The neurodegeneration therapeutic landscape in 2025 is defined by a sharp contrast between the progress of antisense oligonucleotide (ASO) and gene therapy approaches on one hand, and the near-total absence of approved mRNA therapeutics for any neurological indication on the other. This is not because mRNA has been overlooked — several programs are advancing — but because the delivery problem for CNS mRNA is substantially harder than the delivery problem that the COVID vaccine platform solved. Understanding where mRNA fits in the broader landscape requires first understanding what adjacent modalities have achieved and why.
What ASO Therapies Have Demonstrated
Antisense oligonucleotides were the first nucleic-acid-based modality to reach clinical approval for neurological disease in a significant way. Nusinersen (Spinraza, Biogen), approved by FDA in December 2016, is an ASO administered by intrathecal injection that modifies the splicing of SMN2 pre-mRNA to increase production of functional SMN protein in spinal muscular atrophy. The mechanism is elegant: the SMN2 gene has a C-to-T mutation in exon 7 that causes exon skipping, and nusinersen binds the intronic splicing silencer to block this skipping, restoring near-normal splicing. The protein is the patient's own SMN2-derived protein — the drug modifies the endogenous mRNA processing rather than delivering a therapeutic mRNA.
Tofersen (Qalsody, Biogen), approved by FDA in April 2023 for SOD1-ALS, is an ASO that targets SOD1 mRNA for degradation by RNase H1 — a knockdown approach. Patients with SOD1-mutant ALS carry a toxic gain-of-function protein, so silencing the mutant allele is directly therapeutic. The intrathecal route deposits the ASO directly into cerebrospinal fluid, bypassing the BBB entirely. This is the key delivery strategy that has made ASOs viable for CNS targets: route around the blood-brain barrier rather than cross it.
The limitations of ASOs are also instructive for thinking about mRNA. ASOs work well for knockdown (via RNase H or steric block of translation) but are fundamentally limited as upregulation tools. You can use an ASO to block a negative regulatory element and thereby increase expression of a target, as nusinersen does, but you cannot use an ASO to deliver a functional gene product. For longevity-protein upregulation — expressing more Klotho, more FOXO3, a variant protein with superior activity — you need a coding sequence. ASOs and siRNA cannot do this; mRNA can.
Gene Therapy in CNS: What Zolgensma Teaches Us
Onasemnogene abeparvovec (Zolgensma, Novartis), approved by FDA in May 2019 for SMA in pediatric patients, delivers a functional SMN1 copy via adeno-associated virus serotype 9 (AAV9). AAV9 has natural CNS tropism — it crosses the BBB with reasonable efficiency (better than most non-viral vectors, though the mechanism is incompletely understood and involves some transcytosis), and after IV administration achieves transduction of motor neurons and other CNS cell types. The viral capsid has been selected through billions of years of evolution to be good at entering cells.
Zolgensma demonstrates what is possible when delivery is solved — the clinical outcomes in treated infants with SMA Type 1 are dramatic relative to the natural history. But the gene therapy platform carries its own constraints. AAV has a packaging capacity of roughly 4.7 kb for the therapeutic sequence, which excludes large genes. Manufacturing is complex and expensive. Immunogenicity to the AAV capsid limits re-dosing — most patients develop neutralizing antibodies after a single administration, meaning you typically get one therapeutic window. And because AAV integrates at low frequency (largely episomal) but can integrate, there are theoretical long-term genotoxicity concerns that become more prominent for chronic indications in non-pediatric patients with many decades of life ahead.
mRNA avoids most of these constraints: there is no DNA integration, no packaging size limit (practical mRNA therapeutics can carry transcripts of 2–8 kb or larger), and dosing can in principle be repeated. The cost is that mRNA expression is transient, which means chronic conditions require repeat dosing and a delivery vehicle that is safe for repeated administration.
mRNA Programs Entering the CNS Pipeline in 2024–2025
Moderna's mRNA-3927 is an mRNA therapy for propionic acidemia, a metabolic disease where two enzymes (PCCA and PCCA) in the liver are deficient. This is a peripheral (liver) mRNA program, not CNS — it is included here because it represents the state of the art for disease-indication mRNA therapeutics entering IND stage. The program demonstrates that mRNA can be used to replace enzyme function in a metabolic disease setting, which is mechanistically analogous to what a longevity protein replacement program would require, just in a different organ.
BioNTech has disclosed a CNS pipeline that includes mRNA-based programs for neurological indications, though detailed pipeline information is limited in public filings as of early 2025. The company's mRNA technology platform (including modified nucleotides and specialized LNP formulations) has been directed at CNS targets, but the specific programs and their stage have not been publicly detailed in peer-reviewed literature. We are noting this as publicly reported pipeline activity, not making claims about specific program data.
Several academic groups have published CNS mRNA delivery data that is relevant to the development of therapeutic programs. Delivery of mRNA encoding anti-inflammatory interleukins to rodent brains via intrathecal administration, intranasal LNP delivery with documented olfactory bulb expression, and receptor-targeted LNPs using transferrin receptor targeting have all been reported in the literature within the last three years. These are rodent studies and the translation to non-human primate or human is not validated, but they establish proof-of-concept for the approach.
The Approval Gap for Neurodegeneration mRNA
As of April 2025, there is no approved mRNA therapeutic for any neurodegenerative indication. The reasons are worth examining because they inform the timeline for the entire class of programs, including caVos's work.
First, the delivery problem is genuinely unsolved at the clinical stage. Every mRNA CNS program must confront the BBB crossing problem or use an intrathecal/intranasal route with its own limitations. Second, neurodegeneration clinical trials are among the longest and most expensive in medicine. The endpoints for Alzheimer's disease, Parkinson's disease, or ALS trials require measuring slow functional decline over 18–36 months in patient populations with variable progression rates. Demonstrating efficacy for a longevity-associated protein upregulation approach requires either very long trials or validated surrogate biomarkers, neither of which is straightforward. Third, the regulatory pathway for mRNA neurodegeneration therapeutics has less precedent than for, say, ASO neurodegeneration programs, and FDA guidance on clinical translation of mRNA CNS delivery is still evolving.
We are not saying these barriers make mRNA neurodegeneration therapeutics unviable. We are saying they impose a realistic timeline: programs entering IND in 2024–2025 should not be expected to reach Phase 2 readouts until 2028–2030 at the earliest, and approval before 2032 for a first-in-class indication would require an unusually clean data package.
Where caVos Fits in This Landscape
caVos is in preclinical stage. We do not have programs in IND, and we are not projecting specific clinical timelines. What we are doing is building the platform foundation — target identification via comparative genomics, mRNA sequence design and optimization, and initial LNP formulation work — that would be the basis for a future IND-enabling program.
Our focus on longevity-associated proteins rather than acute disease targets reflects a specific strategic bet: the targets most likely to show meaningful benefit in neurodegeneration are those that slow or reverse fundamental aging processes in neurons and supporting cells, not those that address a single disease-specific protein (like mutant SOD1 in ALS). This is a longer-horizon bet — the clinical trials would be harder to design and run — but the potential magnitude of benefit if it works is qualitatively different. A Klotho upregulation approach that reduces age-related neurodegeneration broadly is more valuable than a therapy that slows one genetic subtype of ALS, even though the genetic ALS target is more tractable to develop.
The competitive landscape we are operating in is therefore not primarily the other mRNA neurodegeneration programs, most of which are targeting disease-specific mechanisms. Our longer-term competitive reference is the longevity/geroprotector space more broadly: senolytics, rapamycin derivatives, NAD+ precursors, and other systemic aging interventions. mRNA upregulation of specific longevity-associated proteins is a more targeted and mechanistically specific approach than these systemic interventions, which is the core of our differentiation thesis. Whether that specificity translates into a better therapeutic index than systemic geroprotectors is a hypothesis we need animal data to test, not a claim we can make now.
What the Next Two Years Require
For the mRNA neurodegeneration field to advance meaningfully, several things need to happen that are not specific to caVos. CNS delivery needs to be validated in a non-human primate model with quantitative data on brain region-specific mRNA distribution and protein expression. The immunogenicity profile of repeat-dosed modified mRNA LNPs needs to be characterized in chronic safety studies. Biomarkers that can serve as early surrogate endpoints for CNS mRNA delivery (cerebrospinal fluid protein levels, neuroimaging-based proxies for neuronal health) need to be validated against clinical outcomes in existing ASO and gene therapy trials that have good biomarker data.
None of this is caVos's work alone — it is a collective scientific infrastructure that the field is building. We track these developments closely because the clinical translatability of our own programs depends on them. The fastest path to a viable caVos clinical program runs through the general CNS mRNA delivery problem being solved by the community, not just by us. We are invested in the field succeeding as much as we are invested in our specific programs advancing.
