CAR-T therapy

一行要約

患者 (または donor) T 細胞を ex vivo で取り出し、scFv-hinge-transmembrane-costimulatory domain-CD3ζ から構成される chimeric antigen receptor (CAR) を retroviral / lentiviral vector で導入後、lymphodepletion conditioning を経て再注入する adoptive cellular therapy。CD19 CAR-T が血液腫瘍で paradigm を確立し、Grupp et al. NEnglJMed 2013 / Maude et al. NEnglJMed 2014 が小児 B-ALL での持続的 CR を実証、DLBCL では ZUMA-1 (Neelapu et al. NEnglJMed 2017) が ORR 82% / CR 54% を達成した。包括的総説は June et al. NEnglJMed 2018 / June et al. Science 2018 / Larson et al. NatRevCancer 2021 を参照。固形腫瘍では mesothelin (mesothelioma / NSCLC、Adusumilli et al. SciTranslMed 2014) / DLL3 (SCLC) / HER2 / GPC3 / claudin18.2 等を標的とした CAR-T が開発中だが承認製品はなく、antigen heterogeneity・TME 抑制・tumor trafficking の 3 大障壁克服が課題である。

メンバー比較表

薬剤	製品名	Target	Costim	適応	備考
Tisagenlecleucel	Kymriah	CD19	4-1BB	B-ALL (小児/AYA), DLBCL	Lentiviral, Novartis
Axicabtagene ciloleucel	Yescarta	CD19	CD28	DLBCL, FL	Retroviral, Kite/Gilead
Brexucabtagene autoleucel	Tecartus	CD19	CD28	MCL, B-ALL	T cell enrichment process
Lisocabtagene maraleucel	Breyanzi	CD19	4-1BB	DLBCL, CLL	Defined CD4:CD8 ratio
Idecabtagene vicleucel	Abecma	BCMA	4-1BB	MM	1st BCMA CAR-T
Ciltacabtagene autoleucel	Carvykti	BCMA	4-1BB	MM	Bi-epitope VHH, J&J/Legend

主要エビデンス

血液腫瘍 pivotal (CD19 CAR-T)

CD19 CAR-T は CAR-T 療法全体の paradigm を確立した landmark 領域である:

• B-ALL: Grupp et al. NEnglJMed 2013 が CTL019 (4-1BB costim、後の Tisagenlecleucel / Kymriah) を 2 例の小児 r/r B-ALL に投与し、初の臨床的 CAR-T CR を報告。Maude et al. NEnglJMed 2014 が 30 例 cohort で CR 90% (27/30) の unprecedented efficacy を示し、B-cell aplasia を managed toxicity として定義。Park et al. JClinOncol 2015 が MSKCC の 19-28z CAR-T (CD28 costim) で成人 ALL CR 83% を報告し、costimulatory domain の違いによる efficacy / toxicity profile の差が明らかになった
• DLBCL: ZUMA-1 (Neelapu et al. NEnglJMed 2017) が CD28 costim Axicabtagene ciloleucel で r/r DLBCL ORR 82% / CR 54% を達成し、2017 年 FDA 承認。JULIET trial (Tisagenlecleucel、4-1BB costim) は ORR 52% / CR 40% と lower response rate だが late relapse が少ない傾向 → costimulatory domain が CAR-T kinetics を規定 (Zhao et al. CancerCell 2015)
• 長期 follow-up: Cappell et al. JClinOncol 2020 が NCI の long-term data で、CD19 CAR-T 後の持続寛解が 10 年超に及ぶ例を報告。Elsallab et al. LancetOncol 2020 が承認前後の real-world performance を systematic review で整理

BCMA CAR-T (Multiple Myeloma)

BCMA (B-cell maturation antigen) CAR-T は myeloma 領域で 2 製品が承認:

• Idecabtagene vicleucel (Abecma) : r/r MM (4+ prior lines) で ORR 73% / CR 33%。KarMMa 試験ベース
• Ciltacabtagene autoleucel (Carvykti) : VHH-based bi-epitope BCMA binder で ORR 97% / stringent CR 67% (CARTITUDE-1)。早期 line (1-3 prior) への展開 (CARTITUDE-4) で PFS HR 0.26 を報告 → MM 治療 landscape を変革中
• 総説は Mikkilineni et al. NatRevClinOncol 2021 を参照

固形腫瘍 CAR-T (early-phase)

Mesothelin CAR-T (NSCLC / Mesothelioma)

• Adusumilli et al. SciTranslMed 2014 が intrapleural mesothelin CAR-T の preclinical model で systemic i.v. 投与を凌駕する efficacy と CD4-dependent long-lasting immunity を報告。Regional delivery が fixed TME barriers を bypass する戦略として確立
• Morello et al. CancerDiscov 2016 が mesothelin-targeted CAR の最適化 (scFv affinity tuning / costimulatory domain selection / tonic signaling 回避) を体系的に整理
• Beatty et al. CancerImmunolRes 2014 が mRNA electroporation による transient CAR expression (安全性向上 vs persistence trade-off) で mesothelioma / pancreatic cancer に activity を示した

DLL3 CAR-T (SCLC)

DLL3 は SCLC に高発現する Notch ligand で、Saunders et al. SciTranslMed 2015 が DLL3 を therapeutic target として同定し ADC (Rova-T) を開発。Rova-T は clinical failure に終わったが、DLL3 CAR-T (AMG 119 / AB-2100) および DLL3-targeted BiTE (AMG 757 / tarlatamab、Giffin et al. ClinCancerRes 2021) が SCLC の次世代免疫療法として開発中。DLL3 CAR-T は tumor-restricted expression (正常組織は低発現) で on-target/off-tumor toxicity が低い利点を持つ

HER2 CAR-T

• Ahmed et al. JClinOncol 2015 が HER2+ sarcoma で安全性・activity を報告
• Morgan et al. MolTher 2010 が HER2 CAR-T (高親和性 scFv + 高用量投与) で致死的 on-target/off-tumor SAE を記録。肺上皮 HER2 低発現への cross-reactivity → scFv affinity tuning / dose escalation protocol の重要性を確立した教訓的事例

GBM CAR-T

• Brown et al. NEnglJMed 2016 が IL13Rα2 CAR-T の intracavitary 投与で GBM の劇的 regression を報告 (N=1 case report) → regional delivery + brain tumor の proof-of-concept
• Sampson et al. ClinCancerRes 2014 が EGFRvIII CAR-T の phase I で safety を確認

安全性: CRS / ICANS

• CRS (Cytokine Release Syndrome) : CAR-T 活性化 → IL-6 / IFN-γ / TNF-α massive release → 発熱 / 低血圧 / 低酸素 / 多臓器不全。Brudno et al. Blood 2016 が CAR-T 毒性を包括的に整理。Teachey et al. CancerDiscov 2016 が ALL cohort で CRS 予測 biomarker (baseline disease burden / IL-6 / IFN-γ kinetics) を構築
• CRS 管理: Grade 2+ で tocilizumab (anti-IL-6R mAb) 投与が標準。Corticosteroid は CAR-T persistence 抑制のリスクあり → grade 3+ に限定
• ICANS (Immune Effector Cell-Associated Neurotoxicity Syndrome) : BBB breakdown → CNS cytokine 流入 → cerebral edema / aphasia / seizure。CD28 costim (Yescarta) で 4-1BB costim (Kymriah) より頻度高い傾向

耐性機構

• Antigen escape: CD19 epitope loss (Sotillo et al. CancerDiscov 2015) が CAR-T 最大の achilles heel。Alternative splicing でエクソン 2 スキップ → truncated CD19 が surface に発現するが scFv 認識エピトープを欠失。Antigen loss frequency は DLBCL 約30%、B-ALL 約15%
• Tandem / Dual CAR: Hegde et al. JClinInvest 2016 が tandem CAR (single construct に 2 scFv)、Zah et al. CancerImmunolRes 2016 が CD19/CD20 bispecific CAR で antigen escape 回避を実証。Ruella et al. JClinInvest 2016 が CD19 + CD123 dual targeting で antigen-loss relapse prevention を示した
• T cell exhaustion: CAR tonic signaling が exhaustion を誘導 (PD-1 / TIM-3 / LAG-3 upregulation)。Long et al. NatMed 2015 が 4-1BB costim domain が tonic signaling-induced exhaustion を ameliorate することを示した
• TME-mediated suppression: 固形腫瘍の TME (MDSC / Treg / hypoxia / metabolic stress) が CAR-T 機能を抑制。Moon et al. ClinCancerRes 2014 が multifactorial T cell hypofunction を整理。Cherkassky et al. JClinInvest 2016 が cell-intrinsic PD-1 blockade (dominant-negative PD-1 co-expression) で TME 抑制を克服する戦略を提示

メカニズム

CAR 構造の進化

CAR は extracellular antigen recognition domain (scFv) + hinge + transmembrane domain + intracellular signaling domain で構成される。設計原理の definitive review は Lim et al. Cell 2017 が提供:

• 1st-gen: scFv-CD3ζ のみ。MHC-independent な antigen recognition を実現したが、costimulation 不在で T cell anergy → persistence 短・activity 弱。Hwu et al. JExpMed 1993 が初期 concept を実証
• 2nd-gen (臨床標準) : + CD28 or 4-1BB costimulatory domain。CD28 costim は rapid effector expansion / higher peak / shorter persistence、4-1BB costim は slower kinetics / longer persistence / less exhaustion (Zhao et al. CancerCell 2015)。Kymriah = 4-1BB / Yescarta = CD28
• 3rd-gen: CD28 + 4-1BB dual costim → 理論的には expansion + persistence 両立だが、clinical benefit の明確な evidence はまだ限定的
• 4th-gen (TRUCK / Armored CAR) : CAR + transgenic cytokine secretion (IL-12 / IL-15 / IL-21) で TME を remodel。Pegram et al. Blood 2012 が IL-12-secreting CAR-T の preclinical proof-of-concept を示した。Zuo et al. CancerCell 2026 が IL-36γ armored CAR-T による neutrophil reprogramming と endogenous immunity 誘導を報告
• Next-gen / logic-gated: SynNotch (sequential antigen recognition) / SUPRA CAR (universal adapter) / Kloss et al. NatBiotechnol 2013 の combinatorial antigen recognition → healthy tissue sparing + tumor selectivity

固形腫瘍 CAR-T の 5 大障壁

Jackson et al. NatRevClinOncol 2016 / Khalil et al. NatRevClinOncol 2016 が整理した固形腫瘍 CAR-T の根本的課題:

1. Antigen escape / heterogeneity: 固形腫瘍は target antigen の heterogeneous expression → single antigen loss で relapse。TIL-therapy の polyclonal TCR repertoire と対照的
2. T cell exhaustion in TME: Chronic antigen stimulation + TME 由来の inhibitory signal (PD-L1 / TGF-β / adenosine / IDO) → PD-1+Tim-3+Lag-3+ exhausted phenotype への分化 (Wherry et al. NatImmunol 2011)
3. Tumor trafficking / infiltration: 固形腫瘍への物理的アクセス不足 → chemokine receptor engineering / regional delivery (Adusumilli et al. SciTranslMed 2014) / Caruana et al. NatMed 2015 が heparanase 発現で ECM penetration を改善
4. On-target/off-tumor toxicity: Target antigen の正常組織発現 → lethal cross-reactivity (Morgan et al. MolTher 2010)。Safety switch (iCasp9: DiStasi et al. NEnglJMed 2011) で対処
5. Immunosuppressive TME: MDSC / Treg / hypoxia / metabolic competition → engineered resistance (PD-1 DN / TGFβ trap / CAR-NK) で克服戦略

Allogeneic / iPSC-derived / in vivo CAR

Autologous CAR-T の製造期間 (3-4 週間) / コスト / vein-to-vein time の限界を克服する次世代 platform:

• Allogeneic CAR-T: Healthy donor T cell に CAR 導入 + TCR / HLA knockout (CRISPR-Cas9) で GvHD 回避。Stadtmauer et al. Science 2020 が CRISPR-engineered T cell の phase 1 safety を報告
• iPSC-derived CAR-T: Themeli et al. CellStemCell 2015 が iPSC → T cell 分化 → CAR 導入の pipeline を提示。Off-the-shelf / unlimited cell supply の potential
• CAR-NK: NK cell に CAR 導入 → GvHD risk なし / CRS 低減 / innate anti-tumor activity 保持
• In vivo CAR (LNP delivery) : T cell を体外操作せず、LNP に CAR mRNA / DNA を封入して in vivo で T cell を transduce する革新的 approach。Manufacturing-free の ultimate off-the-shelf 戦略

CAR-T と endogenous immunity の連携

CAR-T は直接殺細胞だけでなく endogenous anti-tumor immunity を boost する:

• Johnson et al. Cell 2021 が immunostimulatory RNA RN7SL1 を搭載した CAR-T が endogenous T cell / DC を活性化し、CAR target 以外の antigen にも immunity を誘導 (epitope spreading)
• Hirschhorn et al. Cell 2023 が CAR-T による neutrophil engagement → antigen escape variant の排除を示した

臨床位置づけ

• B-ALL: CD19 CAR-T (Kymriah / Tecartus) が r/r B-ALL の standard salvage。CR 80-90% だが、long-term relapse (antigen escape + bone marrow relapse) が 40-50% → post-CAR alloSCT consolidation の議論
• DLBCL / aggressive NHL: Yescarta / Breyanzi / Kymriah が 3L+ standard。2L front-loading (ZUMA-7 / TRANSFORM / BELINDA) で Yescarta / Breyanzi が 2L standard 化
• MM: Carvykti が 1-3 prior lines (CARTITUDE-4) で lenalidomide-refractory MM の standard 候補に。Early line への展開が進行中
• 固形腫瘍 (NSCLC / SCLC / mesothelioma) : 承認製品なし。Mesothelin CAR-T (intrapleural delivery) / DLL3 CAR-T (SCLC) が最も advanced だが phase I/II 段階。抗原選択 / delivery 最適化 / TME 克服が breakthrough 条件
• Safety management: CRS / ICANS の grading system 標準化 (ASTCT consensus)。Tocilizumab 常備 / ICU backup が投与施設要件。Outpatient CAR-T (low CRS risk product + step-up dosing) は将来的展望
• Cost / Access: ~$400K-500K/dose(製品価格のみ)。入院費・supportivecare含め$1M+ → payer negotiation / outcome-based pricing / allogeneic off-the-shelf で cost reduction を目指す

Open Questions

• 固形腫瘍 CAR-T の breakthrough: Antigen heterogeneity / TME suppression / trafficking の 3 大障壁同時解決。Multi-antigen targeting + armored cytokine secretion + regional delivery の統合 approach が必要。NSCLC mesothelin CAR-T / SCLC DLL3 CAR-T の pivotal trial 到達が指標
• Off-the-shelf allogeneic / iPSC-derived CAR-T: CRISPR-edited allogeneic (TCR KO / HLA KO / PD-1 KO) の persistence / GvHD balance。iPSC-derived CAR-T の scalability と functional maturity。CAR-NK の固形腫瘍 efficacy data 蓄積
• In vivo CAR generation: LNP-mediated in vivo CAR-T transduction の phase 1 safety・efficacy data 待ち。Manufacturing-free paradigm が実現すれば cost / access 革命
• CRS / ICANS の outpatient 管理: Step-up dosing / split dosing / prophylactic tocilizumab / reduced lymphodepletion で入院期間短縮。Risk-stratified management algorithm の standardization
• Logic-gated CAR (固形腫瘍 safety) : SynNotch / AND-gate / NOT-gate CAR の clinical translation。Normal tissue sparing と tumor selectivity の両立
• Post-CAR-T relapse strategy: Antigen escape 後の rescue — bispecific antibody / alternative antigen CAR-T / TIL therapy への sequencing
• T cell fitness optimization: Manufacturing 段階での T cell phenotype optimization (less-differentiated / stem cell memory enrichment) / metabolic reprogramming (Hirabayashi et al. NatCancer 2021)
• CAR-T + IO 併用: CAR-T + anti-PD-1 (TME exhaustion 解除) / CAR-T + anti-CTLA-4 (priming enhancement) の safety / efficacy data

重要論文 Top 10

1. ★★★★★ Grupp et al. NEnglJMed 2013 — CD19 CAR-T 臨床有効性確立、CAR-T paradigm 起点
2. ★★★★★ Maude et al. NEnglJMed 2014 — B-ALL CR 90% — 持続寛解で細胞療法標準化牽引
3. ★★★★★ Neelapu et al. NEnglJMed 2017 — ZUMA-1 — Yescarta DLBCL 承認、CD28 costim
4. ★★★★ Adusumilli et al. SciTranslMed 2014 — 固形腫瘍 mesothelin CAR-T regional delivery 基盤
5. ★★★★ Sotillo et al. CancerDiscov 2015 — CD19 antigen escape 機構解明 — 耐性理解の基盤

関連エンティティ

• 関連 class: BiTE-bispecific (off-the-shelf T cell engager 代替), TIL-therapy, ADC (antigen targeting alternative)
• target: mesothelin / CD19 / BCMA / MAGE-A4 / DLL3 / GPC3
• ドメイン: novel-cancer-modalities

関連論文

直近 30 件のみ表示中 (関連 321 件)。 全件は MOC ページ または Obsidian (source Markdown) から辿れます。

Basic-Brain metastasis

• Mavrikios et al. AnnOncol 2026 — Lung cancer brain metastases management at the dawn of personalized medicine are we ready to break the barriers

Basic-Extracellular vesicles

• Ji et al. CancerCell 2026 — T cell extracellular vesicles go nuclear to turn cold tumors hot
• De et al. JExtracellVesicles 2026 — Characterizing the fate of anti-CS1 nanobody displaying extracellular vesicles in multiple myeloma

Basic-Immunotherapy

• Wang et al. Cell 2026 — CLIM-TIME identifies metastatic microenvironment modulators for T cell therapy response
• Yue et al. Cell 2026 — Expansion and CAR engineering of granulocyte-monocyte progenitors for cellular immunotherapy
• Zhang et al. Cell 2026 — A convergent uPAR-positive tumor ecosystem creates broad vulnerability to CAR t cell therapy
• Jung et al. FrontImmunol 2026 — ATG5-mediated inducible autophagy sustains CAR-T cell durability under solid tumor stress
• Mao et al. FrontImmunol 2026 — MicroRNA-targeted reprogramming of CD8+ T cells against cancer

Basic-Neutrophil and myeloid cell

• Zuo et al. CancerCell 2026 — IL-36gamma armored CAR T cells reprogram neutrophils to induce endogenous antitumor immunity
• Hao et al. CellRepMed 2026 — Non-genetically reprogrammed meta-neutrophils potentiate chemo-immunotherapy against lung metastatic triple-negative
• Nemilostiva et al. ClinTranslOncol 2026 — CAR-engineered neutrophils derived from induced pluripotent stem cells a new frontier in cellular immunotherapy
• McGarrity-Cottrell et al. FrontImmunol 2026 — Physiological, patho-physiological, and potential therapeutic roles for neutrophils in cancer & beyond

Basic-Others

• Tang et al. CancerCell 2026 — Mapping cell type-resolved transcriptomic profiles to patient survival in pancreatic cancer
• Hanahan et al. Cell 2026 — Hallmarks of cancer-then and now, and beyond
• Hinterleitner et al. Cell 2026 — Senescence in cancer hallmarks, paradoxes, and therapeutic promise

Basic-Tumor microenvironment

• Subudhi et al. Cell 2026 — Targeting angiogenesis lessons from 25 years of normalizing tumor vasculature

Clinical-NSCLC-Driver negative

• Li et al. FrontImmunol 2026 — The application and challenges of immune checkpoint inhibitors in lung cancer therapy

Clinical-NSCLC-EGFR

• Niu et al. FrontImmunol 2026 — Efficacy and mechanisms of immune checkpoint inhibitors in late-stage EGFR-mutated non-small cell lung cancer following

Clinical-Novel modalities

• Kubo et al. BioProtocol 2026 — Cite as kubo, N. et al. (2026). A feeder cell-free system for chimeric antigen receptor gene transduction into
• Chen et al. CancerCell 2026 — Overcoming functional and translational challenges of cellular immunotherapies for solid tumors
• Amor et al. Cell 2026 — CAR T targets AI takes the wheel
• Baker et al. Cell 2026 — AI-driven discovery of GPNMB CAR T cells as a multi-cancer therapy
• Freeburg et al. Cell 2026 — The critical role of the endogenous immune compartment after CAR T cell therapy in recurrent GBM
• Bajgain et al. CellRepMed 2026 — Engineering functionality-optimized fully human B7-H3 CAR T cells for enhanced solid tumor therapy
• Pham et al. FrontImmunol 2026 — Adapter-based allogeneic CAR t cells to overcome antigen escape in solid tumors
• Zhan et al. FrontImmunol 2026 — Ionizing radiation enhances CAR T-cell infiltration and efficacy in solid tumors
• Qi et al. JClinOncol 2026 — Guanylyl cyclase 2c-targeted chimeric antigen receptor T-cell therapy in patients with metastatic colorectal cancer
• Abrahimi et al. JExpMed 2026 — Intravesical mesothelin-based CAR T cells targeting MUC16 effectively control bladder cancer in preclinical models

Clinical-SCLC

• Guan et al. CritRevOncolHematol 2026 — Overcoming immunotherapy resistance in extensive-stage small cell lung cancer insights into mechanisms and therapeutic

Clinical-Supportive care

• Prescott et al. IntensiveCareMed 2026 — Surviving sepsis campaign international guidelines for management of sepsis and septic shock 2026

合成 (Outputs)

• antigen-presenting-tan-endogenous-t-cell-response
• car-cell-tme-resilience-engineering
• hpcs-plasticity-hub-preemptive-targeting-translation
• kras-g12c-crispr-modality-synthesis
• novel-modality-sequencing-driver-resistance