Maciocia Lab
@maciocialab.bsky.social
Bluesky account of Maciocia Lab, University College London working on CAR-T cells
1/5 Here we have a paper looking at CAR-NK cells, trying to give them a boost with different IL-15 variants, like the IL-15/IL-15Rα heterodimer, RLI. They built anti-CD19 CAR-NKs with these different cytokine 'armorings'.
May 5, 2026 at 7:00 AM
1/5 Here we have a paper looking at CAR-NK cells, trying to give them a boost with different IL-15 variants, like the IL-15/IL-15Rα heterodimer, RLI. They built anti-CD19 CAR-NKs with these different cytokine 'armorings'.
1/7 Interesting paper out this week on pirtobrutinib for relapsed/refractory Waldenström macroglobulinaemia. This is a non-covalent BTK inhibitor, which is important for patients who've already progressed on the more common covalent BTKi drugs.
May 4, 2026 at 7:01 AM
1/7 Interesting paper out this week on pirtobrutinib for relapsed/refractory Waldenström macroglobulinaemia. This is a non-covalent BTK inhibitor, which is important for patients who've already progressed on the more common covalent BTKi drugs.
1/4 Here's a paper addressing the perennial problem of CAR immunogenicity. This group has used phage display to generate fully human B7-H3 scFVs, and then designed a CAR-T using their lead candidate, Y111.
May 4, 2026 at 7:01 AM
1/4 Here's a paper addressing the perennial problem of CAR immunogenicity. This group has used phage display to generate fully human B7-H3 scFVs, and then designed a CAR-T using their lead candidate, Y111.
CAR-PRISM out in Nat Med. Cilta-cel in 20 high-risk smoldering myeloma patients. Single infusion, no induction, no bridging. 100% MRD-neg at 10⁻⁶ by 2 months. 16/16 with >6m follow-up in CR. Zero progressions. #CART #Myeloma https://www.nature.com/articles/s41591-026-04365-y
April 22, 2026 at 9:11 AM
CAR-PRISM out in Nat Med. Cilta-cel in 20 high-risk smoldering myeloma patients. Single infusion, no induction, no bridging. 100% MRD-neg at 10⁻⁶ by 2 months. 16/16 with >6m follow-up in CR. Zero progressions. #CART #Myeloma https://www.nature.com/articles/s41591-026-04365-y
1/ TAK-007 Phase 2 just published in Blood Cancer Discovery. Takeda's off-the-shelf CAR-NK for NHL. ORR 61%, but median PFS just 2.0 months for LBCL. 73% of patients needed subsequent therapy. Takeda discontinued the program in April 2024.
April 20, 2026 at 11:26 PM
1/ TAK-007 Phase 2 just published in Blood Cancer Discovery. Takeda's off-the-shelf CAR-NK for NHL. ORR 61%, but median PFS just 2.0 months for LBCL. 73% of patients needed subsequent therapy. Takeda discontinued the program in April 2024.
April 17, 2026 at 4:02 PM
https://doi.org/10.1126/science.adz8994
1/3 Nussenzweig lab in Science. Gene-edited haematopoietic stem cells programmed to produce B lymphocytes as long-lived antibody factories against HIV, malaria and influenza. In mice. The self-amplifying bit on antigen exposure is very clever.
1/3 Nussenzweig lab in Science. Gene-edited haematopoietic stem cells programmed to produce B lymphocytes as long-lived antibody factories against HIV, malaria and influenza. In mice. The self-amplifying bit on antigen exposure is very clever.
April 17, 2026 at 10:41 AM
https://doi.org/10.1126/science.adz8994
1/3 Nussenzweig lab in Science. Gene-edited haematopoietic stem cells programmed to produce B lymphocytes as long-lived antibody factories against HIV, malaria and influenza. In mice. The self-amplifying bit on antigen exposure is very clever.
1/3 Nussenzweig lab in Science. Gene-edited haematopoietic stem cells programmed to produce B lymphocytes as long-lived antibody factories against HIV, malaria and influenza. In mice. The self-amplifying bit on antigen exposure is very clever.
1/3 UCL co-author on this (Sergio Quezada). Pollard lab Edinburgh engineered synthetic super-enhancers that only switch on in glioblastoma stem cells via SOX2/SOX9. Packed into AAV with two payloads: kill switch HSV-TK and IL-12 to wake up the immune system.
April 16, 2026 at 3:03 PM
1/3 UCL co-author on this (Sergio Quezada). Pollard lab Edinburgh engineered synthetic super-enhancers that only switch on in glioblastoma stem cells via SOX2/SOX9. Packed into AAV with two payloads: kill switch HSV-TK and IL-12 to wake up the immune system.
1/3 This one is genuinely left-field. Rust et al put a CAR on neural stem cells, not T cells. Target is integrin αvβ3, upregulated on blood vessels around stroke damage. The CAR is not there to kill anything. A molecular GPS to anchor transplanted stem cells to injured tissue.
April 16, 2026 at 3:03 PM
1/3 This one is genuinely left-field. Rust et al put a CAR on neural stem cells, not T cells. Target is integrin αvβ3, upregulated on blood vessels around stroke damage. The CAR is not there to kill anything. A molecular GPS to anchor transplanted stem cells to injured tissue.
1/4 This account is rapidly becoming a synbio one tbh 😂 Cool paper in Nature Biotech from Alexandrov at QUT with Baker lab. ML-designed binders repurposed as allosteric switches. Circularly permute a binder, fuse into an enzyme. Ligand binds, enzyme activates. Built 47.
April 16, 2026 at 9:39 AM
1/4 This account is rapidly becoming a synbio one tbh 😂 Cool paper in Nature Biotech from Alexandrov at QUT with Baker lab. ML-designed binders repurposed as allosteric switches. Circularly permute a binder, fuse into an enzyme. Ligand binds, enzyme activates. Built 47.
1/9 Two computational binder design papers worth discussing together. NVIDIA screened 467,000 designed proteins against 127 targets. Amazon/MSK designed 288,000 nanobodies against a novel DSRCT target. We tinker in this space so very relevant to us.
April 16, 2026 at 7:45 AM
1/9 Two computational binder design papers worth discussing together. NVIDIA screened 467,000 designed proteins against 127 targets. Amazon/MSK designed 288,000 nanobodies against a novel DSRCT target. We tinker in this space so very relevant to us.
1/4 New preprint from Cheung lab at MSK and Amazon Science. Computational design of nanobody binders against a DSRCT target with zero prior structural data, no existing antibodies. True de novo design, not optimisation of something that already works.
April 15, 2026 at 7:53 PM
1/4 New preprint from Cheung lab at MSK and Amazon Science. Computational design of nanobody binders against a DSRCT target with zero prior structural data, no existing antibodies. True de novo design, not optimisation of something that already works.
1/4 New preprint from Cheung lab at MSK and Amazon Science. Computational design of nanobody binders against a DSRCT target with zero prior structural data, no existing antibodies. True de novo design, not optimisation of something that already works.
April 15, 2026 at 7:53 PM
1/4 New preprint from Cheung lab at MSK and Amazon Science. Computational design of nanobody binders against a DSRCT target with zero prior structural data, no existing antibodies. True de novo design, not optimisation of something that already works.
1/5 This is a nice paper in Cell and second one of the day not really (or at all) CAR-T related. Controlling genes with magnets, specifically a synthetic switch where electromagnetic fields activate gene expression as a "remote control".
April 15, 2026 at 6:19 PM
1/5 This is a nice paper in Cell and second one of the day not really (or at all) CAR-T related. Controlling genes with magnets, specifically a synthetic switch where electromagnetic fields activate gene expression as a "remote control".
1/3 Carl June has been making this case for years: CAR-T follow-up requirements are disproportionate to actual risk. Foss-Campbell, June et al. now put registry data behind that argument in J Immunother Cancer.
April 15, 2026 at 1:49 PM
1/3 Carl June has been making this case for years: CAR-T follow-up requirements are disproportionate to actual risk. Foss-Campbell, June et al. now put registry data behind that argument in J Immunother Cancer.
1/3 Marzuttini et al in Blood: a single TCR that recognises cathepsin G through both HLA-A*24:02 and C*07:02. Dual restriction from one receptor is unusual and immediately widens the eligible patient pool for TCR-T therapy in AML.
April 15, 2026 at 1:48 PM
1/3 Marzuttini et al in Blood: a single TCR that recognises cathepsin G through both HLA-A*24:02 and C*07:02. Dual restriction from one receptor is unusual and immediately widens the eligible patient pool for TCR-T therapy in AML.
1/5 Left-field (for a CAR-T account) paper from Nature Comms where the authors engineered programmable inhibitors of CRAC (Ca2+ release-activated Ca2+) channels.
April 15, 2026 at 8:21 AM
1/5 Left-field (for a CAR-T account) paper from Nature Comms where the authors engineered programmable inhibitors of CRAC (Ca2+ release-activated Ca2+) channels.
1/7 A longer post on this one. Allogene ALPHA3 interim futility analysis is out. First randomised data for allogeneic CAR-T as first-line consolidation in LBCL. 58.3% MRD clearance with cema-cel vs 16.7% observation (n=12 per arm). Passed the (quite low) futility bar.
April 14, 2026 at 9:15 AM
1/7 A longer post on this one. Allogene ALPHA3 interim futility analysis is out. First randomised data for allogeneic CAR-T as first-line consolidation in LBCL. 58.3% MRD clearance with cema-cel vs 16.7% observation (n=12 per arm). Passed the (quite low) futility bar.
1/6 Dual CAR approach for allogeneic rejection: CD19 CAR for the tumour, CD70 CAR to kill host alloreactive T cells. Multiple companies are already pursuing CD70 as their allo strategy so this is not a new concept, but useful data.
April 12, 2026 at 2:42 PM
1/6 Dual CAR approach for allogeneic rejection: CD19 CAR for the tumour, CD70 CAR to kill host alloreactive T cells. Multiple companies are already pursuing CD70 as their allo strategy so this is not a new concept, but useful data.
1/3 The authors made NKG2D CAR-T cells to target senescent astrocytes in the SOD1 mutant ALS mouse model. These cells did what they were supposed to do, clearing senescent populations and reducing SASP markers.
April 11, 2026 at 7:00 AM
1/3 The authors made NKG2D CAR-T cells to target senescent astrocytes in the SOD1 mutant ALS mouse model. These cells did what they were supposed to do, clearing senescent populations and reducing SASP markers.
1/4 The problem with CAR-T in GBM is not just the target, but also the tumour microenvironment, T cell trafficking, antigen escape, and poor persistence.
April 11, 2026 at 7:00 AM
1/4 The problem with CAR-T in GBM is not just the target, but also the tumour microenvironment, T cell trafficking, antigen escape, and poor persistence.
1/4 The authors make a panel of CARs with different affinities to the same epitope and test them individually and in combination. Low affinity CARs cause less CRS in humanised mice but are also less effective at tumour control, which is not exactly surprising.
April 11, 2026 at 7:00 AM
1/4 The authors make a panel of CARs with different affinities to the same epitope and test them individually and in combination. Low affinity CARs cause less CRS in humanised mice but are also less effective at tumour control, which is not exactly surprising.
1/2 10kb targeted integration at TRAC in primary T cells at 20% efficiency. Impressive technical work from Theo Roth as ever. Of course the real challenge is deciding what to do with all that packaging capacity, what you put in probably matters more than how you do it.
April 10, 2026 at 8:02 PM
1/2 10kb targeted integration at TRAC in primary T cells at 20% efficiency. Impressive technical work from Theo Roth as ever. Of course the real challenge is deciding what to do with all that packaging capacity, what you put in probably matters more than how you do it.
1/3 An in vivo gene KO screen in CAR-T where they took a curated gene list and stuck them in a lung tumour NSG xenograft (immunodeficient model) then fish out the persisting T cells.
April 10, 2026 at 8:53 AM
1/3 An in vivo gene KO screen in CAR-T where they took a curated gene list and stuck them in a lung tumour NSG xenograft (immunodeficient model) then fish out the persisting T cells.
1/4 The problem with CAR-T in GBM is not just the target, but also the tumour microenvironment, T cell trafficking, antigen escape, and poor persistence.
April 10, 2026 at 8:53 AM
1/4 The problem with CAR-T in GBM is not just the target, but also the tumour microenvironment, T cell trafficking, antigen escape, and poor persistence.