Introduction
Pathogenic variants in SCN2A and SCN8A have been linked to a wide variety of neurodevelopmental disorders (NDDs), including infantile seizures, epilepsy, autism spectrum disorder (ASD), intellectual disability (ID), and developmental and epileptic encephalopathy (DEE) [Reference Berg, Thompson and Myers1,Reference Meisler, Hill and Yu2]. SCN2A and SCN8A encode for voltage-gated sodium channels Nav1.2 and Nav1.6, respectively. These channels mediate sodium ion flux in neurons, and pathogenic variants that disrupt channel activity impair neuronal excitability and overall brain function. Hundreds of unique pathogenic variants with unique effects on channel function have been identified, contributing to a wide spectrum of clinical presentations [Reference Brünger, Pérez-Palma and Montanucci3]. Pathogenic variants are largely de novo, spontaneous, new mutations occurring in a child on one of the two SCN2A or SCN8A alleles. Variants that eliminate channel protein or impair channel function are termed “loss-of-function” (LoF), whereas variants that result in an over-active channel are termed “gain-of-function” (GoF). In some cases, variant effects on channel function are more complex, with a constellation of properties that on their own would be classified as either loss- or gain-associated. These variants are currently termed “mixed function.” For both SCN2A and SCN8A, there is a general genotype-phenotype correlation: LoF variants often present with ASD and/or ID with or without seizures, whereas mixed and GoF variants are more often associated with DEE [Reference Berg, Thompson and Myers1,Reference Conecker, Xia and Hecker4]. However, this categorization is not definitive and GoF, LoF, and mixed variants present with clinically overlapping features.
Thus, SCN2A and SCN8A-related disorders (SRDs) are medically complex, and disease management is often incomplete and misguided, contributing to a lack of answers for families caring for individuals with SRDs. Advocacy for improving disease management, clinical outcomes, and advancing therapeutic development has been an ongoing effort since the establishment of the FamilieSCN2A Foundation and the International SCN8A Alliance.
The FamilieSCN2A Foundation was established in 2015 by caregivers whose children are affected by an SCN2A-related disorder [Reference Schust, Burke, SanInocencio, Bryan, Ho and Egan5]. With a focus on building a connected and purpose-driven ecosystem, the FamilieSCN2A Foundation serves as a central hub for patients, caregivers, scientists, clinicians, and industry. Guided by core values of urgency, integrity, collaboration, and inclusion, the Foundation is driven by a shared vision of a world with effective treatments and cures for all SCN2A-related disorders [Reference Schust, Burke, SanInocencio, Bryan, Ho and Egan5,Reference Sanders, Campbell and Cottrell6]. To foster multidisciplinary collaboration, the foundation began to host a Family & Professional Conference every summer, the first taking place in Chicago, IL, in 2016, with the goal of connection at the forefront.
Similarly, the International SCN8A Alliance was founded by caregivers in 2014 to bring together a broad spectrum of stakeholders and collaborate on advancing the science of SCN8A-related disorders [Reference Warren, Patel and Helen Cross7]. The Alliance aims to accelerate scientific progress by hosting scientific meetings, funding early investigators, building collaborations between researchers and clinicians, and engaging the global SCN8A community, with notable contributions including the first consensus guidelines for SCN8A and the first SCN8A Research Roadmap [Reference Conecker, Xia and Hecker4,Reference Berg, Burge and Conecker8].
In conjunction with the 2025 FamilieSCN2A Foundation Annual Family and Professional Conference, these advocacy groups led a Professional SCN2A & SCN8A Research Roundtable with goals to advance therapeutic development for SRDs. Researchers, clinicians, industry professionals, and caregivers convened for a drug development workshop, with the ultimate objective of establishing a therapeutic development framework spanning from preclinical research to clinical trial readiness. The therapeutic modalities explored were allele-specific antisense oligonucleotides (ASOs), gene editing, and small molecules. The goal was to collaboratively define scientific and strategic milestones for distinct therapeutic modalities, with an emphasis on caregiver-informed priorities.
Materials and methods
Conducting this roundtable jointly across the FamilieSCN2A Foundation and the International SCN8A Alliance had multiple benefits. First, the SCN2A and SCN8A genes both encode for sodium channels that play important roles in cellular excitability in the brain and are often co-expressed in the same neurons;[Reference Lorincz and Nusser9,Reference Lorincz and Nusser10] thus, there are common disease manifestations and potential treatments across both groups. Much of the translational work at the heart of this roundtable, including ASO development, gene editing approaches, regulatory pathways for N-of-1 trials, and outcome measure development, is driven more by therapeutic modality than by the specific gene involved. Second, because each condition is individually ultra-rare, pooling stakeholder bases and expertise to support a single convening rather than two separate ones yielded more breakout groups, broader industry pipeline representation, and richer cross-disease comparison than either community could have achieved alone.
At the same time, the cross-disease design could result in some tradeoffs for each group. SCN2A and SCN8A differ in their variant mechanisms, with SCN8A skewed toward GoF and SCN2A spanning a gain and loss of function spectrum, along with differences in phenotype severity and existing therapeutic pipelines. There is also a risk that caregivers or scientists more invested in one condition may feel their priorities were diluted. Finally, building shared preparatory materials, such as the roundtable’s information packet, for two distinct but related communities added logistical complexity, requiring content specific enough to be useful for each condition without becoming too broad in scope.
Ten breakout groups were formed to reflect the full spectrum of stakeholder perspectives, ensuring that caregivers, clinicians, researchers, industry partners, and foundation staff were represented in each discussion. Total attendance distribution was as follows: 22% industry, 16% caregivers, 16% clinicians, 41% researchers, and 6% foundation staff, with 102 total attendees (percentage values rounded). Groups of 9–11 individuals were populated according to attendance distribution & known research/therapeutic interest, with the “average” group comprising 21% industry, 15% caregivers, 16% clinicians, 40% researchers, and 7% foundation staff (Figure 1).
Breakdown of participant affiliation by therapeutic modality focus. Each row represents one breakout group (numbered 1–10) with 9–11 participants. Boxes on right show the expected therapeutic modality focus based on the current pipeline of the industry representative and which therapeutic modality was actually discussed by the group. Two groups that we expected to discuss small molecules and two groups that we expected to discuss gene editing strategies instead ended up discussing ASOs. In addition, one group expected to discuss gene editing strategies instead discussed small molecules, and vice versa. These changes in therapeutic modality focus were driven by caregiver input at the beginning of the sessions.
Figure 1. Long description
Heatmap showing distribution of five stakeholder groups across three therapeutic modalities (Antisense Oligonucleotides, Small Molecules, Gene Editing) and 11 participants. Stakeholder groups by color: Purple=Industry, Yellow=Caregivers, Blue=Clinicians, Green=Researchers, Pink=Staff. Left bars indicate intended modality focus; right bars show actual focus. Participants 1-2 predominantly Industry/Caregivers; Participants 3-9 show increased Researcher/Clinician representation; Participants 10-11 show Staff and Caregiver engagement across approaches.
Each drug developer attending the conference was designated to 1–2 groups based on their current pipeline. The intended therapeutic modality categories were broadly defined intention as ASO’s, Gene Editing, and Small Molecules. Based on the current drug developer pipelines, we intended to have 2 groups discuss ASO’s, 4 groups discuss gene editing, and 4 groups discuss small molecules (Figure 1, right).
Regardless of intent in group organization, each group chose their own therapeutic modality for the workshop, which was informed by a caregiver perspective at the beginning of the session. Somewhat surprisingly, most groups chose to discuss ASO’s (6/10 groups), and 2 groups each chose to discuss gene editing and small molecules (Figure 1, right). Groups were provided with a large-format blank research roadmap template with fillable domains representing key translational phases and a meaningful impact/outcome prompt (see Figure 2). These roadmap domains were: (1) Preclinical Discovery, (2) In Vitro Studies, (3) In Vivo Studies, (4) Preclinical Development, (5) Clinical Design I, and (6) Clinical Design II.
Roadmap template. This template with fillable domains representing key translational phases and a meaningful impact / outcome prompt was provided to each breakout group to facilitate their discussion.
Figure 2. Long description
This diagram shows seven sequential drug development stages connected by a central pathway: (1) Pre-Clinical Discovery—therapeutic class, delivery, natural history; (2) In Vitro Studies—cell lines and assay systems; (3) In Vivo Studies—animal models; (4) Pre-Clinical Development; (5) Clinical Design I—trial design, eligibility, GMP manufacturing, family input; (6) Clinical Design II—pivotal and secondary endpoints; (7) Parking Lot/Considerations—regulatory and stakeholder concerns. Information gaps are marked at each stage. Color-coding: green (pre-clinical), tan (biomarkers), purple (animal studies), blue (clinical), orange (regulatory/stakeholder).
Each roadmap also included a “Parking Lot” section for capturing cross-cutting regulatory issues, family-centered insights, and unresolved questions. Blank boxes under each domain were labeled with subcategories based on prior stakeholder input and best practices in translational research planning, such as Therapeutic Class, Delivery, Cell Line(s), Assay Systems, Animal Models, Good Manufacturing Practice (GMP) Manufacturing, Trial Design, and Endpoints.
To ensure a shared foundation of knowledge across participants with diverse expertise, each table was also provided with an approximately 40-page information packet containing curated SCN2A/SCN8A background materials. This packet included summaries of available cell lines and in vitro models, descriptions of existing animal models, lists of clinical studies, highlights and summaries of key academic publications, data-sharing resources, and emerging outcome measures. The packet was designed to ground discussions, minimize time spent on basic clarification, and support more efficient roadmap development, particularly for attendees whose primary expertise did not span the full breadth of translational research or therapeutic modalities represented in the workshop.
At the beginning of the breakout session, caregiver participants provided narrative input on lived experience, symptom prioritization, and treatment burden that was shared with the group to inform downstream decision making. Teams collaboratively filled in the roadmap templates using markers and sticky notes, working to align therapeutic design elements with feasibility, readiness, and patient-centered outcomes. Each breakout group presented a short summary of their discussion to the entire room.
All completed roadmaps were photographed and transcribed for qualitative analysis. Outputs were synthesized using thematic coding to identify recurring trends, translational gaps, and actionable insights across therapeutic modalities. These were cross-referenced with transcript excerpts from live group discussions and plenary debriefs to enrich interpretation.
Results
Thematic analysis of the roadmaps and group discussions revealed several cross-cutting themes spanning therapeutic modality selection, outcome prioritization, and translational infrastructure. Participant quotes, drawn directly from the workshop’s audio recording and woven throughout this manuscript, further illustrate how these themes emerged in real time. Across stakeholder groups, participants emphasized the prominence of allele-specific ASOs and emerging gene-editing platforms, the need to move beyond seizure-centric endpoints toward caregiver-prioritized domains, and the central role of variant stratification and human-relevant models in de-risking development. These themes were consistently shaped by caregiver perspectives on treatment burden, acceptable risk, and what constitutes meaningful clinical benefit.
Therapeutic modality trends
Discussions around therapeutic strategies for SRDs centered on the rapidly evolving therapeutic landscape [Reference Robinson, Wettschurack and Halurkar11], with particular attention to ASOs, gene editing platforms, and small-molecule approaches. Across these modalities, participants weighed the scientific promise against practical considerations including safety, feasibility, allele specificity, delivery challenges, and the lived experiences of families navigating these disorders.
ASOs
Across all breakout groups, ASOs emerged as the most actively explored therapeutic modality (6 out of 10 groups). ASOs are short nucleic sequences that bind directly to mRNA sequences to modify protein expression. In theory, ASOs can be designed to up- or down-regulate protein expression, but challenges remain in developing ASOs that can selectively target alleles harboring missense variants. Down-regulating ASOs (RNAseH or gapmer ASOs) can induce mRNA degradation and reduction of protein expression by targeting the wild-type gene and a specific variant. Up-regulating ASOs, including splice-switching ASOs, can be used to enhanced protein levels by targeting repressor exons in cases of LoF and haploinsufficiency or by modulating exon inclusion to alter protein function. Currently, ASOs are delivered to the brain via monthly intrathecal injections to achieve long-term efficacy. Clinical development is already underway: a non-allele specific, down-regulating SCN2A ASO is being evaluated in an ongoing clinical trial of early onset SCN2A-related DEE. SCN8A programs remain in preclinical stages [Reference Hill, Jafar-Nejad, Rigo and Meisler12,Reference Dame, Kopalle and Miñana13]. N = 1 personalized ASO trials for two SCN2A cases are also underway, and early results show improvements in seizure burden and neurodevelopmental skills [Reference Kim-McManus14,Reference Kim-McManus, Mignon and Douville15].
Participants frequently cited the growing accessibility of ASO design, the expanding regulatory precedent, and the alignment of ASO strategies with current disease biology as reasons for choosing this modality over others. Participants emphasized allele specificity as a critical design consideration for down-regulating ASOs, as suppression of both the healthy and missense allele could mirror LoF conditions or, at high doses, reduce channel numbers to levels not observed in the patient population. Such conditions have been modeled in mice, resulting in more severe phenotypes [Reference Eaton, Zhang and Ma16–Reference Ma, Eaton and Liu18].
Delivery considerations shaped much of this discussion. Standard ASOs require repeated intrathecal injections of the “naked” nucleotide, and clinicians and caregivers voiced concerns about the cumulative burden of lumbar punctures, particularly because many individuals with SRDs have an elevated level of anesthesia-related respiratory risks. These practical, lived considerations shaped enthusiasm for next-generation delivery technologies such as vectorized ASOs and oligonucleotide transport vehicles (OTVs), which allow intravenous administration. In a vectorized ASO, the naked nucleotide is placed in a plasmid that is then reproduced by the body for life, eliminating the need for repeated intrathecal injections. There was excitement at the possibility of using vectorized ASOs as a potential bridge to gene therapy, highlighting both the modular potential of ASO platforms and the desire to incrementally scale up therapeutic intervention. Similarly, an OTV could be designed to deliver the naked nucleotide across the blood brain barrier. This vector could be delivered intravenously, allowing for continual dosing and titration to optimal effect. While vectorized ASOs and OTVs may ease delivery challenges, they still face significant hurdles including remaining questions regarding effectiveness in the CNS, immunogenicity risks, balancing effective dosing with unintended side effects, and the prospect of more extended regulatory approval timelines.
One group described a stepwise strategy: starting with intrathecal dosing to establish safety and efficacy, followed (if warranted) by a transition to a vectorized ASO that only needs to be given once.
As one industry professional theorized,
We would administer it to patients as an ASO IT a few times and look at safety and efficacy, and then, if everything looks good … we will start using the vectorized ASO as a one-and-done.
Additional ASO directions generated interest, including splice-switching approaches for exons that are differentially expressed across development. For SCN2A, this would be appropriate for variants contained within exon 5, which switches from a neonatal form to an adult form during development. Such modulation could reduce expression of pathogenic isoforms and represents a potentially powerful mechanism for SRDs[Reference Dame, Kopalle and Miñana13] in exon 5 or with biophysical effects that are modulated by exon 5 splice isoform type [Reference Thompson, Ben-Shalom, Bender and George19].
Base and prime gene editing
Gene editing was also discussed robustly, albeit with greater concern for safety, dosing, and long-term expression (2 of 10 groups). Base and prime editing were repeatedly highlighted as promising emerging technologies that could serve as an alternative to traditional CRISPR/Cas9 systems. Unlike CRISPR approaches that introduce double-stranded breaks, base and prime editors nick only one DNA strand, reducing the risk of unintended insertions and deletions [Reference Anzalone, Randolph and Davis20,Reference Komor, Kim, Packer, Zuris and Liu21]. Because most SCN2A and SCN8A variants arise de novo, these technologies may require N = 1 or ultra-small trials tailored to individual pathogenic variants. Base editing currently allows correction of 4 out of 12 possible base substitutions and is limited by risks of bystander edits within the editing window [Reference Wang, Zhang, Zhang and Zhao22]. Still, preclinical work on an adenine base editor targeting a recurrent SCN8A variant showed improvements in seizures, SUDEP, and non-seizure comorbidities, with minimal off-target effects in a mouse model [Reference Reever, Boscia and Deutsch23]. Recently, base editing was used to successfully treat a patient with severe liver-related CPS1-deficiency [Reference Musunuru, Grandinette and Wang24], though successful application for a brain-related disorder has yet to be reported. This approach may represent a framework for rapid deployment of a gene editor, though more work is needed to translate this approach to brain disorders [Reference Yang25].
Prime editing offers greater versatility, enabling correction of virtually any single-nucleotide variant as well as small insertions and deletions. Participants noted, however, that gains in efficiency often come with higher rates of unintended edits, requiring careful optimization [Reference Chen and Liu26]. A persistent translational barrier across gene-editing approaches is size: CRISPR/Cas9-type editors exceed the packaging capacity of standard AAV vectors, necessitating minimized editors or dual-AAV systems for human delivery [Reference Sousa, Hemez and Lei27,Reference Mittas, Riedmayr and Gavrilov28]. In addition, broader regulatory issues surrounding criteria for initiating IND trials for brain disorders have yet to be clarified, though there are ongoing discussions regarding the possibility of platform-based approval of editing systems [Reference Devinsky, Coller and Ahrens-Nicklas29].
Two other therapeutic modalities, CRISPR activation and full gene replacement, have also seen recent preclinical advances, but, due to time, were not a major part of the roundtable discussion. CRISPR activation is an approach where a “nuclease-deficient” form of CRISPR Cas9 that is incapable of editing DNA is used to shepherd a transcriptional activator to a gene’s upstream regulatory elements, including promoters or enhancers. This approach can thus increase expression of the target gene, which would be useful for SRDs where one allele harbors a protein-truncating variant. Recent work has highlighted efficacy at the cellular and behavioral level mice for a promoter-targeting CRISPRa approach [Reference Tamura, Nelson and Spratt30,Reference Wang, Derderian and Hamada31], and at the cellular level in hiPSCs for both promoter- and enhancer-targeting approaches [Reference Tamura, Nelson and Spratt30,Reference Chen, Nair and Osorio32].
Gene replacement, where an exogenous copy of a gene is introduced to the system, has recently been advanced for a related sodium gene, SCN1A [Reference Mich, Ryu and Wei33]. This gene, like SCN2A and SCN8A, is too large to be packaged into current AAV delivery vectors. Thus, this group split the gene into two viruses, with the split site flanked by machinery that allows for reconstitution of the full-length gene in cells. This approach was viable when targeted exclusively to inhibitory interneuron populations, minimizing seizure burden in an SCN1A Dravet mouse model, but was lethal when delivered without interneuron specificity [Reference Mich, Ryu and Wei33].
Small molecules
Small molecule strategies were comparatively less discussed but were recognized for their potential advantages in systemic delivery and regulatory familiarity (2 of 10 groups). Several participants emphasized drug repurposing, including fenfluramine—a serotonin 5HT2 agonist approved for Dravet syndrome—which has shown meaningful seizure reductions in a case series of three individuals with SCN8A-related DEE [Reference Aledo-Serrano, Cabal-Paz and Gardella34]. Meanwhile, bexicaserin (LP352), a potent and selective 5HT2C agonist, has completed an open-label Phase 2 trial for DEE [35] and is now in Phase 3 testing [36].
Notably, discussions broadened beyond seizure control to include autonomic dysfunction, a major but historically overlooked challenge across SRDs and other DEEs. Participants, especially caregivers, repeatedly underscored gastrointestinal (GI) dysfunction as a priority non-seizure outcome with profound implications for daily life [Reference Hosie, Ellis and Swaminathan37–Reference Hosie, Abo-Shaban and Lee39]. These experiences highlighted the urgent need for targeted small-molecule or biologic strategies addressing GI and autonomic pathways. Other small-molecule avenues, including neuroinflammatory modulation [Reference Hammer, Bahramnejad, Watkins and Ronaldson40], GABAergic approaches, and transcriptomics-guided drug discovery, were identified as promising but early-stage directions.
Caregiver-informed priorities
Caregiver insights shaped roadmap development in direct and tangible ways. Participants highlighted the appeal of one-time therapies (such as gene therapies or long-acting vectors) over chronic, invasive options like repeated intrathecal ASO delivery. As one SCN2A caregiver explained:
[ If ] they offered me the option of an ASO, where you can get the lumbar puncture treatment every two months or so … not just now but as he continues to grow through his childhood and then possibly into adulthood. That’s just a lot to put on him.
Convenience and reduced procedural burden were emphasized as key considerations for families navigating complex care routines. However, despite the treatment burden of multi-use gene therapies, when considering one-time gene therapies like gene editing, many caregivers were concerned with the novelty of these treatments and their safety, along with accurate dosage, reliability, and immunogenic responses that may preclude one’s ability to receive a different gene therapy in the future, if needed.
Caregivers also emphasized the cyclic and interrelated nature of symptoms, such as vomiting leading to aspiration, sleep disruption, and behavioral regression, and urged stakeholders to develop therapies that address the full arc of disease burden. One SCN2A caregiver described this as:
One thing that came up was there’s stably bad, and then there’s unpredictably bad. And those are very different things.
Finally, participants noted that caregiver risk tolerance varied by age, disease severity, and prior treatment experience, underscoring the importance of shared decision-making frameworks in future trial design.
Cross-cutting gaps
Several recurring challenges were identified across therapeutic modalities. One of these challenges is the lack of validated outcome measures for non-seizure domains. While seizure reduction was acknowledged as a pragmatic and measurable primary endpoint, stakeholders repeatedly emphasized the importance of targeting a broader set of outcomes reflective of lived experience. GI issues, behavioral dysregulation, cognitive function, and developmental trajectory were frequently brought up as critical areas of concern. These domains, often overlooked as endpoints in clinical trials, were highlighted by caregivers and clinicians alike as central to quality of life. As such, these areas are important to measure as part of a clinical trial to capture meaningful change. Caregiver narratives reframed the definition of treatment success beyond seizure frequency to include reduced hospitalizations, improved communication, fewer feeding complications, and restoration of developmental challenges. Both caregivers and clinicians acknowledged that a disease-modifying therapy trial carries higher risk than a traditional epilepsy medication trial, as greater improvements in all domains are expected from a disease-modifying therapy trial.
Participants across all groups also emphasized the necessity of variant-informed basic research and clinical care. The lack of robust stratification tools, including availability of functional testing, was identified as a bottleneck in early development planning, with downstream implications for trial design and patient recruitment. Precise genotype-phenotype relationships have been difficult to elucidate in SRDs because of GoF, LoF, or mixed-effect variants that cause complex alterations to channel function. High rates of mosaicism (∼6% of patients) also contribute to phenotypic variability among patients [Reference DiStefano, Cooper and Elisha41]. Functional characterization of variants was viewed as essential to guiding therapeutic eligibility and mechanism matching. In SCN2A-RDs, recent efforts to develop computational tools to segregate protein-truncating variants, which tend to be associated with ASD and behavioral abnormalities, and missense variants, associated with neonatal onset, epileptic spasms, and seizures, may improve the ability of clinicians to quickly and confidently treat patients [Reference Crawford, Xian and Helbig42]. Several teams recommended incorporating long-read sequencing and single nucleotide polymorphism (SNP) phasing to support allele-specific approaches, particularly for ASO strategies targeting pathogenic variants on a single allele. One investigator summarized the shared view across groups:
We realized there is a very clear point of interaction between all of the groups here—we need targeted long-read sequencing on all gain- or mixed-function variants in all the genes.
Other challenges that were discussed briefly include the limited availability of translatable biomarkers for target engagement, including cerebrospinal fluid (CSF) circulating RNA and EEG-based digital endpoints, and regulatory uncertainty surrounding novel ASO chemistries and gene editing platforms, particularly in relation to manufacturing standards and long-term monitoring.
Disease model systems
In vitro models
Participant discussions on model systems revealed a marked shift toward human-derived platforms. Induced pluripotent stem cell (iPSC) lines and cerebral organoids were prioritized for their translational relevance and capacity to model human neurodevelopmental phenotypes. It was also noted that iPSC models can be made from patient and diverse cells to enable studies of how genetic background interacts with specific variants.
In vivo models
The relevance and utility of in vivo models were also discussed briefly. These models have been instrumental in elucidating how SCN2A or SCN8A dysfunction alters cellular and circuit physiology, revealing widespread effects on dendritic excitability, glial activity, dopaminergic signaling, myelination, and synaptic plasticity [Reference Wang, Derderian and Hamada31,Reference Thompson, Miralles and Wengert43–Reference Li, Huang and Hu49]. Moreover, findings from these preclinical systems are beginning to inform potential biomarkers in humans with SRDs. Emerging candidate biomarkers beyond seizure burden include alterations in the vestibulo-ocular reflex (VOR) [Reference Wang, Derderian and Hamada31], quantitative EEG measures [Reference Hudac, Dommer and Mahony50], noninvasive “wearable” neurophysiological monitoring [Reference Chalmers, Hickey and Newton51], and sleep disturbances associated with Scn2a loss-of-function variants [Reference Ma, Eaton and Liu18]. Participants also identified sensory processing, working memory, and motor control as promising domains for biomarker discovery and translational alignment between basic and clinical research.
However, GI dysfunction and autonomic dysregulation, which were highlighted by caregivers during the session as high-priority clinical concerns, remain unstudied in animal models of SRDs. Studies of Dravet syndrome mouse models have begun to look to at dysregulated autonomic function, including core body temperature, heart rate, and breathing rate [Reference Sahai, Bard, Devinsky and Kalume52]. Extending these studies to SCN2A and 8A models would help address these non-seizure outcome priorities.
Recommendations
Building on the workshop findings, several key recommendations emerged to accelerate translational progress and improve therapeutic readiness across the field (Figure 3). First, there is a clear need to invest in the validation of non-seizure outcome measures, particularly those that capture meaningful changes in cognition, behavior, and GI function. Personalized outcome measurements are currently in development for patients severely affected with SRDs [ Reference Sevinc, Knox and George53,Reference Downs, Ludwig and Wojnaroski54]. Utilization of these outcome measures as clinical trial endpoints for FDA approval was noted as a challenge, with some stakeholders emphasizing seizure reduction as necessary for regulatory and payer alignment. Future measures should be complemented by biomarker correlation studies to strengthen the link between clinical outcomes and underlying disease mechanisms. Participants also emphasized the importance of developing translational models that reflect GI and autonomic symptoms, which remain underrepresented in current preclinical systems. Such models would enable a more holistic understanding of disease pathology and therapeutic effects beyond seizure reduction.
Key findings and recommended strategies. Four key findings from the meeting and the strategies that were recommended by the participants. SNP = single-nucleotide polymorphisms.
Second, there was much excitement surrounding possible ASO treatments, but concerns about delivery strategies and wider applicability of treatments were expressed. Stakeholders recommended focusing on next-generation delivery strategies (such as vectorized ASOs) and initiating long-read sequencing efforts to more precisely define single-nucleotide polymorphism (SNP) targets for allele-specific therapeutic platforms, including ASO and gene editing strategies.
Third, the exercise highlighted important differences in caregiver risk tolerance for N-of-1 approaches. Caregivers highlighted the need for clearer risk frameworks, especially around irreversible modalities such as gene editing. Caregiver risk tolerance varied with age, disease severity, and prior treatment experience. To address this, stakeholders recommended focusing on de-risking precision genetic treatments by focusing on variant stratification as well as developing and testing better human-relevant models.
Fourth, the field should prioritize decentralized, caregiver-informed clinical trial designs that reflect the lived experiences of families and reduce participation barriers. Early regulatory engagement with the FDA was also identified as critical, particularly for programs involving novel chemistries and mixed-function variants that may not fit existing trial frameworks.
Finally, the group encouraged the publication of case-based therapeutic roadmaps to capture lessons learned across modalities and to inform future platform-based therapeutic development.
Feedback and lessons learned
A post-hoc survey was distributed to participants to solicit feedback and evaluate the roundtable’s effectiveness across stakeholder groups. Overall, the day-long session was rated highly, although response was low (10 of ∼100 participants), likely due to the delayed distribution of the survey. Among respondents, six identified primarily as researchers, one as an industry representative, one as a patient/family advocate, and two as holding dual researcher-advocate identities. The relevance of topics discussed and the overall organization were rated Excellent or Good by all respondents. The roundtable workshop itself was rated Excellent by 8 respondents and Fair by 2 respondents, and all respondents indicated they would recommend this meeting to a colleague. Feedback on the inclusion of patient and caregiver perspectives was viewed positively by 9 of 10 respondents; they described these perspectives as “invaluable,” “humanizing,” and “centering,” highlighting the importance of firsthand insights into SRD manifestations for guiding research priorities and risk assessment. One researcher noted that the discussion with the caregiver limited time for deeper technical exchanges, stating, “our group spent too much time explaining things to the family advocate,” describing it as “a missed opportunity to discuss aspects of the therapeutic plan.”
The post-hoc survey also asked how respondents felt about the cross-disorder approach combining SCN2A and SCN8A. Overall, the joint nature of the roundtable was viewed positively by all respondents—researchers called it “a big win,” “positive,” and “a very nice addition,” and family advocates were similarly positive, calling it “very timely” and “successful.” The industry professional also felt it was a very positive and useful exercise, stating, “I loved it and hope you continue. So much overlap and learnings from one another.” Ultimately, the benefits of combining the two communities outweighed any disadvantages and allowed roadmap-building to transfer across conditions rather than duplicating efforts.
The integration of caregiver perspectives into this workshop consistently reshaped scientific priorities toward domains reflective of lived experience, including GI function, sleep, autonomic dysregulation, communication, and developmental progress. The exercise also surfaced important differences in stakeholder assumptions, such as varying tolerance for N-of-1 approaches, differing views on patient and caregiver tolerance of monthly intrathecal injections, and reliance on seizure reduction as a regulatory endpoint. Caregiver perspective also had a strong influence over the therapeutic modalities that were discussed. Based on the focus of the industry participants, we had expected that 2 groups would discuss ASOs, 4 groups would discuss gene editing strategies, and 4 would discuss small molecule development (Figure 1). Instead, 6 groups discussed ASOs. This reflects the strong influence of the caregiver perspective in each group, as their preference ultimately drove the discussion. Future roundtables will maintain caregiver participation at approximately 10%, with one caregiver represented at each table to ensure lived-experience input without overwhelming technical discussion. Although comprehensive booklets were provided at each table, participants noted that caregivers would benefit from receiving these materials in advance in a more accessible, lay-friendly format, including a glossary of scientific and regulatory terms. Scientist participants should also receive brief guidance on how to engage sensitively and constructively with caregiver perspectives during discussions.
Finally, several practical and logistical lessons emerged. Participants expressed interest in more time for modality-specific deep dives, including short primers on technologies such as vectorized ASOs, base and prime editing, and decentralized trial designs before the breakout work begins. They also recommended assigning a trained facilitator to each group, using mixed-media tools (large-format templates, sticky notes, and digital capture), and administering surveys in person at the end of the session to improve response rates and reduce recall bias. Together, these adjustments are expected to improve the balance between scientific rigor and patient-centeredness, make discussions more efficient, and enhance the reproducibility of this roadmapping model for other rare epilepsies and neurodevelopmental disorders.
Conclusions
This multi-stakeholder roadmapping workshop underscores both the accelerating momentum and the persistent challenges in advancing therapies for SRDs. Participants converged on allele-specific ASOs, emerging gene editing platforms, and targeted small molecules as the most promising therapeutic modalities yet emphasized that progress will continue to be constrained without improved variant stratification, human-relevant preclinical models, and validated non-seizure outcome measures. These insights refine and extend previous SCN2A and SCN8A research roadmaps by translating broad priorities into concrete, modality-specific milestones that span discovery, preclinical development, and clinical trial readiness.
This proceedings report offers a practical model for accelerating translational progress in rare epilepsies: convening diverse stakeholders, embedding caregivers as partners, grounding discussions in shared reference materials, and using structured roadmap templates to identify gaps and actionable next steps. While the findings reflect the perspectives of a single convening with a modest post-hoc survey response rate, the approach provides a replicable framework for other developmental and epileptic encephalopathies.
Author contributions
Christina SanInocencio: Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing, Raquel Miralles: Conceptualization, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing, Gabrielle Conecker: Supervision, Writing – original draft, Hannah Monday: Writing – original draft, Writing – review & editing, Leah Schust: Conceptualization, Resources, Writing – original draft, Megan Abbott: Writing – original draft, Morgan Robinson: Writing – original draft, Writing – review & editing, Yael Weiss: Resources, Morgan Weberg: Writing – original draft, Writing – review & editing, Vanessa Bender: Project administration, Visualization, Writing – original draft, Writing – review & editing.
Funding statement
This work was supported by the FamilieSCN2A Foundation. The authors received no external grants or funding for the preparation of this manuscript.
Competing interests
Y. Weiss, CEO Mahi Therapeutics, no other conflicts to declare.
Open access
