Jun. 24 at 4:55 PM
$KYTX TLDR: A novel in vivo CAR-T approach could reshape the autoimmune treatment landscape. For both investors and researchers, it’s a space worth watching. KYTX’s ex vivo lead remains relevant, but the field is clearly evolving.
A recent Science publication by Hunter et al. (June 2025) introduces a potentially disruptive approach to CAR-T therapy: the generation of CAR-T cells directly in vivo via lipid nanoparticle (LNP) delivery of mRNA targeting CD8+ T cells. In both rodent and non-human primate models, this method achieved effective tumor control and deep B-cell depletion without the complex infrastructure and lymphodepletion required for traditional ex vivo CAR-T manufacturing. Importantly, the resulting B cell repopulation in primates skewed naïve, suggesting a possible "immune reset" that could be relevant to both cancer and autoimmune indications.
This platform could mark a significant paradigm shift, especially for autoimmune applications, which have long been constrained by manufacturing limitations, cost, and patient eligibility. Rather than harvesting, engineering, expanding, and reinfusing cells -- as with traditional autologous CAR-T -- this method introduces the potential for a single-step, off-the-shelf therapy that is scalable, repeatable, and more broadly accessible.
For those who follow KYTX closely, you already know their lead clinical programs focus on Stiff Person Syndrome (SPS) and Myasthenia Gravis (MG), both B cell–mediated autoimmune diseases. While lupus nephritis and systemic sclerosis are part of their broader pipeline, the clinical spotlight remains on SPS and MG, where KYV-101 (a CD19-directed autologous CAR-T) is already in active trials. In that context, the in vivo approach presented by Hunter et al. is worth watching -- it may not immediately displace existing strategies, but it introduces a potential technological pivot that could influence how future therapies are developed.
Key Difference:
Ex vivo CAR-T (like KYV-101 / KYTX)
Cells are removed from the patient, genetically modified in a lab (often using viral vectors), expanded, then re-infused into the patient.
Requires external cell processing infrastructure, GMP manufacturing, cryopreservation, and often lymphodepleting chemotherapy before reinfusion.
It’s a personalized product -- unique to each patient (autologous).
In vivo CAR-T (like the Hunter et al. LNP model)
Instead of removing cells, the therapy is delivered directly into the patient via lipid nanoparticles (LNPs) carrying mRNA that reprograms the patient’s T cells inside the body.
No external cell manipulation, no expansion, and no complex lab infrastructure -- a single administration, off-the-shelf solution.
KYTX has built its platform around the ex vivo model, with strong scientific rationale, early clinical traction, and significant institutional support. But should the LNP in vivo model mature into a new standard of care, companies like Kyverna may face pressure to adapt their development pipelines accordingly.
That said, KYTX currently remains ahead in clinical development and regulatory positioning. Their construct has already demonstrated encouraging safety and efficacy in early autoimmune trials, and their commercial roadmap is based on real-world infrastructure already validated in oncology CAR-T. In addition, their management team brings deep M&A and clinical execution experience that could prove essential if the field shifts dramatically.
Bottom line: this new research doesn’t diminish the clinical or near-term investment case for KYTX. But for long-term, platform-level investors, it raises an important question -- will the field evolve away from ex vivo approaches, and if so, who’s best positioned to pivot or integrate? These are the kinds of inflection points that shape the next generation of biotech winners. I share this as part of my ongoing process -- constantly stress-testing assumptions, revisiting data, and staying open to where the research leads. As always, this reflects my own views. Do your own due diligence before making decisions. Cheers.
