MAIN LESSONS FROM THE 2020 CDKL5 FORUM

2020 is the year when the world discovered the urgency that the rare disease community feels to develop effective medications and cures.

It is also the year that prevented us from hosting large conferences, so the 6th edition of the CDKL5 Forum, which would have taken place in London, UK, instead took place October 12-14 in a virtual conference center. The Forum is the annual meeting hosted by the Loulou Foundation where scientists and drug developers working on CDKL5 Deficiency Disorder (CDD), together with representatives from patient organizations, meet to discuss the latest advances. This was the fourth Forum I attended, and the third since joining the Loulou Foundation.

Screen Shot 2020-10-12 at 14.53.54.png

Every year, the CDKL5 Forum allows us to look back in time to how the field was just a couple of years ago, review all the progresses made in the last year, and anticipate how the next year of research into CDD and drug development is likely to look like. And if we are to learn from our predictions from the past years, we will probably underestimate how much everything will move in just 12 months.

Dr Orrin Devinsky, Chairman of the Loulou Foundation Advisory Committee, summarized the 2020 Forum stating how “what we have seen these 3 days is breathtaking”. For those of you who were not in the room, these are my main conclusions from the CDKL5 Forum and how far along we have come in the CDD field.

1. UNEXPECTED WONDERFUL BIOLOGY

The Forum always starts by updating our growing understanding of what the CDKL5 protein does and why mutations in the CDKL5 gene lead to such a dramatic neurodevelopmental condition.

I am a cell biologist and I am used to thinking about genetic conditions by knowing what the missing (or mutated) protein does. For example, “this is an ion channel so when it is missing neurons are hyperexcitable”, or “this protein is essential for protein degradation so when it is mutated there is abnormal buildup of toxic proteins”. But CDKL5 deficiency is challenging all those simple schemes. The CDKL5 protein is a kinase involved in so many cellular processes that learning about what happens when CDKL5 is missing is not only teaching us about how the deficiency disorder works, it is also teaching us new biology previously unknown to science.

In the words of Dr Victor Faundez from Emory University, which capture very well this take-home message:

““One mutant opened the door to wonderful biology”

We saw much of this during a pre-meeting with the Loulou Foundation awardees as well as during the first day of the Forum. Scientist from all around the globe shared their findings about how CDKL5 modulates microtubule dynamics (like the skeleton within cells), and how this is so critical for neuronal plasticity and formation of new synapses (neuronal connections). And CDKL5 also seems to have important functions at the cellular headquarter where DNA is located: the nucleus. They also told us how CDKL5 also regulates motile cilia, like in the epithelial cells that surround the brain ventricles and move CSF around, and about functions outside of the brain. And CDKL5 starts looking like a kinase that probably controls many other kinases, like a critical node in a network, so it takes us in many directions. From some of these pathways we start identifying “actionable” targets, which suggest possible drugs that might be used to at last ameliorate some of these consequences of missing CDKL5.

That said, it is clear that with CDKL5 controlling so many processes it will be hard to “bypass it” or achieve a full rescue with treatments that target individual pathways. We really should try to replace the protein or the gene to achieve the full recovery. The good news is that these treatments are all in development.

Some of the consequences of missing CDKL5 remind me to what goes wrong in other neurodevelopmental disorders. For example, all of the microtubule dynamic discussions make me think of KIF1A associated neurological disorder, and a recent publication showing that CDKL5 could be flagging proteins for ubiquitination in the kidney reminds me to Angelman syndrome, where a missing ubiquitin ligase is the culprit. So these all make sense as individual pathways that once de-regulated by the lack of CDKL5 might lead to the neurodevelopmental disorder, and they happen to all be engaged at the same time in CDD.

All because of the one mutant that opened the door to wonderful biology.

2. PACKING THE TRANSLATIONAL TOOLBOX

One of the reasons why so many scientists have come to work on CDKL5 is the biology that it uncovers. Another reason is research tools.

toolbox.jpg

Scientist cannot research a protein or a disease if they don’t have the tools to produce that protein, or to see its activity, or to model the disease in cells or animals. A big focus of the Loulou Foundation and our partner organizations has been to build “the translational toolbox” so that we have all we need to take a therapy all the way from understanding what happens in the cell to completing clinical trials. Last year the Forum dedicated the parallel breakout sessions to map out the needs for tools and to focus our efforts to create them. A year later, those efforts have paid off.

I will only highlight three of those tools, there were more presented:

Dan Lavery, CSO of the Loulou Foundation and our Forum host, announced that Coriell (a biomedical tool repository) will soon start offering iPSCs from CDD patients. This is the result of years of work in collaboration with the Boston Children’s Hospital that led to the creation of several lines of iPSCs together with isogenic controls. For the non-scientists: iPSCs are inducible pluripotent stem cells, a sort of stem cell that is obtained by taking some cells from the blood or the skin of a person (in this case CDD patients) and “reverting” them to stem cell state. They are very useful because scientists can then make them go forward in development, this time to neurons, and be able to have CDD neurons in a dish. And because they will be available from this repository, getting them will soon be as easy for scientists as placing an order on Amazon.

Another important tool to research CDKL5 is CDKL5 itself. Brian Ranes from Amicus Therapeutics offered to share purified CDKL5 protein with the research community, opening the door to exciting science that was until now very difficult to do because producing and purifying CDKL5 is remarkably complicated. But Amicus has developed considerable expertise around protein production, and showed once again their commitment to advancing research in CDD beyond just working on their own therapeutic programs.

One critical challenge in the development of therapies is reproducibility, in particular reproducibility of the preclinical (mouse) drug studies. A year ago, the Loulou Foundation signed a joint collaboration with Baylor College of Medicine and the Texas Children’s Hospital to establish a pipeline of CDD mouse models and behavioral assays to be able to evaluate (and confirm) potential therapeutics for CDD in a rigorous and unbiased way. The work is led by Dr Rodney Samaco, who was the recipient of the Lab of the Year 2020 award by the Loulou Foundation for the transformational value that this mouse behavioral assay platform will have for the development of therapeutics in CDD. This platform has already been instrumental for the preclinical proof-of-concept for the gene therapy from Ultragenyx that was also presented at the Forum (more information below).

Building the research toolbox takes many years, many scientists, much funding and much collaboration. The CDD translational toolbox is a reflection of the maturity of the CDD field.

3. CDD IS TECHNICALLY NOT A NEURODEVELOPMENTAL DISORDER

This provocative statement is one of the major conclusions that I take from the 2020 CDKL5 Forum.

Coming into this Forum we already knew that there are no large changes in neuronal wiring in CDD, or in brain anatomy, and we have not seen signs of it being a neurodegenerative disease. So it would appear that CDKL5 is constantly needed for the very dynamic process of synaptic plasticity and formation. This means that bringing CDKL5 expression back is likely to provide a benefit even in the more mature brains, while in diseases that affect neuronal migration, or that lead to neuronal death, we only have a narrow time window to replace the protein and see any improvement. But we really had no evidence for this yet.

During this Forum we learned from mouse studies that the CDKL5 protein remains highly expressed throughout life (so it is not only there during early development). We also learnt that if we remove it from the brain of an adult mouse that has otherwise developed completely normal, as if CDD could be started in an adult individual by flipping a switch, this results in all of the same problems that mice have if they are born with CDD. Joe Zhou from the University of Pennsylvania showed us how removal of CDKL5 expression in adult mice resulted in hyperactivity, reduced sociability and other autism-spectrum behaviors, memory problems and poor motor coordination (this is how CDD looks in mice, remember that it is very hard and rare to see seizures in CDD mice). Because removal of CDKL5 at a later stage is enough to produce the disease, Joe’s conclusion was that “CDKL5 is indispensable for life”.

But what about the reverse? What is the consequence of bringing CDKL5 expression back to a brain that has developed without it? Joe told us how he plans to do the reverse mouse, one where the animals are born without CDKL5 and then he can start the gene at different ages, and we all look forward to seeing those results. In the meantime, the first partial answers to these questions came from the gene therapy programs. Sharyl Fyffe-Maricich from Ultragenyx showed how administering AAV carrying a copy of CDKL5 to mice with CDD aged 3 to 5 weeks, which is the equivalent of children 4 to 11 years old, was able to improve several of the phenotypes of these mice, including motor function and cognition. While she made it clear that they were only able to restore CDKL5 expression to 35% of its usual levels, which is just a partial rescue, these experiments showed that at least in mice even partial rescue of CDKL5 expression is sufficient to reduce the disease symptoms, and this is the first time that we have been able to measure this directly.

What this tells us is that CDD is not strictly a neurodevelopmental disorder, where the CDKL5 protein is needed for brain development and no longer needed afterwards. Instead, CDD is a neuromaintenance disorder, where CDKL5 protein expression is needed throughout life to ensure proper neuronal functioning. We don’t appear to be fighting against a transient neurodevelopmental window when CDKL5 is needed and where the damage is already done if we come too late. And this is much more hopeful for treatments, because all of these experiments tell us that CDD is a good candidate for gene therapy or enzyme replacement therapy even in adult patients.

“CDD is not a neurodevelopmental disorder, but a neuromaintenance disorder, and this is much more hopeful for treatments.

4. GENE THERAPIES (IN PLURAL) ARE COMING

The topic of gene therapies was a major one in day 2 in the 2020 Forum. Jim Wilson, a veteran from the gene therapy field, explained us how gene therapies have to progress roughly through 3 stages before they can be used to treat large groups of patients in clinical trials. I will give them a name based on my own words:

–       Stage 1 – Prove that you can treat a mouse: the first treatment studies are often done in mice that carry the same genetic deficit that the patients have, to test the ability to correct the disorder in these animals with the experimental therapy and get an early feeling of potential toxicity issues. In gene therapy scientists use a virus that they have emptied from virus DNA and replaced by the therapeutic gene that is needed for the patient, so that when the virus infects the patient it is actually bringing the gene to the cells. So the biggest challenge of this stage is to produce that “transgenic therapeutic virus” which involves several pieces, and to treat the CDD mice with it.

–       Stage 2 – Prove that you can treat larger brains: these are the toxicology studies which are done in large animals, often non-human primates, looking for safety and ability to bring the virus to large parts of the brain. This is often the biggest challenge, the “biodistribution” of the virus, in particular for diseases that need to treat large areas of the brain (or other organs) as opposed to very focal diseases.

–       Stage 3 First in Human studies: this is when the therapy reaches clinic for the first time, and the first studies focus on safety and toxicology before risking exposing too many patients to the new experimental gene therapy.

Jim reviewed how the challenge for many gene therapies is step 2, and figuring out how to scale up from very small mice to large animals. For example, figuring out the route of administration. In the case of CDD, could we use intravenous administration or should we deliver the virus directly into the brain? Earlier in the program we had learned from Amicus that they are working on a project to boost the number of neurons that can benefit from gene therapy by using a virus to deliver not the plain CDKL5 gene but a gene encoding for a secretable CDKL5, so that the neurons that receive the virus and can express CDKL5 will also act as local factories to secrete CDKL5 that can then get into neighboring cells, an approach that they called “cross-correction”.

And then we got to one of the most exciting moments in the Forum, when we got to see two proof-of-concept in mice from two companies developing gene therapies for CDD.

The first one was from Jim Wilson’s group, who is running a gene therapy for CDD program for a company called Elaaj Bio. Ralf Smichd presented a very large collection of experiments optimizing the gene therapy for CDD. He showed us how much work went into creating the “transgenic therapeutic virus”, from seeing which of the different isoforms of CDKL5 is needed (there are 4!) to checking what happens if you overexpress CDKL5 in the brain (to know if we face an overexpression risk) to confirming activity in both males and females. And to run the proof-of-concept, this group chose to go for the best-case scenario, which is to treat CDD mice at birth. This ensures the greatest brain coverage for the virus as well as treating as early as possible. The best-case scenario in CDD mice was amazing. Ralf showed complete rescue of CDKL5 protein expression AND activity, because we can monitor the kinase activity looking at a reporter phospho-target called EB2. He showed us beautiful brain expression of CDKL5, and rescue of all the symptoms in these mice when examined months after the virus administration into their CSF. He even showed some preliminary EEG improvement, and had reproduced these data in a separate lab with two additional CDD mouse strains. You can hardly get more solid science than here.

This gene therapy is now being optimized for the stage 2 that Jim explained to us before, which is to tweak it until it has good biodistribution in the non-human primate (large) brain so that they can confirm toxicity and safety and best dosing in these large animals before thinking about trials.

And as I commented before, Ultragenyx also presented a mouse proof-of-concept with their own gene therapy, which earned them the 2020 CDKL5 Forum Award for the “Company making a difference” in the preclinical space. And I found very interesting that Ultragenix had chosen a different approach for designing their mouse proof-of-concept. Instead of going for the best-case scenario, as in the previous therapy, Ultragenyx chose to go for the real-life-scenario, which was to treat older mice with a partial rescue of gene expression. And Sharyl showed us how even partial rescue with the virus administered directly into the CSF of CDD mice at the equivalent human age of roughly 4 to 11 years old resulted in partial efficacy across different symptoms. The results were therefore tremendously encouraging because we are likely to achieve only partial rescue in the clinic and we will be treating patients that are not newborns anymore. So the question of “is this sufficient” is indeed very relevant, and the answer – at least in mice – is “yes”.

So in 2020 we have proof by two companies that we can rescue CDD in mice, with improvements even in older mice, and the discussion has now moved to how to scale this up to work in larger brains. We have made it past stage 1.

These programs open the path for follow up therapies based on gene editing, the technology that earned the Nobel Prize for Chemistry in 2020. At the Forum we reviewed several of these programs, including a modified CRISPR approach to try to “unlock” the inactive CDKL5 in the second X chromosome of each cell, as well as “prime editing” which is a field pioneered by David Liu who explained us that CDD is one of the first diseases that they are trying to correct with prime editing. There are so many therapeutic programs in development for CDD that it would be hard to list them all in this type of update. What is important to know is that pretty much every therapeutic technology that we know already has an active program (or more) to apply  it to CDD, so the question is no longer whether we will ever get a curative treatment, but instead it is “which of them” and “when”.

5. 2020 THE YEAR OF CLINICAL TRIALS (ALSO IN PLURAL) SUCCESS

Day 3 of the 2020 CDKL5 Forum came with many great news. Dr Orrin Devinsky summarized the data with Epidiolex in CDD as part of an expanded access program that showed efficacy, and then announced for the first time the results of two small investigator-initiated phase 2 studies (those are “one hospital” clinical trials) that he has led:

1.     Ataluren did not have efficacy in CDD. Orrin showed us the data of a study with 18 patients, some of them with CDD some of them with Dravet syndrome, in which some patients started with ataluren and then continued with placebo and others did it the other way around (a placebo-controlled crossover study). Ataluren is a molecule that helps the cell skip premature stop codons caused by non-sense mutations, so all of these patients had non-sense mutations. Ataluren did not show an effect in reducing seizure frequency or improving cognitive, motor, or behavioral function or quality of life in subjects with either Dravet Syndrome or CDKL5 deficiency due to nonsense mutations. This randomized, placebo-controlled trial was limited by a small sample size and a treatment-phase of 12 weeks, which may be too short to identify a disease-modifying effect. .

2.     But fenfluramine does have efficacy in CDD! Fenfluramine is an anti-epileptic drug that works through the serotonin system and that has been approved in the US and Europe for the treatment of Dravet syndrome, another complex neurological condition. Orrin is running a 10-patient study with fenfluramine in CDD at NYU, and was able to present the results from the first 6 patients. And the results are very good, with a 90% reduction in generalized tonic-clonic seizures that are the strongest seizure type, and also efficacy in other seizure types. He still has 4 slots open for patients with CDD ages 2 to 18, and encouraged the company developing fenfluramine to consider CDD for a phase 3 trial.

And the phase 2 study with fenfluramine was only the first out of THREE positive clinical trials in CDD that were presented that afternoon. Dr Julia Tsai from Ovid Therapeutics presented positive data with soticlestat in CDD, where a large majority of the parents reported positive improvements in different disease symptoms and where seizure improvement (the primary endpoint) continued increasing over time. Soticlestat acts through a completely novel mechanism of action that includes indirect glutamate modulation and other potential therapeutic benefits, so these different therapies are good candidates for being used in combination in the future if they get approved. Ovid also run other phase 2 studies with this drug in other genetic syndromes with epilepsy, and has not yet announced how many will be progressing to phase 3 trials or when.

And taking about phase 3 trials, the next presentation was by Alex Aimetti from Marinus who walked us through the positive results of the phase 3 trial with ganaxolone, the very first pivotal trial ever run in CDD, and that recently announced it had met its primary endpoint. The Marinus team explained how until recently the regulators were not even sure if CDD was a separate disorder, and it was thought that there were so few patients in the world that a phase 3 trial could only have 50 patients, but a couple of years later they have the Orphan Drug Designation by both the EMA and the FDA, the Rare Pediatric Designation by FDA, and a 100-patient positive Phase 3 trial in CDD. Ganaxolone led to a significant reduction in major motor seizure frequency and was well tolerated, so the company will apply for marketing authorization and plans to initiate an Expanded Access Program first in the US, that might potentially will get to additional territories later on. Marinus was awarded the 2020 CDKL5 Forum Award for the Company making a difference in the clinical space, and the company has played a unique role at helping de-risk clinical development in CDD and bringing visibility to the disorder.

three-gold-cups.jpg

I want to stress again what we just saw at the Forum: three positive clinical trials in CDD. One pivotal trial, the first ever, which was positive and will hopefully lead to the first drug approved for treating CDD, and two phase 2 studies each of which looks good enough to lead to a new phase 3 study for CDD. I don’t believe there are many rare diseases that get that type of pipeline progression in one year.

6. PREPARING FOR OBSERVATIONAL STUDIES

In the last years we have been talking much about getting the field ready for more complex studies, and by that I mean studies that will look at symptoms beyond seizures. This is particularly critical for the gene therapies. A small molecule can get away with measuring seizures and getting approved for “the treatment of seizures in CDD” even if it has efficacy beyond that symptom, but when treatments are more invasive they need to demonstrate that they have broader benefit.

In preparation for that, we had launched a series of programs including obtaining a new ICD-10-CM diagnostic code for CDD and co-hosting a PFDD meeting with FDA together with IFCR, to help identify the main symptoms that impact patients with CDD. The next step is to translate those symptoms into “clinical outcome measures” which are clinical scales that assess either one main symptom or a collection of symptoms, and that are used in trials to quantify the efficacy of treatments.

IFCR recently announced the award of an NIH U01 grant to a network of CDKL5 Clinical Centers of Excellencefor the characterization over the next 5 years of a collection of clinical outcome measures that they are developing specifically for CDD by running observational studies through their network.

And the Loulou Foundation announced at the CDKL5 Forum the immediate launch of a large international observational study, in the US and Europe, in collaboration with several bio-pharma companies, to determine the feasibility of using in CDD trials a collection of clinical outcome measures that haves already been used successfully in advanced clinical trials for related disorders. The industry partners hope to be able to apply the learnings from this study to their clinical trials in the near future.

This means that the CDD community, both clinicians and patients, will soon start hearing more about the concept of observational studies and the central role that they play in getting the field ready for more complex studies.

7. BIGGER AND GLOBAL

Before concluding, I would like to highlight a couple of additional observations beyond these scientific presentations.

One is how much the idea of how rare CDD is has changed. Not so long ago it was through that we only had a couple of hundreds of patients in the world, and now we know it is one of the most common genetic causes of epilepsy. In their latest corporate presentation, Marinus now estimates about 12,500 children in the US and EU, and the actual numbers could be higher with an estimated incidence of 1 in 42,000 births.

This change in how we see CDD became particularly clear when during the Forum, the speaker from Ultragenyx, a company that was started to treat ultra-rare diseases, spoke about “serving this large patient population” in reference to CDD. And the International CDKL5 Alliance already includes 20 members, also showing the growth of the patient community and their willingness to work closer together.

And the second one is the wonderful keynote speakers that we had in the Forum. On day 1, Antonino Caridi, Ariadna’s grandfather and one of the pillars of the CDD patient community, told us about how life some time takes a turn and places you at war against a monster, and how he found in the CDKL5 Alliance a real alliance beyond just words. He spoke about how we all play a role in the movie about CDD, including the families, and encouraged all “actors” to make the best in the roles they are playing.

On day 2, Yann Le Cam, Founder of EURORDIS, told us about his own journey as a father of a now adult child with a rare disease, and the need to think global and be strategic. In his words: “think about the ecosystem, and not just your organization”. Antonino and Yann both echoed the need to respect the challenge (the monster) ahead, while creating the focus and determination to bring forward a change which requires thinking and acting globally.

And on day 3, Jeremy Levin, CEO of Ovid Therapeutics and Chairman of the Board at the Biotechnology Innovation Organization (BIO), provided us with his unique view, from his global pharma position, of what is happening in biomedical research in 2020 and how the CDD field is doing. He acknowledged the strangeness and difficulties of 2020 for everyone due to COVID-19, and the enormous challenge and disruption that it has brought to the pharma industry. Although we have seen an immediate slowdown of clinical trials, he envisions that many of the changes in trial design and regulatory policies that have been put in place to speed up the development of COVID-19 therapeutic and vaccines will also speed rare disease therapies from now on, including adopting telemedicine. These great changes are not going away, he said, they are here for the future. Jeremy also praised the CDD community for helping Marinus and Ovid deliver late-stage trial results in the middle of the pandemic.


To close, I like to end the Forum updates by reviewing the goals that the Loulou Foundation set when it started, which are to get treatments (to reach trials) by 2020 and cures by 2025. Now we know that by 2020 we’ve had not one trial, but several trials leading to three positive clinical trials including a 100-patient pivotal trial. And based on what we are seeing in the preclinical space, I believe that by 2025, we will have multiple cures (gene therapies and similar) in advanced clinical trials. As Orrin said referring to the progresses during the last year, “what we have seen these 3 days is breathtaking”. We just need five more years like this one.

I hope you enjoyed this summary! let me know your thoughts in the comments. Here are my articles on the 2018 and 2019 Forum meetings.

Ana Mingorance, PhD

Disclaimer: These are my own impressions from the presentations that I was most interested in as a scientist and patient advocate, and not an official text about the Forum by the Loulou Foundation.  I write these texts with the parents of individuals with CDD in mind, so excuse also my lack of technical accuracy in parts ;-)