Cell therapy roadmap: CD19 CAR-T → solid tumours → in vivo CAR
From a cure for some leukaemias in 2017 to the first solid-tumour CAR-T and the prospect of making CAR-T inside the body with an injection.
Blood cancers were the easy case: a clean antigen and accessible cells. Solid tumours required new antigens, TCRs, TILs, and armour. The next inflection is manufacturing: allogeneic and in vivo approaches to make cell therapy scalable and affordable.
- 1988-2016historic
Foundations
Rosenberg's TIL therapy in melanoma (1988); first-generation CARs (Eshhar, 1989); 4-1BB and CD28 costimulation (June, Sadelain, Campana); Emily Whitehead treated with CTL019 (2012).
- 2017-2022historic
CD19 and BCMA CAR-T approved
Tisagenlecleucel and axicabtagene (2017), then five more products; ZUMA-7 puts CAR-T ahead of transplant in second-line lymphoma; ide-cel and cilta-cel in myeloma; blinatumomab and teclistamab establish off-the-shelf T-cell engagers as competitors.
- 2024-2026current
Solid tumours: TIL, TCR-T, CAR-T
Solid-tumour cell therapy arrives: lifileucel (melanoma TIL, 2024), afamitresgene autoleucel (MAGE-A4 TCR-T, synovial sarcoma, 2024; age ≥12 in 2026), satri-cel (CLDN18.2 CAR-T, gastric, China 2025), GD2 CAR-T in neuroblastoma and glioma, GPC3 CAR-T in HCC. CAR-T expands to autoimmune disease.
- 2025-2028emerging
Manufacturing revolution: allogeneic and in vivo
Gene-edited allogeneic products (cema-cel, ALLO-316) show feasibility with persistence limits. In vivo CAR-T via targeted LNPs (Capstan/AbbVie, Orna) and lentivirus (Umoja, Interius) enter first-in-human trials with B-cell depletion and no lymphodepletion. iPSC-derived CAR-NK. CAR-T cost and access become policy issues.
- 2028+speculative20%–45%likely
Speculative
Logic-gated and armoured CARs for solid tumours; neoantigen-specific TCR-T at scale via AI-predicted TCRs; in vivo CAR as an outpatient injection for lymphoma and autoimmune disease; CAR-T against fibroblast and myeloid targets; regional delivery for glioma, mesothelioma, and pancreatic cancer.
Probability ranges are named estimates that the claim is borne out on roughly a five-year horizon. They are meant to be argued with: propose a revision with your name and reasoning via a pull request to src/data/confidence.ts.
Story
topFoundations
Rosenberg's TIL therapy in melanoma (1988); first-generation CARs (Eshhar, 1989); 4-1BB and CD28 costimulation (June, Sadelain, Campana); Emily Whitehead treated with CTL019 (2012).
CD19 and BCMA CAR-T approved
Tisagenlecleucel and axicabtagene (2017), then five more products; ZUMA-7 puts CAR-T ahead of transplant in second-line lymphoma; ide-cel and cilta-cel in myeloma; blinatumomab and teclistamab establish off-the-shelf T-cell engagers as competitors.
A CD19 CAR-T that cures about 40% of patients with large B-cell lymphoma who had failed everything, and beat transplant in second line.
Ciltacabtagene autoleucel is a one-time BCMA CAR-T for myeloma that, in CARTITUDE-4, cut the risk of death by about 45% compared with standard regimens.
Blinatumomab was the first T-cell engager (2014), and is now given to children and adults with leukaemia even when in remission, because it improves survival.
Teclistamab was the first off-the-shelf bispecific for multiple myeloma, and is now approved after just one prior line of therapy.
Solid tumours: TIL, TCR-T, CAR-T
Solid-tumour cell therapy arrives: lifileucel (melanoma TIL, 2024), afamitresgene autoleucel (MAGE-A4 TCR-T, synovial sarcoma, 2024; age ≥12 in 2026), satri-cel (CLDN18.2 CAR-T, gastric, China 2025), GD2 CAR-T in neuroblastoma and glioma, GPC3 CAR-T in HCC. CAR-T expands to autoimmune disease.
Lifileucel was the first approved TIL therapy: the patient's own tumour-fighting immune cells are expanded to billions and given back.
Afamitresgene autoleucel was the first engineered T-cell receptor therapy approved for a solid tumour, synovial sarcoma.
Satricabtagene autoleucel (satri-cel) is the first CAR-T therapy approved for a solid tumour (gastric cancer), in China.
A tight-junction protein normally hidden in the stomach lining that becomes exposed in gastric and pancreatic cancers.
Glypican-3 is a fetal liver protein that reappears in liver cancer, giving CAR-T and bispecifics a clean target.
A protein normally only in testis that some sarcomas and other tumours switch on; T cells can be engineered to recognise fragments of it.
Manufacturing revolution: allogeneic and in vivo
Gene-edited allogeneic products (cema-cel, ALLO-316) show feasibility with persistence limits. In vivo CAR-T via targeted LNPs (Capstan/AbbVie, Orna) and lentivirus (Umoja, Interius) enter first-in-human trials with B-cell depletion and no lymphodepletion. iPSC-derived CAR-NK. CAR-T cost and access become policy issues.
Instead of engineering T cells in a factory, an injection reprograms them inside the patient's body.
Cell therapies made from healthy donors in advance, so patients do not have to wait for their own cells to be engineered.
Putting the cancer-seeking receptor on natural killer cells or macrophages instead of T cells, which could be safer and off-the-shelf.
AbbVie owns Venclexta and Elahere, and bought Capstan, the leading in vivo CAR-T company.
Speculative
Logic-gated and armoured CARs for solid tumours; neoantigen-specific TCR-T at scale via AI-predicted TCRs; in vivo CAR as an outpatient injection for lymphoma and autoimmune disease; CAR-T against fibroblast and myeloid targets; regional delivery for glioma, mesothelioma, and pancreatic cancer.
Upgraded CAR-T cells that also secrete immune boosters, resist exhaustion, or only fire when two signals are present.
T cells engineered with a receptor that sees fragments of proteins inside the cancer cell, reaching targets CAR-T cannot.
FAP is a protein on the scaffolding cells that surround many tumours, so it lights up almost any solid cancer on a PET scan.