The Company That Invented CAR-T Is Watching From the Sidelines

Why In Vivo CAR-T Has Everyone So Excited

To understand the frenzy, you need to understand the problem. Current CAR-T therapy is like ordering a custom suit that takes weeks to tailor, costs a fortune, and only a handful of shops in the world can make it.

The process works like this: doctors extract a patient's T cells, ship them to a specialized facility, genetically engineer them to recognize cancer, grow them in bulk, then ship them back and infuse them into the patient. The whole cycle takes two to four weeks. It costs between $373,000 and $475,000 per treatment. And many patients with aggressive diseases get sicker (or die) while waiting.

In vivo CAR-T skips almost all of that. Instead of modifying cells in a lab, it delivers the genetic instructions directly into the patient's bloodstream using viral vectors or lipid nanoparticles (tiny fat bubbles that carry genetic cargo into cells). Your body's own T cells get reprogrammed on the spot. No extraction, no factory, no weeks of waiting.

It's the difference between that custom suit and buying one off the rack that somehow fits perfectly.

Still Very Early, Still Very Expensive

The excitement is real, but so is the risk. Only a handful of in vivo CAR-T programs have reached Phase 1 clinical trials so far. None have advanced to Phase 2. The early results are intriguing; Kelonia's lead candidate showed "highly encouraging" safety data, and EsoBiotec reported an 80% response rate in a small group of multiple myeloma patients. But Leerink analyst Daina Graybosch has cautioned against "overreactions to small datasets," noting that initial patients in these trials tend to be heavily selected.

The market projections tell you where Wall Street thinks this is going, and they explain why companies are paying billions for Phase 1 assets.

The Shrinking Shopping List

Novartis has a timing problem. The company previously explored in vivo CAR-T back in 2024 through a collaboration with Vyriad's lentiviral vector platform. Nothing came of it. And the number of independent targets worth acquiring is shrinking fast.

Zercher flagged Umoja Biopharma as one of the prime remaining acquisition targets, though it already has ties to AbbVie. Analysts have also pointed to offshore options like China's Starna Therapeutics. But the menu is getting shorter by the quarter.

Multiple analysts have explicitly predicted that Novartis (along with Amgen and J&J) will eventually need to make a deal. Narasimhan's M&A strategy has historically favored sub-$2 billion upfront deals for early- to mid-stage assets, which could limit the company's options as valuations in the space climb higher.

The Pioneer's Dilemma

Novartis finds itself in a familiar corporate paradox. The company that created the first generation of a technology is often the slowest to embrace the next one. There's an installed base to protect (Kymriah and its successors), an existing manufacturing infrastructure built around the old approach, and an institutional belief that incremental improvement beats radical reinvention.

But the whole promise of in vivo CAR-T is that it could make the current manufacturing model obsolete. If the technology works as hoped, the multimillion-dollar cell therapy factories that Novartis operates become expensive relics.

Narasimhan clearly knows this. His comments suggest a company that's paying close attention but hasn't decided whether to leap. The question is whether "continuing to evaluate" turns into action before the best opportunities are gone.

Because in this particular feeding frenzy, the pioneer doesn't get extra time at the buffet.