For a decade, CAR-T meant a factory: collect the patient's T cells, ship them to a clean room, engineer and expand them for weeks, ship them back and infuse after chemotherapy. In 2026 a new generation of therapies is trying to skip almost all of that by engineering the T cells inside the patient. For CMC, quality and supply chain teams, that is not a small change. It moves the product from a living cell to a vector.

$7B
Lilly's deal for Kelonia (up to, April 2026)
$2.1B
AbbVie's acquisition of Capstan (up to, 2025)
$1B
AstraZeneca's acquisition of EsoBiotec (up to, 2025)
0
lymphodepletion in Legend's LB2501 phase 1 at EHA 2026

From ex vivo to in vivo

Classic autologous CAR-T is a personalised manufacturing chain. Every batch is one patient, every batch has its own chain of identity, and vein-to-vein time is measured in weeks. In vivo CAR-T replaces the chain with an off-the-shelf medicine: a vector that carries the CAR gene, or CAR mRNA, to T cells after an intravenous infusion.

EX VIVO CAR-T · WEEKS Apheresispatient cells Cold chainto factory Engineer & expand1 batch = 1 patient QC releaseper batch Cold chainback to clinic Lymphodepletionthen infusion IN VIVO CAR-T · OFF THE SHELF Vial: vector or LNP IV infusion Targeted particles find T cells and deliver the CAR gene or mRNA CAR-T made in the body
Figure 1. In vivo CAR-T removes the patient-specific manufacturing chain. The engineering step happens inside the patient.

Why 2026 is the inflection point

Big pharma has placed its bets. AstraZeneca completed its acquisition of EsoBiotec and its lentiviral ENaBL platform in May 2025, for up to $1 billion. AbbVie agreed to buy Capstan for up to $2.1 billion, gaining CPTX2309, an anti-CD19 CAR mRNA in a targeted lipid nanoparticle designed for autoimmune disease. In April 2026, Eli Lilly announced it would acquire Kelonia Therapeutics for $3.25 billion upfront and up to $7 billion in total.

The clinical signals are early but striking. Kelonia's KLN-1010, a one-time intravenous gene therapy that generates anti-BCMA CAR-T cells in multiple myeloma, was presented in the plenary session at ASH 2025. Updated data reported around ASCO 2026 showed objective responses in all 18 treated patients. In June 2026, Legend Biotech presented LB2501, a CD19/CD20 dual-targeting in vivo CAR-T, at EHA: at the higher dose, 6 of 6 patients with relapsed or refractory B-cell lymphoma responded and 5 of 6 achieved a complete response, without lymphodepletion and with no dose-limiting toxicities reported.

These are phase 1 numbers in small cohorts. Durability and safety still have to be proven. But the manufacturing question has already changed.

What changes for CMC and quality

When the product is a vector or an LNP, the critical quality attributes move with it. Teams that grew up on cell therapy release need to plan for a different control strategy.

Product definition
The drug substance is the vector or mRNA-LNP, not cells. In the EU, an in vivo gene-transfer product is regulated as a gene therapy medicinal product (an ATMP); mRNA-LNP products need case-by-case classification.
Potency and targeting
Potency must show selective T-cell transduction or transfection and CAR function. Expect orthogonal assays and an early, clear link between targeting attributes and clinical activity.
Safety attributes
Off-target transduction, replication-competent virus for lentiviral platforms, integration profile, empty or partial particles and lipid impurities become central release and characterisation topics.
Scale and supply
One batch can treat many patients. Chain-of-identity logistics fade; batch consistency, stability, cold chain and vector capacity become the bottlenecks.

What it could mean for the Arab world

Ex vivo CAR-T has been hard to scale in the region: few qualified treatment centres, long shipping routes to manufacturing sites abroad, and weeks of vein-to-vein time. An off-the-shelf vial changes that equation. It could be stocked, distributed and administered through ordinary specialist hospital pathways. It also makes in vivo platforms a realistic candidate for regional fill-finish and, later, local vector manufacturing, which fits national localisation agendas.

Regulators will still need ATMP-specific capability for review, GMP inspection and long-term follow-up. Saudi Arabia's SFDA has already issued dedicated GVP guidance for ATMPs, taking effect in April 2027, which shows where the region is heading.

WHAT TO DO NOW
1Map your current cell-therapy control strategy against vector and LNP attributes: which assays transfer, which are new.
2Review classification and regulatory pathways early in each target market, including scientific advice.
3Plan vector capacity and cold-chain distribution as part of the development programme, not after it.
4Build long-term follow-up and pharmacovigilance for gene therapy into the RMP from the first clinical study.

Molgenium supports ATMP and mRNA developers with CMC strategy, GMP readiness and regulatory planning for EU and MENA markets. Talk to our experts if you are preparing an in vivo programme.

Sources