Browse
Log in / Register
Share
May 8, 2026 — A comprehensive review in Bioengineering & Translational Medicine is refocusing the cell therapy field on an overlooked immune cell: the macrophage. Engineered to carry chimeric antigen receptors, these so-called CAR-M cells are being positioned as a viable alternative to CAR-T for solid tumors, where T-cell therapies have repeatedly disappointed. Early human trials, the review finds, have posted reassuring safety results and preliminary signs of efficacy.
Chimeric antigen receptor T cells have transformed the treatment of blood cancers, driving durable remissions in leukemias, lymphomas, and multiple myeloma. But when pointed at solid tumors, the platform has struggled. Three obstacles stand in the way: a dense extracellular matrix that physically shields the tumor, immunosuppressive cytokines that fatigue arriving T cells, and antigen heterogeneity that permits escape from single-target recognition.
Macrophages are a different story. These innate immune cells are already present in high numbers within solid tumors—often recruited by the tumor itself to promote growth. The review lays out how CAR-M flips that relationship, converting a tumor's allies into its attackers.
The concept is not new. In 2006, researchers transduced human monocytes with a chimeric receptor targeting carcinoembryonic antigen and demonstrated tumor suppression in mice. What is new is the pace of progress. Over the past five years, the field has expanded from proof-of-concept studies into a pipeline of clinical candidates—including MT-101, SY001, and CT-0508—with favorable early safety data across multiple studies.
CAR-M construction follows familiar CAR-T principles: an extracellular single-chain variable fragment that binds antigen without MHC restriction, a hinge for flexibility, a transmembrane anchor, and intracellular signaling domains. First-generation CARs rely on a single activation domain such as CD3ζ, FcRγ, or Megf10, driving phagocytosis. Second-generation designs pair those with costimulatory molecules—4-1BB, CD28, or DAP10—or with cytokine-inducing modules like IFN-γ, CD147, and TLR4 to boost inflammatory output. The emerging third generation uses chimeric cytokine receptors that fuse IL-10 or TGF-β receptor sensing domains to IFN-γ receptor signaling, flipping immunosuppressive environmental cues into antitumor activation.
Producing CAR-M at scale remains difficult. Laboratory workhorses like RAW264.7 and THP-1 cell lines are practical for validation, but clinical manufacturing depends on autologous peripheral blood mononuclear cells. The standard workflow—G-CSF mobilization, CD14+ monocyte isolation, GM-CSF-driven differentiation, and adenoviral CAR delivery—takes roughly nine to thirteen days. CT-0508 manufacturing data show a mean viability of 86.39 percent and transduction efficiency of 79.28 percent. Alternatives in development include induced pluripotent stem cells and hematopoietic stem/progenitor cells, which could provide off-the-shelf scalability, as well as biomaterial-based platforms that edit macrophages directly at disease sites in animal models.
Macrophages resist viral transduction, a barrier that nearly killed the field in its infancy. Lentiviral systems carrying the HIV-2 protein Vpx now achieve about 70 percent transduction in monocyte-derived macrophages, and adenoviral Ad5/F35 vectors reach up to 80 percent in clinical settings. Paradoxically, viral transduction activates inflammasomes, which enhances the proinflammatory and phagocytic profile of the final product. Nonviral delivery is accelerating: lipid nanoparticles—the workhorse of mRNA vaccines—can deliver CAR mRNA to tumor-associated macrophages directly in vivo, eliminating ex vivo manipulation. Other platforms, including fluorinated ionizable lipids, metal-organic frameworks, engineered small extracellular vesicles, erythrocyte carriers, and enucleated mesenchymal stem cells that home to glioblastoma, have all produced functional CAR-M in living tissue.
CAR-M bring a more diverse antitumor toolkit than CAR-T. They clear targets through antigen-specific phagocytosis, release reactive oxygen and nitrogen species, and secrete IL-6 and TNF-α in a self-amplifying loop. They also act as professional antigen-presenting cells, cross-priming CD8+ cytotoxic and CD4+ helper T cells for durable systemic immunity. And they tear down the physical barriers CAR-T cannot: in pancreatic cancer models, CAR-M engulf fibroblast activation protein-positive stromal cells and reduce collagen; in breast cancer models, HER2-triggered CD147 signaling upregulates matrix metalloproteinases that degrade the extracellular matrix, improving T-cell infiltration and drug penetration.
The review counts twelve CAR-M studies worldwide, most targeting HER2- or mesothelin-expressing solid tumors. The fact that more than a dozen studies have reached the clinic within roughly five years of the concept's revival underscores a field moving at unusual speed.
CT-0508, a HER2-directed product, provided the first human phase 1 readout. Among fourteen patients with HER2-overexpressing cancers, there were no grade 3 or 4 cytokine release syndrome events and no neurotoxicity, and the trial did not require lymphodepleting chemotherapy. In thirteen radiographically evaluable patients, 40 percent saw reductions in target lesions. Biopsy data showed CAR-M present in 92 percent of on-treatment samples and 27 percent at week four, with single-cell sequencing revealing remodeling of the tumor microenvironment toward a more adaptive immune state. SY001, targeting mesothelin, reported no grade 3 or higher adverse events in two ovarian cancer patients, both of whom maintained stable disease at day 28.
CAR-M technology has moved into other disease areas. FAP-targeted CAR-M have reduced cardiac fibrosis after myocardial infarction by engulfing activated fibroblasts, with related designs showing benefit in liver, lung, and kidney fibrosis. Aβ-targeted CAR-M engineered to secrete M-CSF for self-maintenance have cut amyloid plaque burden in Alzheimer's disease models. CD47-directed CAR-M with reactive oxygen species-responsive nanoparticles have boosted cholesterol efflux in atherosclerosis. In infectious disease, SasA-targeted CAR-M generated by implantable nanoparticle coatings have reduced Staphylococcus aureus infection, and CR3022-based CAR-M have engulfed SARS-CoV-2 particles in vitro. Anti-TNF CAR-M carrying an IL-4 signaling switch have shown efficacy in inflammatory liver and kidney injury.
The field's central anxiety is runaway activation—cytokine release syndrome and off-tumor toxicity in organs like the liver, lung, and kidney. Synthetic biology offers several controls: logic-gated CARs (AND, OR, NOT) restrict killing to cells that express multiple antigens or spare healthy tissue carrying protective markers; switchable universal CARs redirect specificity on demand with soluble adapters; hypoxia- and lactate-responsive circuits keep cells silent in normal tissue. Inducible-caspase-9 suicide switches provide pharmacologic elimination; one biodistribution study showed AP1903 could control CAR-M persistence even after sixty days.
The review's authors propose a "4S framework"—specificity, switchability, synergy, and safety—for next-generation development. Combinations look especially attractive: chemotherapy creates immunogenic cell death that feeds macrophage phagocytosis; checkpoint blockade protects the T-cell compartment; CD47 blockade lifts the "don't eat me" brake on engulfment; and CAR-M plus CAR-T coadministration could create a positive feedback loop of mutual amplification.
Current evidence is limited to observational and phase 1 studies, and the field must still validate efficacy endpoints, scale manufacturing, and standardize response criteria. Yet with early safety signals, a mechanism built to overcome the very barriers that defeated CAR-T in solid tumors, and applications reaching beyond cancer into fibrosis and neurodegeneration, CAR-M is emerging as one of the most serious frontiers in cellular immunotherapy.
