South Korea's KAIST Develops an Injection That Reprograms a Tumour's Own Immune Cells to Destroy Cancer

Researchers at the Korea Advanced Institute of Science and Technology (KAIST) in South Korea have developed a new cancer treatment strategy that injects a specially engineered drug directly into tumours and instructs the tumour's own immune cells to become cancer-targeting killers from within. The research, published in ACS Nano in November 2025, offers a potential solution to one of the most persistent barriers in solid tumour immunotherapy: the inability of immune therapies to penetrate and function within densely packed tumour tissue.

Tumours contain significant populations of macrophages, white blood cells whose natural function includes detecting and eliminating abnormal cells. Under normal circumstances, macrophages should attack cancer cells. However, the tumour microenvironment actively suppresses macrophage activity, releasing chemical signals that silence them. These tumour-associated macrophages remain present in the tumour but dormant, unable to carry out their anticancer function.

Current CAR-macrophage therapies have attracted significant research interest because macrophages can directly engulf and destroy cancer cells and stimulate surrounding immune cells to amplify the overall response. However, the standard process for creating CAR-macrophages requires extracting cells from a patient's blood, genetically modifying them in the laboratory, expanding them in culture, and then reinfusing them. This process is slow, expensive, and technically demanding, limiting its scalability.

The KAIST team, led by Professor Ji-Ho Park of the Department of Bio and Brain Engineering, bypassed this limitation by developing a method to reprogram macrophages already inside the tumour, directly in the body. Their approach uses lipid nanoparticles engineered to be selectively absorbed by macrophages, loaded with mRNA that encodes the CAR protein responsible for cancer recognition, and an immunostimulant that activates the macrophage's anticancer capabilities. When the drug is injected into a tumour, the nanoparticles are taken up by tumour-associated macrophages. Those macrophages then produce the CAR protein on their own surface, transforming themselves into active cancer-targeting cells without ever leaving the body.

In laboratory experiments and mouse models of melanoma, the approach demonstrated significant suppression of tumour growth and activation of a systemic immune response that extended beyond the local tumour environment. The research is still at a preclinical stage and has not yet entered human trials. Clinical translation will require safety evaluation, optimisation of nanoparticle targeting specificity, and confirmation of effectiveness across different solid tumour types.

For the oncology research and clinical trials community, this approach is notable on several levels. First, it addresses the access and scalability problem of ex vivo CAR therapy by performing the cell engineering process inside the patient, potentially dramatically reducing cost and complexity. Second, it uses mRNA-loaded lipid nanoparticles, a delivery mechanism validated at scale by COVID-19 vaccine development. Third, it represents a distinct approach from CAR-T cell therapies, which have shown extraordinary results in blood cancers but limited effectiveness in solid tumours due to difficulty penetrating the dense tumour stroma.

Sources: ScienceDaily January 2026 | ACS Nano (2025, DOI: 10.1021/acsnano.5c09138) | KAIST Official Announcement