1 2 aim

#apaperaday: Potential limitations of microdystrophin gene therapy for Duchenne muscular dystrophy

In today’s #apaperaday, Prof. Aartsma-Rus reads and comments on the paper titled: Potential limitations of microdystrophin gene therapy for Duchenne muscular dystrophy

Today’s pick is a mouse study to compare 3 micro-dystrophins that are currently in clinical development for Duchenne by Hart et al in JCI insight DOI 10.1172/jci.insight.165869

Gene therapy aims to provide a functional genetic code for a protein that is missing in a genetic disease. For Duchenne this would be dystrophin. However, the genetic code of dystrophin is rather big and the only viral vector able to deliver to muscle and heart (AAV) is small.

Thus micro-dystrophins were engineered that contain crucial domains of dystrophin and lack ‘redundant’ domains. Some personal comments about this. Authors suggest micro-dystrophins aim to convert Duchenne into a Becker type of pathology. However, this likely won’t happen.

For one Becker dystrophins are much larger and also intervention happens at a later timepoint for now, when damage has already accumulated. Likely the micro-dystrophins will still be somewhat functional in slowing pathology and as such there would be therapeutic benefit.

However, inflating potential therapeutic effects leads to unrealistic expectations with Duchenne families. Micro-dystrophins will at best be partially functional: ‘redundant’ domains have specific functions as fulllength dystrophin is more functional than Becker/micro-dystrophin.

Back to the paper: authors argue correctly that there will be dilution of microdystrophin protein levels when patients grow and when there is muscle turnover (given that the micro-dystrophin is only partially functional there will be muscle turnover just like in Becker).

To allow having functional levels of microdystrophin the goal therefore is to deliver as much micro-dystrophin gene codes as possible to skeletal muscle and heart. Authors here wanted to study the longer term effects on this in the d2/mdx mouse model.

This mouse model is more severely effected at early stages of life, but in our hands skeletal muscle pathology recovers with time and mice also develop cardiac problems – even wild types. However, the wild type colony authors use do not seem to develop cardiac pathology with age.

Authors study 4 different micro-dystrophins: the Sarepta, the Solid, the Pfizer and a shorter version of the Pfizer one (lacking hinge 3). For all micro-dystrophins they use AAV10 to deliver (different from clinical development: Sarepta uses AAV74 and Solid and Pfizer AAV9).

Authors use the CK8 promotor, which is used by Solid, but not by Sarepta & Pfizer. Authors use the dose they claim is used in clinical trials, i.e. 2.10(14) genome copies per kg. However, I believe that what is used in clinic(al)trials is viral particles per kg is used.

As not all viruses will contain a micro-dystrophin gene this means that the authors here use a higher dose than in patients. Mice were treated with intravenous infusion of the different micro-dystrophins at 1 month of age.

The mice with Sarepta and small Pfizer micro-dystrophins (I’ll call this small MD from now on, as it is otherwise confusing) were analyzed after 12 months, as they started dying, while those with Solid and Pfizer micro-dystrophins were analyzed after 18 months.

Expression levels of the micro-dystrophins were very high: 5 times more than full length dystrophin in skeletal muscle and 55 times more in heart, with the Sarepta micro-dystrophin expressed even higher. This resulted in improvement in pathology and function, with better diaphragm function & reduced fibrosis & reduced damage after eccentric exercise in the EDL muscle. The Pfizer micro-dystrophin seemed less effective than others. However, none of the micro-dystrophin resulted in full normalization to wild type levels (as expected).

For the heart the Sarepta and small MD were not effective and might actually do worse than the untreated d2/mdx mice, and mice started dying. The Solid and Pfizer micro-dystrophins did improve cardiac pathology and reduced fibrosis.

Authors studied the reason for the Sarepta and small MD resulting in poor heart performance. They found these micro-dystrophins outcompeted utrophin that is also expressed & this reduces functionality (as full length utrophin apparently is more functional than micro-dystrophin).

Furthermore, the huge overexpression of micro-dystrophin causes problems with protein waste disposal of the cells (the excess amount of micro-dystrophins and misfolded micro-dystrophins), which is known to cause toxicity.

So what does this mean for micro-dystrophins used in Duchenne patients? Authors suggest longer term follow up may reveal cardiac problems for Sarepta micro-dystrophin (note that for the Pfizer one they saw only problems with the smaller version which is not used in the clinic).

Monitoring patients who received gene therapy is a good idea as this is a pioneering therapy and we do not know the long term effects. However, I do not know whether the cardiac problems authors see here will translate to patients or translate to the same extent.

Several reasons:

  1. Humans are not mice, but also:
  2. may differences between what authors tested and what is used in trials:
    1. The serotype of AAV is different.
    2. The d2/mdx mouse is very prone to developing cardiac problems even the wild types, so they may be more sensitive.
    3. The micro-dystrophin levels expressed in heart are extremely high in the mice. Likely this will be much lower in humans, because: the dose is lower (about 50% of viruses are empty for Elevidys, so the dose is at least 50% reduced compared to what authors use).

Furthermore, heart is transduced very efficiently by AAV IN MOUSE, but we know this is not so in non human primates (same level as skeletal muscle). For humans we have limited data, but we know from https://pubmed.ncbi.nlm.nih.gov/38055269/ that AAV9 levels in heart & skeletal muscle were similar.

So this may be something mouse specific and likely micro-dystrophin levels in humans are more similar to what is seen in skeletal muscle. Again, monitoring patients is a good suggestion always because we do not know what to expect long term.

However, authors could have elaborated somewhat more on the limitations on their study and how they differed from the human situation to explain why the severe cardiac pathology they saw in the mice might not translated to the human situation.