For the past ten years, the Loulou Foundation has hosted an annual meeting, the CDKL5 Forum, where scientists and drug developers working on CDKL5 Deficiency Disorder (CDD), together with representatives from patient organizations, meet to discuss the latest developments in the field and to advance towards treatments and cures. You can find summaries from some of the last meetings here: 2018, 2019, 2020, 2021, 2022 and 2023.
The 2024 CDKL5 Forum edition took place October 28-29 in Boston, and will stay in Boston next year for the 2025 edition. This was the largest Forum to-date, with over 200 participants including representatives from 33 companies and 18 national patient groups. This was also a special edition, as this was the 10th CDKL5 Forum (with the first one taking place 9 years ago). I will try to summarize the main take-home messages from this year’s Forum. This won’t cover all the presentations, but rather focus on the main themes that we saw and the main progresses – which are always many.
1. New interesting learnings about CDKL5 biology and CDD
Image: Researcher Maria del Carmen Martin Carrascosa (CIPF), Forum Director Dan Lavery (from Loulou Foundation), and research doctor Josephine Thinwa (UTSouthwestern).
Sila Ultanir introduced a session on the biology of CDKL5 saying that a few years ago we only knew that CDKL5 was missing, so we could only think of putting back CDKL5 as a way to cure the disease. But today we know much more about what it does in cells, so this opens up different entry points to try to correct some aspects of the disease without needing to put back CDKL5 yet.
We learnt from Oguz Kanca from Baylor that there are patients with gain of function mutations in CDKL1 and CDKL2 that have neurodevelopmental syndromes so CDKL5 is not the only important gene in this family. And we learnt from Maria del Carmen Martin Carrascosa that fruit flies without CDKL5 have spontaneous epilepsy and repeated stereotypies so they seem to capture better the lack of CDKL5 that mice do. However in any animal that came in evolution before jaws appeared (yes, jaws), CDKL5 is acting almost like a CDKL1-4 which don’t have the full scope of functions of our CDKL5.
One of those potential redundant CDKL genes is CDKL2, so last year we considered if increasing CDKL2 could be a possible therapeutic approach when CDKL5 is missing. This year Kevin Dempster presented what happens if you get a mouse missing both CDKL5 and CDKL2… does it get even worse? The answer is yes but interestingly CDKL2 seems to only do SOME of the jobs that CDKL5 does in the cell, not all of them, notably not the job of changing activity of calcium channels and glutamate receptors. So in a way, increasing CDKL2 has the potential to rescue some of the consequences of missing CDKL5, but other consequences still need to be addressed separately, as is the case of still needing to inhibit Cav2.3 (more on this later).
On the biology side, we learnt from Josephine Thinwa that CDD patients might be more sensitive to viral infections, causing a larger inflammatory response, and from Lauren Orefice that lack of CDKL5 in different sensory neurons makes people with CDD more sensitive to textures, sensations (tactile sensitivity, like a tag in your clothes) and more prone to GI pain. All these seem to be sensory problems caused by CDKL5 lack in sensory neurons.
There were also different presentations during the Forum and pre-Forum meeting that point to the thalamus as an important brain part in CDD, as part of the thalamocortical connection. For example Rachel Oren presented that the cortical Visual Impartment is not in the eye (which is fine) or in the neurons that go from the eye to the thalamus, but somewhere later by the time the signal gets to the visual cortex. Another important brain component seems to be interneurons, shown by Tim Benke using a CDKL5 inhibitor in brain slices, which could help explain why some of the GABAergic drugs including ganaxolone are the best for CDD.
2. Beyond the research toolbox: the year of platforms
a. Cells and mice
One of the main goals of the Loulou Foundation early on was to make sure that any scientist anywhere in the world can work on CDD. This meant creating a collection of research tools like animal models and making them open-access. Can we see where CDKL5 is in this cell? Can we see if it is active? Can we see the consequences of missing CDD? And what does this mean about this organ physiology? we can’t answer this without cells and antibodies and mice.
This has led to a growing collection of animal models and cell models and antibodies which are available to the community. But this was perhaps the year of “platform as a service”, which is the next step once tools exist.
Image: slide from CDD mouse testing platform, slide from CDD iPSC-derived neuron platform, and researchers Liz Buttermore (BCH) and Jasmine Carter (Loulou Foundation) during the breakout session on iPS cells.
I want to highlight the iPSC-derived neurons from CDD patients which are running as a service at Boston Children’s Hospital Neuron Core led by Liz Buttermore, and the breakout session that Liz and Loulou Foundation’s Jasmine Carter led about the iPSC-derived models.
Liz walked us through all the crazy work that led to their very good in vitro epilepsy assay with patient-derived iPSC lines, and it was a great example of how hard is to standardize an assay of this type of have it be robust and reliable. Companies can now send their compounds to Boston to be run in this epilepsy-in-a-dish CDD model using microelectrode arrays.
We had a breakout session in the afternoon led by Liz and Jasmine to see how the academic and biotech community will want to work together in setting up these in vitro models. There were discussions about learning from each other successes and mistakes, and having a shared channel to ask questions since troubleshooting these models can get so tricky. The discussion was fantastic (and I’m sorry I missed the parallel breakout sessions).
Just like with cells, The Jackson Labs has set up a panel of mouse models of neurodevelopmental diseases that are now available for drug testing in a fee-per-service way. Rajat Puri presented data about their CDD mouse preclinical testing service, showing robust signal across mouse ages, mouse batches and even across lab researchers. Again, these standards are very tricky and it is powerful to have them available as a service as opposed to having to go through all the troubleshooting in each individual lab. The tests that I liked the most showed how bad CDD mice are at burying marbles and building nests, which are the mouse equivalent to activities of daily living for people.
And why are the services so important for companies working on treatments? Because having a mouse or a cell line available is like buying ingredients at the supermarket, we still need to cook them in the lab, but these in vitro patient models and complex behavioral mouse models are very elaborate and easy to get wrong, so having them running as a service is like offering restaurants where you can get the complex dish ready. It is much easier to go to a restaurant and order a paella than to spent two years learning how to cook a good one.
b. Patients
There is an observational clinical study in CDD designed to learn about clinical scales to measure symptoms other than seizures in trials, and to collect data that might serve as a control dataset for future clinical trials in CDD. It is called the CANDID study and I am very very proud of it, not only because I was part of the early team that saw the need for such study and got it started, but because it has evolved and it is now owned by the community. At the Forum, Xavier Liogier from the Loulou Foundation presented the results from the scales assessments in the CANDID study, which has more than 100 patients from the US, EU and MENA, and we learnt not only which scales could be used in a trial today for a gene therapy, but also the fact that 83% of patients would qualify for a trial if it asked for a minimum of 16 seizures per month. 83% IS A LOT! In other syndromes 25% or less of patients would have enough seizures for trials, and seizures are very likely to remain as the primary endpoint for trials, so in a way we have the perfect disease for clinical trials. This is important to know!
Image: Xavier Liogier (Loulou Foundation) presents an update on CANDID, and group discussion with doctors and patient representatives.
There was also a meeting for the CANDID study where Xavier and Maria Makarovskaya, the two professionals from the Loulou Foundation in charge of CANDID, asked clinicians in the study and representatives from the patient community about their experience in the study and their feedback. And this is where it became clear to me that CANDID was started by a small group of people, but it is now owned by a much much larger group, including the 112 patients that signed up and their parents, and all of the study investigators and their team of nurses and psychologists. This is huge, and a reason for many to be proud.
There is also a large fluid biomarker effort that was discussed in one of the breakout sessions. A fluid biomarker is a way to measure something in blood or CSF or other body fluid that shows that a therapy is working. For example sugar levels in blood serves as a biomarker for diabetes, while cholesterol in blood serves as a biomarker for cardiovascular risk. Maurizio Giustetto from Italy is looking at saliva to measure little vesicles that neurons spit out to talk to each other and that could potentially be picked in saliva (crazy!). And there is an international collaboration called ELPIS led by Massimiliano Bianchi that is collecting patient blood samples to check for potential biomarkers in plasma. They already found some potential plasma biomarkers, I wrote about that last year, and one year later the program is expanding to more countries and is considering a sample collection of blood and saliva from patients and siblings (as controls) during the Rome 2025 patient conference. Stay tuned for more information about that.
3. Therapies: the now, the near now and the blurry future
At the beginning of the therapies session, Omar Khwaja who is a doctor who has treated kids with neurodevelopmental disease and then became a drug developer, provided an overview of the CDD pipeline today.
We have ganaxolone approved in the US and Europe, and the CDD patients also use often Epidiolex off label, because it is approved for LGS and because there is good Phase 2 data for CDD. So we “almost” have two drugs approved. We then have clinical trials with fenfluramine (in Phase 3) and bexicaserin (about to start Phase 3 for DEEs including CDD). After that, there is a group of four “almost ready for trial” programs including three gene therapies.
In Omar words, we’ve come really far in these 9 years since the first Forum, with an entire landscape of programs in development including disease-modifying genetic therapies in late preclinical stage and multiple drugs in trials or approved. I invite you all to see the older summaries of the Forum to see this progress, including former trials that didn’t progress like ataluren (negative data in Phase 2) and soticlestat (positive data in Phase 2 but not chosen by the pharma for a Phase 3).
a. The now
Images: Sabine Bongardt from UCB Pharma, Randall Kaye from Longboard, and a slide from Elisa Borghi’s presentation on the pre-biotic and pro-biotic supplement trial.
UCB Pharma presented the progress with the fenfluramine Phase 3 trial in CDD, which is probably closing recruitment at the end of this year. They are particularly asking for children ages 1 to 2 to participate, with room for about 20 in the trial. This is very exceptional, because trials often only start at 2 years of age and in a disease like CDD that can start as early as in the first weeks of life, those two years can mean over one thousand seizures before the child could start taking fenfluramine (or any other drug with a minimum age of 2 years!). So if you know any family with a one year old with CDD and uncontrolled seizures please let them know about this trial (more info about locations in the trial website) so that this trial will complete recruitment at the end of the year.
And as one trial ends another one starts. Longboard Pharma is developing a drug called bexicaserin that is a second-generation fenfluramine, and it is starting a Phase 3 trial combining all epilepsy syndromes and in particular the large ones (including CDD) for patients ages 2-65 with at least 4 countable seizures a month. They are aiming for 80 trial hospitals, potentially over 100, and to start recruiting really soon. So check with your doctors if Longboard has reached out to them about this upcoming trial with bexicaserin called DEEp OCEAN.
By the way, Longboard has been quite popular recently because of choosing to combine all the epilepsy syndromes (all the DEEs) into a single trial and they even recently got bought by Lundbeck, the pharma company that makes clobazam. So there has been a lot of momentum around bexicasein. And Longboard received the 2024 CDKL5 Forum Company Making a Difference Award and CMO Randall Kaye talked about seeing an amber light as a sign to speed up, not to stop. Longboard is speeding up to treatments for all DEEs, even though the difficulty of the task ahead could have made others want to slow down (or turn away). For us in CDD, their Phase 3 means another chance to access a novel molecule in development before it gets approved.
And at the Forum we also had a very interesting presentation from Elisa Borghi about a clinical trial with a mix of pre-biotic and post-biotic supplements, so “food for your microbiome” and “products that good microbiome would release”. Their research was started by the Italian patient family group, and it is now in a Phase 2 trial with CDD patients to see how it might improve GI function and epilepsy, which seems to have a GI component. What I like of this program is that if the trial is positive, families can directly access the supplement and don’t need to wait for lengthy regulatory approval and price reimbursement negotiations as is the case for drugs. And there was another company in attendance although not presenting that is also targeting the microbiome, but using the good bacteria (so it is not a pre- or post-biotic but a biological treatment!). I hope that next year we also have them present and learn more about clinical interventions in this gut-brain area.
b. The near now
I don’t know if you have realized that we have now three gene therapies and one small molecule program all past clinical candidate nomination. “Clinical candidate” is how we call a drug or therapy that is already done with optimization and tweaking, and that scientists then transition to safety and toxicology testing in animals other than mice so prepare to file and IND (a permission to start trials).
Images top: Ultragenyx’s Emil Kakkis and slide from their gene therapy program. Slide from Lario Therapeutics.
Images bottom: UCDavis’ gene therapy slide and Kyle Fink. Slide from the Loulou Foundation and Penn gene therapy presentation.
One of these is the gene therapy for Ultragenyx, which CEO Emil Kakkis explains that already reaches enough number of neurons in pigs (and is safe) to look ready for trials, and in a few months they hope to make a decision about potential methods to improve the brain distribution of the virus prior to then progressing to an IND. He also showed us great results that Ultragenyx is having in other three neurological diseases like Angelman syndrome where Ultragenyx is progressing to a Phase 3, and left us with the hopeful learning that “the brain is more plastic and more capable than we think”.
A second program in this near now category is the gene therapy from the Loulou Foundation, funded via its subsidiary Elaaj and created by the Gene Therapy Program at UPenn under the direction of gene therapy super-expert Jim Wilson. Jim recently left Penn to start a company called Gemma Therapeutics so the presentation of this gene therapy was done by Janine Lamonica, the leading scientist for this program at Gemma. This gene therapy looks like a solid clinical candidate that is ready to take forward. Both the Ultragenyx and the Loulou Foundation gene therapies use a copy of human CDKL5 gene inside of an adeno-associated virus (AAV) and are administered via injection into the cisterna magna, the big pool of CSF that you can access at the nape, in between your neck and your skull (via ICM). So these gene therapies are fairly similar.
A third gene therapy that is now moving into toxicology is the X-reactivation CRISPR 2.0 approach from Kyle Fink at UCDavis that uses a modified CRISPR to find the second CDKL5 gene in cells, which is inactivated as part of the inactive X chromosome, and makes the cell read it. Instead of find-and-cut this is a find-and-activate type of CRISPR, and to get it to neurons the scientists are using two AAV virus, each carrying half of the CRISPR sequence (actually millions of virus of each, not only one of each, so that when they infect the same neuron they can make the find-and-activate type of CRISPR). Their mouse rescue data is beautiful and they are seeing success in adult female mice, again showing that it is probably never too late for CDD!
And the small molecule program that has already nominated a candidate (a final molecule) is the Cav2.3 inhibitor program from Lario Therapeutics. Two years ago we learnt that Cav2.3 is a channel that makes glutamate neurons fire and that is usually controlled by CDKL5 as a break, bringing down its activity. In the absence of CDKL5, Cav2.3 is too active and excitatory neurons go crazy, so we would need inhibitors. Last year we saw some early data with those inhibitors made by Lario, and this year they showed us a beautiful dose-response activity in glutamatergic neurons made from CDD patient cell lines, so the more drug the more correction of crazy excitation. This is a drug, not a gene therapy, which makes it very attractive to combine with other treatments.
c. The blurry future
Here is where we have all the other therapies including gene editing (the one that fixes your DNA sequence, by David Liu), or the gene therapy that fixes non-sense mutations (by Tevard), or the early-stage ASO approaches that are being attempted to rescue CDKL5 or increase CDKL2, and even the efforts to make CDKL5 protein in the lab to then give it to the brain (enzyme replacement therapy). There are also talks about changing the virus type to the ones that can be administered via blood injection and then cross to the brain, so that we don’t need to inject gene therapies directly into the brain. Many people are working hard in all those avenues, but we don’t know IF any of them will eventually get to trials or WHEN they might get to trials. So what we know is that today, the best gene therapy type that medicine knows how to do is to put a gene copy inside of an AAV9 virus or similar and to administer this gene therapy directly into the brain, like the three CDD gene therapies in the “near now” category. Anything else, is still a blurry (but hopeful) future.
4. What’s good enough and the big head problem
One thing that became clear in the discussions with the speakers and the industry panel is that because AAV gene therapies can only be done once, we really need to be sure that we got the right dose and distribution before trials. In other diseases where the main “gene therapy” type is an Antisense Oligonucleotide (ASO), which is like a piece of DNA that you inject into the spine once every few months, you can go to trials earlier and figure out the right dose there, but you cannot do this in viral-delivered therapies which are once-and-done. I will give you two real examples for this go-to-trials speed advantage with ASOs: (1) in Angelman, Ultragenyx went into the clinic initially with a dose of ASO too high, they saw some localized inflammation in the spine and scary leg paralysis but that was transient, and they could continue dosing those same patients with lower doses. (2) In Dravet syndrome, Stoke Therapeutics started trials thinking that the good dose would be 30mg, it wasn’t, so they went up to 45mg to try to see efficacy and still wasn’t enough, so then they started using 70mg as the dose and finally got impressive efficacy. In these two cases, the first dose was “wrong”, but it wasn’t a big problem because you keep dosing ASOs every few months so you can always go up or go down. But in CDD we haven’t had good shots on goal until recently to use ASOs, and we are instead using viral-delivered gene therapies as our first gene therapies, and because in those you literally only get one shot we are taking much longer to start trials because they have to be right from the start.
So how do we know when it is good enough? In the case of CDD the main challenge seems to be being confident that we are reaching enough neurons, because mice have small brains that are easy to cover with the virus but we have much much larger brains (a 5 year old has a brain 2,500-times larger than a mouse brain). So that is our challenge, to have so much virus spread throughout the brain in pigs and monkeys (mid-size brain) that we can trust we can jump to humans and get enough brain infection to see good efficacy.
There was quite a bit of debate about “what is good enough” to take those gene therapies to trials, which the consensus answer being that we need to be “comfortably sure” before starting trials. It still looks like it will come in the near future, but this is why it is so hard to pin-point exactly when. And the feedback from the industry experts in the panel was that even though we know the future will bring better gene therapies than the current most-advanced ones, this should not prevents us from starting trials with the almost-ready ones as soon as they are ready. Because patients don’t have so much time to wait in particular if they are young, and because we don’t know what future technologies will bring us and when.
As I write this, I am also thinking that we need to work hard on getting non-viral therapies to move to trials because they can be combined with the viral gene therapies, while scientists have not yet found a way to be able to give more than one viral gene therapy to the same person (the “redosing” problem, see also the next section). There are projects that were not presented at the Forum because they are early stage but they fit this need very well, such as the development of ASOs to increase CDKL2, ASOs to fix CDKL5, and making the CDKL5 enzyme in the lab to then infuse it into the brain (enzyme replacement therapy). I hope we can review these next year.
5. How the patient community can help
Industry panel, from left to right: Ralf Schmid (Novartis), Sharyl Fyffe-Maricich (Ultragenyx), Randall Kaye (Longboard), Brian Moseley (UCB) and moderator Majid Jafar (Loulou Foundation)
We had a panel with industry experts in drug and gene therapy development to see how we, the community, can help get new treatments forward faster. Some messages were:
EDUCATE/RESEARCH: Help families understand that being part of research should be part of the treatment paradigm. It doesn’t need to be get into clinical trials with drugs, but perhaps observational clinical studies like CANDID, or biomarker studies like ELPIS. If patients don’t participate then we can’t get that next generation of treatments.
RESEARCH: Help validate scales and outcome measures that can go beyond seizures, so that the new therapies can show that they are better than the current seizure therapies (I think we are already doing this well with the U01 and the CANDID studies and with ELPIS)
EDUCATE families about gene therapies: what can we realistically expect, what does the journey of a trial for a gene therapy looks like…
ADVOCACY: Some challenges that we can’t fix but can perhaps influence are getting regulators to understand what we need for gene therapy clinical trials, and also the problem of pricing for gene therapies.
RESEARCH: also beyond our reach but where we can join researchers and other communities, is the problem of figuring out a way to make gene therapies redosable. If we know that we have a second-shot, and it is not only a one-shot goal, it will make it easier for people to participate in trials and for companies to run those trials. One way that we could help this is for example to have a focused workshop or session at the next Forum about the immune side of gene therapy and how to be able to one day re-dose patients.
I will add two things that they didn’t say but that I think are important roles that the community also plays towards getting new treatments:
RESEARCH/ COMMUNITY: I noticed that several of the presentations were on research projects that had initially been seeded by some of the national patient groups, like Canada, Italy and Spain. So this goes to show that seeding research is another important role of the patient community that can start the building of a mountain with the first grain of sand. I think this is a very important function of patient groups also, to seed CDD research in their countries.
COMMUNITY: We really need to find more patients. It is clear that there are more patients living with CDD that don’t know their diagnoses yet, than those who already have the correct diagnosis. An undiagnosed patient is a person who might be in the wrong treatment, who will miss all these trials, and who can’t help us solve the science questions (including the accurate epidemiology of the disease). I believe this needs to be addressed at a country level so I am adding it to our to-do lists!
6. The patient community: belonging, believing, and getting work done
The Forum traditionally opens and closes with the voice of the patient, and this year opened with Lily Howard telling us her experience as a sister to Harper, a little girl with CDD. Lily spoke about isolation and forced maturity, and feeling different from other kids and not able to connect. That’s why she has created an event for siblings, so that they get to create a sense of belonging.
Image: Majid and Lynn Jafar (Loulou Foundation, parents to Loulou); Lily Howard (Hope-4-Harper, sister to Harper); Andreas and Ana Born (CURE5; parents to Siena); and Heike Knip (CDKL5 Alliance President, dad to Valentina)
During the big gala dinner night, the patient award went to Ana Carolina and Andreas Borg, the parents of little Siena, who talked about the initial shock of the diagnosis, discovering a world that they didn’t know existed, to then finding purpose and a sense of agency in advocacy. I will share one beautiful sentence from Andreas: “all solvable problems can be solved if enough talented people put their energy into them”, and one from Ana: “we always look for heroes, and I’m proud I’m raising mine”. That’s the magic night of the Forum where we all end up crying and at the same time believing in a wonderful future that we can build.
To close the Forum, the outgoing Chair of the CDKL5 Alliance, Heike, spoke once again about the need for Unity in the patient community and shared the highlights from the last year work by each patient group in the Alliance. He also reminded us that we should try to improve quality of life, not just focus on gene therapies, with the example for how his daughter Valentina has gone from having 3-4 seizures a day to having 3-4 seizures a week in one clinical trial, so every bit that we can do helps a lot. This call echoes the words from Lily at the opening of the Forum, when she told us that even assistive technologies (not even medicines) can improve quality of lives for patients and their families. It is a beautiful reminder that at the end of the day, we are treating a person, not just a gene.
Last, Majid Jafar, co-Founder of the Loulou Foundation, wrapped up the Forum asking us all to meet at the CDKL5 Alliance meeting in Rome in June 2025, and announcing that the next CDKL5 Forum will return to Boston in October 27-28 of 2025. And he left us with a call to action: “we’ve come so far… and there is still so much to do. Let’s get it done”.
So we get back to work with our heads full of ideas and our hearts full of friends. I look forward to seeing you all in Rome and in Boston next year.
Ana Mingorance, PhD
Disclaimer: This is my own summary and key learnings, and not an official text about the Forum by the Loulou Foundation. I write these texts with the parents of people with CDD in mind, so excuse also my lack of technical accuracy in parts.