treatment due to the risk of an anti-micro-dystrophin auto-immune response.

Clinical trials :

Sarepta/Roche (SRP-9001/delandistrogene moxeparvovec/Elevidys)

• Several clinical trials have been done with ElevidysIn a clinical trial in Colombus, Ohio (coordinated by Sarepta Therapeutics) AAV-microdystrophin was delivered intravenously in young Duchenne patients together with high doses of corticosteroids (200,000 billion viral particles per kg bodyweight). Four patients have been treated and long term follow up results have been published. No serious side effects were observed, but vomiting shortly after infusion was observed in 2 patients. Analysis of muscle biopsies 90 days after treatment revealed that for each patient the majority of muscle fibers expressed the micro-dystrophin. The total levels were ~75% of dystrophin levels expressed in healthy muscles.

• A placebo controlled trial In this trial 20 patients received SRP-9001 and 21 received placebo (to receive SRP-9001 after 48 weeks). Treatment was well tolerated, with most patients experiencing mild or moderately severe side effects. Severe side effects were reported for 4 treated patients (a severe breakdown of skeletal muscle (rhabdomyolysis – see ‘Immune response to micro-dystrophin) and increased transaminases (a sign of liver damage)) and 1 placebo patient (rhabdomyolysis). A biopsy taken 12 weeks after treatment revealed treated patients on average produced 20% of micro-dystrophin. The patients receiving SRP-9001 with a year delay (i.e. those starting as the placebo group) produced 40% of dystrophin. It was later discovered that this group received the intended dose, while those receiving SRP-9001 initially received a reduced dose, which likely explains why they produce lower amounts of micro-dystrophin. No significant improvement in the North Star Ambulatory Assessment (NSAA) was observed for treated patients compared to placebo.

• An open label study with commercially produced delandistrogene moxeparvovec (ENDEAVOUR study). This open label study is ongoing to confirm the commercially produced delangistrogene moxeparvovec. Results are published for 12 months analysis for the first 20 patients (4-8 years old). A biopsy taken after 12 weeks revealed 54% of micro-dystrophin. Function could only be compared to natural history studies. This suggests a slower disease progression. However, caution has to be taken with interpreting this, as the patients treated with delandistrogene moxeparvovec are temporarily treated with higher doses of corticosteroids, which can also slow down disease progression.

• A placebo-controlled trial tested safety and efficacy in ambulatory patients age 4-8 years (EMBARK). This is the ‘pivotal’ study that had to confirm that treatment with melandistrogene moxeparvovec slows down disease progression, as also outlined in the conditions from FDA when they approved the approach for 4-5 year olds. Results revealed no significant improvement on the primary end point, i.e. North Star Ambulatory Assessment scale for treated patients compared to placebo treated patients after a year. Significant improvements were seen for some of the secondary outcomes, such as the time to rise from the floor, 10 meter walk run and the stride velocity test.

Despite the confirmatory trial not meeting its primary endpoint, Elevidys received full approval from the FDA. Treatment is also available in several countries in the Middle-East. Evaluation by the European Medicine Agency is ongoing.

Several clinical trial are ongoing for Elevidys:

• A Placebo-controlled phase 3 clinical trial is ongoing in ambulatory and non-ambulatory patients (ENVISION).An open label study is ongoing in Europe in patients under the age of 4 coordinated by Roche.An open label study is ongoing in patients with preexisting immunity to AAV74 where extra immune suppression (imlifidase) will be used.For long term follow-up, all patients previously treated with delandistrogene moxeparvovec in a clinical trial setting are enrolled in an open label extension study. Here long term effects (beneficial and side effects) will be monitored.

Pfizer

In parallel, Pfizer is developing fordadistrogene movaparvovec (previously PF06939926). Multiple clinical trials have been done with this compound:

• A clinical trial assessing 2 different AAV-microdystrophin doses (using 0.5 and 1.5 times the dose used in the Sarepta trials) was completed. Three severe adverse events were reported, 1 patient experienced nausea and vomiting and 2 an immune response to the viral particles. One patient had TMA, which resulted in kidney injury The patient was hospitalized and required dialysis, but recovered and the kidney function is now normal. Based on these events, the monitoring protocol has been amended for this and other trials with fordadistrogene movaparvovec.

• Dystrophin staining analysis on a biopsy obtained 2 months after treatment revealed that ~38% of fibers expressed micro-dystrophin in patients treated with lowest dose and ~69% of fibers expressed micro-dystrophin for patients treated with the higher dose.  Mass spectrometry analysis revealed dystrophin levels of ~24% and 30% for the low and high dose, respectively. Since no western blot analysis was done a direct comparison with the results from Solid, Genethon, Regenxbio and Sarepta is not possible. Compared to an external natural history cohort, 16 treated ambulatory patients had better North Star Ambulatory Assessment scale and rise from floor performance 2 years after treatment.

• A phase 3 placebo-controlled  trial in 99 patients to further evaluate their micro-dystrophin AAV has been completed as well. This trial failed to meet its primary endpoint: after 1 year, treated patients had the same scores on the North Star Ambulatory Assessment scale as the placebo group. There was also no difference between secondary endpoints. Treatment patients did produce micro-dystrophin in their skeletal muscle. Based on these results, Pfizer decided to discontinue the clinical development of fordadistrogene movaparvovec. They will continue to follow the patients who received gene therapy treatment for assessment of long term safety in an open label study.

• The clinical trials in older ambulant and non-ambulant patients and in 2-3 year old patients were stopped as well with the discontinuation. Patient who had already been treated will be followed up for long term safety assessment in an open label study.

Solid

A trial coordinated by Solid Ventures  was put on hold after the first patient experienced serious adverse events after infusion of the AAV-micro-dystrophin. The trial reinitiated in June 2018 and the biopsies of the 3 patients receiving the lowest dose (4 times lower than the Sarepta trial) have been analyzed as well as biopsies from 3 patients receiving a higher dose (same as Sarepta trial). This revealed low levels of dystrophin in up to 10% of fibers at levels of less than 5% of normal dystrophin for the lower dose and dystrophin in up to 50% of fibers in levels of ~10% for the higher dose. Adverse events were reported for the first patient treated with the higher dose, including a decline in platelet counts and markers suggesting liver damage. Things normalized after the dose of corticosteroids was increased. The trial was put on hold for a second time by FDA in November 2019, but this has been lifted in October 2020. Since then a 3 more patients have been treated without experiencing severe adverse events.

Solid has prepared a new AAV-micro-dystrophin product (SGT-003), where the AAV vector has been modified for increased delivery to muscle. This is now evaluated in a clinical trial.

Genethon

Genethon has initiated their clinical trial with AAV8 micro-dystrophin gene therapy in Europe in 2021. So far 5 patients have been treated. In biopsies, micro-dystrophin levels of 15-85% were observed, and preliminary results suggest stabilization on the North Star Ambulatory Assessment scale for 1-2 years, while a decrease would be expected from natural history controls. Genethon expects to start a pivotal study with this compound in 2025

Regenxbio

Regenxbio has initiated clinical trial with their AAV8 micro-dystrophin gene therapy product RGX-202. So far 3 patients have been treated with a low dose and 2 with a high dose. The micro-dystrophin levels varied between 11 and 83%. Patients show an improvement in the North Star Ambulatory Assessment scale compared to baseline, while a decrease would be expected from natural history controls.

Regenxbio is planning a pivotal trial with their compound.

To Consider:

Discrepancy between functional effects in open label studies and placebo controlled studies

The phase 1/2 trials with gene therapy are generally open label studies where patients’ function is compared to baseline and to natural history controls, while in a phase 3 trial patients are compared to a placebo group.

When an improvement in the North Star Ambulatory Assessment scale is seen in the phase 1/2 trials, we need to take into account 2 things: 1. Young Duchenne patients will often improve in the NSAA scale due to growth and development; 2. Patients in gene therapy clinical trials will receive higher than normal steroids, which will also have an impact on muscle function. As such, the improvement on the NSAA scale may not be due to the gene therapy and definitely is not only due to the gene therapy.

In placebo-controlled trials the placebo group also receives high dose steroids so this is a better comparator. So far, micro-dystrophin gene therapy did not show an improvement of treatment on the NSAA scale. However, the trials were short (1 year) and also done in young patients, where improvements on the NSAA scale are expected. The therapeutic effect of the micro-dystrophin, however, is not to improve function, but to slow down disease progression. As such, the fact that no treatment effect was observed in 1 year trials in young patients does not mean that micro-dystrophin gene therapy is not therapeutic.

Longevity of AAV micro-dystrophin gene therapy

AAV does not integrate in the DNA. This is good from a safety perspective. However, because the microdystrophin is not fully functional, the muscles with microdystrophin will still be damaged. This means that over time the micro-dystrophin gene may be lost. Studies in dystrophic dogs suggest that most of the delivered gene is lost after 5 years. It is not known whether this is also the case in humans and if so how long the microdystrophin protein will persist. It is unlikely however that AAV-mediated microdystrophin delivery will be a ‘one and done treatment’ for Duchenne.

Past clinical trial results if intramuscular injections

A first clinical trial where patients received local AAV-microdystrophin injections in the arm muscle was performed in the USA (Mendell, Xiao Xiao and Samulski). Results of this trial have been published. The authors report very poor expression of their microdystrophin version, and the anticipated immune response to AAV.