A narrative review of barrier-matched engineering of macrophages for solid tumor immunotherapy: design logic, translational barriers and clinical outlook.

Background and Objective: Macrophages are being investigated as adoptive cell therapies for solid tumors because they can migrate into tissues, phagocytose tumor cells, present antigen, secrete cytokines, and interact with stromal and other myeloid cells. However, their plasticity is also a liability. Once transferred into solid tumors, macrophages are exposed to hypoxia, metabolic stress, suppressive cytokines, and tumor-derived signals that may reduce phagocytosis, dampen inflammatory activation, and redirect them toward tumor-associated macrophage (TAM)-like, tumor-supportive phenotypes. This review examines engineered macrophage therapy through a barrier-matched framework. Methods: PubMed, Web of Science, Scopus, and ClinicalTrials.gov were initially searched from database inception to 24 December 2024, with a supplementary search update conducted through 21 July 2026. Search terms covered engineered macrophages, chimeric antigen receptor macrophages (CAR-M), solid tumors, the tumor microenvironment, phagocytosis checkpoints, macrophage backpacks, nanoparticle macrophages, and stem-cell-derived macrophages. Peer-reviewed original studies, clinical trial reports, high-quality reviews, and registered trials directly relevant to macrophage- or myeloid-cell-based cancer immunotherapy were selected. No formal risk-of-bias assessment or meta-analysis was performed. Key Content and Findings: Solid tumors impose five interacting barrier layers on macrophage therapy: physical barriers, metabolic barriers, immunosuppressive barriers, anti-phagocytic barriers, and phenotypic-plasticity barriers. These include restricted infiltration through dense extracellular matrix and abnormal vasculature, hypoxia, lactate accumulation, extracellular adenosine (eAdo)-A2A receptor (A2AR) signaling, suppressive tumor microenvironmental cues, do-not-eat-me checkpoints such as CD47-SIRPα, and TAM-like reprogramming after transfer. Engineering strategies have therefore shifted from simple cell transfer toward barrier-matched design. Some approaches add tumor-directed recognition, as in CAR-M or ligand-modified macrophages. Others aim to preserve inflammatory or phagocytic activity, deliver local payloads, attenuate inhibitory signaling, define scalable cell sources, or improve release and safety testing. The evidence base remains uneven. CAR-M has entered early clinical testing, whereas many nanoparticle-armed, backpack-based, stimulus-responsive, and genome-edited systems remain largely preclinical. Conclusions: Engineered macrophage therapy should be guided by the dominant barrier layer in a given solid tumor setting rather than by the number of modifications added to the cell product. Future development requires clearer links between engineering modules, barrier-adapted translational readouts, manufacturing feasibility, persistence, safety, and rational trial design.