Here’s the thing about Immunofoco that most people missed in the AACR noise this morning.

They’re not some new entrant chasing the in vivo CAR-T hype.

The Suzhou-based company already has IMC002, a CLDN18.2 CAR-T, in a Phase 3 gastric cancer trial. The Phase 1/2a readout was a 66.7% objective response rate in advanced gastric and GEJ cancer. That’s actually some serious data, in a real commercial indication, from an ex vivo cell therapy program that’s actually working.

So when I saw Immunofoco presented in vivo CAR-T data at AACR this morning, I figured I’d give it a look…and I was surprised at what I saw.

It wanted to just write them off as another company hopping on the in-vivo CAR-T bandwagon, but I think it’s deeper than that.

It’s seems more like a company that knows how to make CAR-T work in patients deciding the next version happens inside the patient.

The mechanism is where this gets interesting.

Most in vivo CAR-T programs are using LNPs with mRNA.

Orna was circular RNA plus LNP, and that’s the approach Lilly paid $2.4B for.

Capstan (acquired by AbbVie last year) used the same general idea.

Immunofoco is going a different direction.

Their iMAGIC platform uses a lentiviral vector with a mutated MxV glycoprotein (MxV-G-mut) paired with a T-cell targeting module called TCM3.

The pitch? They have a new approach to T cell-targeting and can selectively transduce T cells  -without the misses of previous platforms. They also integrate the CAR construct into the genome to get durable expression instead of the transient expression you get with mRNA.

Other companies – I’m thinking of Interius (acquired by Gilead, Kite) – use lentiviral approaches for full integration. I happen to agree with this approach over LNP or AAV’s transient nature.

Durable matters.

Part of the pitch for in vivo CAR-T over traditional autologous is eliminating manufacturing.

The other part is getting something that lasts.

mRNA and AAV approaches are great for safety (the CAR goes away if something goes wrong) but that transience is also a limitation for cancers where you need sustained tumor control.

And after all, ex vivo CAR-T use integrating viruses … so why can’t we do it with in vivo? I get there’s a greater risk for non-T cell infection (and then you get a muscle cell or something expressing a CAR), but Immunofoco’s enhanced T cell targeting should – at least partially – solve this.

The preclinical data from today was what you’d expect at this stage.

Tumor burden reduction and durable tumor control across multiple xenograft models, selective T-cell transduction in vitro, cytotoxicity against myeloma cells, IFN-γ upregulation.

Safety looks good and there was no percent-reduction or survival numbers disclosed — both to be expected for AACR preclinical posters.

Now to the target: BCMA. I think this is a smart target for a second act.

AstraZeneca’s ESO-T01 just read out Phase 1 data in Nature Medicine — 4 of 5 heavily pretreated myeloma patients responded, three achieved stringent complete remissions with MRD negativity.

So, the target is validated in vivo.

The question is who gets there with the best combination of manufacturing simplicity, durability, and safety.

Immunofoco’s bet is that lentiviral integration beats mRNA for durability, and that the MxV-G-mut envelope gives them the T-cell selectivity that’s been the hardest problem to crack in this field.

No IND timing has been disclosed. But a company that’s already running a Phase 3 CAR-T isn’t fumbling for trial design. The infrastructure is there. What’s new is whether the delivery platform holds up in humans.

That’s the story to watch for the next 12 months.