Autologous chimeric antigen receptor (CAR) T-cell therapy reaches patients weeks after the decision to treat. Patients with rapidly progressing disease often need bridging therapy during that interval or become ineligible before infusion, and manufacturing failures, limited production slots, and cost restrict access further. Some first-in-human trials test a way to skip the factory step: generating CAR-T cells inside the patient.
In vivo CAR-T therapy replaces leukapheresis and ex vivo manufacturing with a single intravenous infusion of genetic instructions. Endogenous T cells take them up, express a CAR, and recognize malignant or autoreactive cells. Several early programs also give the infusion without lymphodepleting chemotherapy.
Two delivery platforms dominate clinical investigation. Engineered lentiviral vectors introduce CAR-encoding genetic material that integrates into the genome and can sustain CAR expression after a single infusion. Lipid nanoparticles deliver messenger RNA without integration, producing transient CAR expression that repeated dosing can renew. Prolonged activity suits sustained cancer control, and temporary, controllable B-cell depletion could suit autoimmune disease.
Making CAR-T cells in the body removes quality checks that conventional manufacturing performs before infusion. Viral vectors must enter T cells selectively, without modifying malignant cells or other populations. Genomic integration requires long-term monitoring for insertional effects, and the number and phenotype of engineered T cells cannot be controlled through ex vivo expansion.
Nonviral delivery avoids permanent genetic modification but faces challenges in tissue distribution, expression duration, and repeated dosing. Both approaches must show reliable CAR-T generation, expansion, and persistence with acceptable immune-related toxicity (Vilcot et al., 2026)
The American Society of Hematology (ASH) Annual Meeting, December 12-15 in New Orleans, will include first-in-human studies of in vivo CAR-T therapy. This preview covers five of them in malignant hematology, four in B-cell lymphoma and one in multiple myeloma, chosen for development stage and scientific rationale. It describes each study’s design and the questions it addresses, and does not disclose results.
LB2501: Dual CD19/CD20 In Vivo CAR-T Therapy in B-Cell Lymphoma
Qu et al., Abstract #4169
December 13, 2026, Poster presentation
Relapse after CD19-directed CAR-T therapy can involve loss or reduced expression of the target antigen, which lets resistant B-cell populations survive. Targeting CD19 and CD20 together maintains recognition when one antigen changes. LB2501, developed by Legend Biotech, uses the TaVec platform, an engineered lentiviral vector carrying a dual-targeting CAR. A modified viral envelope and a CD3-binding component direct gene delivery to endogenous T cells. Earlier findings presented at EHA 2026 supported continued investigation in relapsed or refractory B-cell non-Hodgkin lymphoma.
The ongoing multicenter phase I trial evaluates a single intravenous LB2501 infusion without leukapheresis, ex vivo manufacturing, or lymphodepleting chemotherapy. The ASH update expands the earlier experience and includes aggressive and indolent B-cell lymphomas. Besides response durability, the trial tests whether direct vector administration achieves consistent T-cell transduction across dose levels and across patients with different immune-cell characteristics.
SNC116: In Vivo CD19 CAR-T Generation in B-Cell Lymphoma
Miao et al., Abstract #704
December 13, 2026, Oral presentation
SNC116 is a replication-incompetent lentiviral vector designed to introduce a CD19 CAR into circulating T cells after intravenous administration. Treatment relies on the patient’s existing T cells to undergo genetic modification and expansion, so vector delivery, transduction efficiency, endogenous T-cell fitness, and antigen-driven expansion all determine how many functional CAR-T cells result from a given dose. These variables matter in heavily pretreated lymphoma patients, whose circulating T cells can be reduced in number and impaired in function.
The first-in-human dose-escalation study enrolls adults with relapsed or refractory B-cell non-Hodgkin lymphoma after at least two prior lines. Patients receive a single infusion without lymphodepletion, with safety and dose-limiting toxicity as primary endpoints. Secondary assessments cover vector kinetics, CAR-T generation, expansion, and preliminary antitumor activity.
IASO208: In Vivo CD20-Directed CAR-T After Previous B-Cell-Targeted Treatment
Xu et al., Abstract #705
December 13, 2026, Oral presentation
Most approved CAR-T therapies for B-cell lymphoma target CD19, so CD20-directed cellular therapy is less developed despite the antigen’s established role in lymphoma treatment. CD20 may remain expressed after CD19-targeted therapy, which offers an alternative target after progression. IASO208 uses a third-generation replication-defective lentiviral vector with modified cocal glycoprotein pseudotyping and additional T-cell-targeting and activating components, designed to generate CD20-directed CAR-T cells in the patient.
The first-in-human phase I study evaluates IASO208 in patients with CD20-positive relapsed or refractory B-cell non-Hodgkin lymphoma, including those previously treated with monoclonal antibodies, T-cell engagers, or cellular immunotherapies. Treatment is a single intravenous infusion without leukapheresis or lymphodepleting chemotherapy. The dose-escalation design focuses on safety, CAR-T expansion, persistence, and preliminary activity.
KSV01: In Vivo CAR-T Generation With Reduced Inflammatory Activation
Zhu et al., Abstract #2350
December 12, 2026, Poster presentation
In vivo CAR-T therapy requires systemic administration of gene-delivery vectors, which can trigger inflammatory reactions before engineered T cells begin expanding. KSV01, developed by TCRX Therapeutics, uses an optimized third-generation lentiviral vector designed to transduce circulating T cells selectively and generate CD19-directed CAR-T cells. Its immuno-quiescent design aims to reduce the early cytokine response to vector administration and could allow treatment without prophylactic corticosteroids.
The first-in-human phase I trial evaluates KSV01 in relapsed or refractory diffuse large B-cell lymphoma after at least one prior line. A single intravenous infusion is given without leukapheresis, lymphodepletion, or corticosteroid premedication. Endpoints are dose-limiting toxicity, infusion-related reactions, CAR-T expansion, and preliminary antitumor activity.
QI-019A: Dual CD19/BCMA In Vivo CAR-T in Relapsed Multiple Myeloma
Xu et al., Abstract #1051
December 14, 2026, Oral presentation
QI-019A pairs BCMA with CD19, which is expressed variably across myeloma-associated cell populations. The ClariT platform uses a lentiviral vector to introduce a nanobody-derived CD19/BCMA CAR into endogenous CD3-positive T cells, with manufacturing modifications intended to reduce premature CAR expression during vector production and improve delivery efficiency.
The phase I study enrolls patients with relapsed or refractory multiple myeloma after at least two prior lines, including those previously treated with BCMA-directed therapy. QI-019A is given intravenously without lymphodepletion, and dose escalation assesses safety, in vivo CAR-T generation, expansion, and antimyeloma activity. Including BCMA-pretreated patients tests whether dual recognition retains activity after antigen-directed immunotherapy.
What Will Determine the Future of In Vivo CAR Therapy?
Every trial in this preview gives a single infusion without lymphodepleting chemotherapy, so CAR-T expansion depends on how many functional T cells a heavily pretreated patient still has and how efficiently the vector reaches them. Expansion consistency across patients, vector selectivity, and persistence of the CAR-T cells are the measures that separate these platforms from conventional products.
Site-specific integration is a further line of work. Preclinical approaches that combine targeted gene-editing particles with donor DNA templates have introduced CAR constructs at defined genomic locations, including the T-cell receptor alpha constant locus. This could reduce the uncertainty of random integration and give better control over CAR expression, but it has no clinical validation yet.


