Capstan Therapeutics: LNP Delivery Beyond the Liver $abbv
In-vivo CAR-T, Targeted LNPs, and Autoimmune Disease
I’ve written previously about CAR-T therapies, including on the drawbacks of existing FDA approved CAR-T treatments, the promise of CRISPR/gene-editing in engineering allogeneic approaches, as well as on an instruments business powering these gene-editing players.
AbbVie made headlines recently with its ~2 billion USD all cash acquisition of Capstan Therapeutics. Capstan’s lead product, CPTX2309, is a CAR-T treatment for B-cell mediated autoimmune diseases that’s in phase 1 trials. What makes Capstan interesting is not that it’s developing an autoimmune disease treatment, nor that it’s developing one that leverages a CAR-T approach, nor that its target antigen is CD19. Rather, what makes the company so fascinating is its potential therapy:
Is in vivo, and so doesn’t require a specialized treatment center, an initial regimen of chemotherapy, harvesting a patient’s cells, a weeks to months period of editing those cells, or an eye-popping sticker price. Critically, it also doesn’t require donor cells, and so unlike other allogeneic treatments in clinical trials one doesn’t have to worry about mitigating graft vs host disease.
Delivers the chimeric antigen receptor non-virally. mRNA coding for the CAR is contained inside an LNP.
Is part of Capstan’s broader tLNP platform. This platform is exciting because it’s an instance of using LNPs to reach something other than the liver. That requires some innovation!
There’s more to be said on the above, starting with (3). Currently, LNPs are limited in that they really can only target the liver.1 Capstan has found a way around this impediment: first, they were able to design an LNP that primarily accumulates in the spleen. The spleen makes white blood cells, so accumulation here is very helpful when trying to affect the immune system! Secondly, Capstan conjugated an antibody fragment to the LNP. This antibody fragment enables the lipid nanoparticle to bind precisely to a specific cell type. In the case of Capstan’s lead product, the fragment facilities LNP binding to CD8+ T-cells. The mRNA is then delivered into these cells, and the chimeric antigen receptor is successfully expressed.2
Capstan’s decision to target CD8+ T-cells differs from typical CAR-T approaches which target both CD8+ cells and CD4+ cells. The rationale for this is twofold: CD4+ cells are often overexpressed in patients with lupus, so you may not want to stimulate additional activity here. Moreover, CD4+ cells are responsible for driving cytokine release syndrome (CRS), an excessive immune response that’s a well known side-effect of CAR-T treatments.3
That mitigation of side-effects is key, and leads to a related one on why point (2), Capstan’s non-viral approach, is so important. Treating autoimmune disease is a very different thing from treating refractory/relapsed cancers. If patients with serious cancer don’t get treatment they die; that is thankfully not the case for those with autoimmune disease. To be sure, autoimmune disease can wreak havoc. Organ damage, especially to the kidneys, is a real risk for those with lupus, and it’s estimated that 10-15% of those with this disease will die prematurely.4 That’s serious, but still quite a different prognosis than for any cancer patient eligible for CAR-T treatment. There’s also a reason that this piece by Capstan researchers mentions the total number of people affected by B-cell mediated autoimmune disease: the goal is to develop a treatment for the whole patient population, not only those with the most severe instantiations of it. For the majority of those with a B-cell mediated autoimmune disease, it wouldn’t make sense to consider CAR-T treatment when the side-effects can be life-threatening. Taking steps to minimize the odds of cytokine release syndrome is a meaningful stride towards widening the potential addressable patient population. The second, and perhaps more important, stride that Capstan has taken here comes back to its decision to deliver the CAR non-virally.
I’ve covered the potential benefits of non-viral CAR delivery previously. The key concern for viral methods centers around the semi-random integration of the lentivirus/retrovirus into the T-cell DNA. This comes with the risk of insertional mutagenesis, whereby the viral DNA integration causes a mutation within the T-cells. These mutations could theoretically cause a second cancer.
I say theoretically, but there have been cases of T-cell lymphoma in patients who received CAR-T treatments, and the FDA now requires that treatment providers list secondary malignancies as a potential side-effect. It’s so far been very difficult to discern precisely what’s causing these secondary cancers. Their incidence is thankfully low, and for a patient with already serious cancer the risk/reward makes sense. But again, that calculation changes when it’s a patient with even a severe case of autoimmune disease. Delivering the CAR via LNP, as Capstan does, avoids that insertional mutagenesis risk altogether. Furthermore, it meaningfully differentiates the company from other players in the in-vivo CAR-T space. Interius (acquired by Gilead), EsoBiotec (acquired by AstraZeneca), Umoja (recently raised 300mm), Novartis, and Kelonia are all working on in-vivo CAR-T treatments, but all of them remain constrained by lentiviral delivery.
There’s a natural question here as to whether Capstan’s approach threatens existing CAR-T cancer treatments. In the near-term I think this answer is no. For those with cancer, the persistence of CAR-T cells in a patient post-treatment is seen as critical to keeping him/her successfully in remission. For those with autoimmune disease that’s not the case, which is exactly why the CAR can be delivered in the form of mRNA. Transient expression is good enough here! The obvious solution to modifying Capstan’s approach for cancer would be repeat dosing, but then one has the added hurdle of worrying about the immune response to repeated mRNA intravenous treatments.
Disclaimer: The information in this post is not intended to be and does not constitute investment or financial advice. You should not make any decision based on the information presented without conducting independent due diligence.
I’m speaking here of LNPs delivered intravenously, rather than LNPs used in the Covid vaccines. In the case of Covid vaccines these reach the muscular cells, but that’s because they’re delivered via injection. LNPs’ delivery limitations are why companies developing gene therapies often are not using this vessel when attempting to reach cells outside the liver. Innovation around LNP delivery gets very exciting in the gene therapy context because it gives one the opportunity to re-dose a patient. Such an opportunity does not exist when one uses a viral vector.
In some ways this approach is conceptually to antibody-drug conjugates used in the treatment of cancer.
This could actually be a reason that Capstan’s approach ends up failing. CAR-T CD4+ and CD8+ cells work together when used to fight cancer, and an absence of CAR CD4+ cells could lead to T-cell exhaustion.