Researchers at the University of Southern California have identified a way to manufacture immune cells at scale, potentially breaking a supply barrier that has hindered cancer immunotherapy for years. In a study published June 19, 2026 in the journal Cell, a team led by Dr. Qi-Long Ying of the Keck School of Medicine reported that a specific class of immune cell precursors, called granulocyte-monocyte progenitors or GMPs, can be grown indefinitely in the laboratory and engineered to hunt tumors.
Macrophage Therapies Face Challenges
Macrophages are naturally suited to fight solid tumors because they infiltrate tissues and consume foreign material. However, current approaches struggle to produce enough of these cells. A Phase 1 trial reported in February 2025 in Nature Medicine showed that macrophage therapy targeting HER2-overexpressing tumors was safe, but it relied on an autologous approach where cells are harvested from each patient. This process is expensive, requiring weeks of production time, and cannot be stockpiled for widespread use.
Unlike CAR-T therapies, which have treated thousands of patients with blood cancers, no CAR-T therapy has been approved for solid tumors, which represent approximately 90 percent of cancer cases globally. The USC team aims to solve this by using a renewable, off-the-shelf source of cells.
Related: Claude Sonnet 5 Default Closes Opus Gap
Challenging Traditional Biology
The key breakthrough involves the developmental stage of the cells used. Classical biology suggests that once cells become progenitors like GMPs, they irreversibly lose the ability to self-renew. The USC team found this assumption is conditional. By using a chemical cocktail to block the programs driving GMPs toward maturity, they kept the cells dividing while maintaining their identity and ability to eventually become functional macrophages.
The cells retained their markers and functional capacity across many rounds of division in the lab.
Engineering Immune Cells
Growing GMPs at scale addresses the supply problem, but the cells also need to recognize cancer. The team engineered their expanded GMPs with a chimeric antigen receptor that directs the cell to recognize specific markers.
Related: Retatrutide matches bariatric surgery as monthly GLP-1 debuts
Testing in Mice and Humans
When the team infused the CAR-engineered GMPs into tumor-bearing mice, the cells demonstrated persistence. Unlike mature macrophage therapies that are cleared quickly, these cells seeded hematopoietic niches in the bone marrow and generated a continuous supply of donor-derived macrophages. This delayed disease progression in mice with leukemia and solid tumors.
They also tested the platform in mice with chronic granulomatous disease, a rare inherited immunodeficiency. The treatment restored the animals’ ability to fight bacterial infection.
A collaborating laboratory at Stanford University independently reproduced the long-term maintenance and genetic engineering of GMPs. Ravi Majeti, MD, PhD, director of the Stanford Institute for Stem Cell Biology and Regenerative Medicine, led the confirmation. Majeti is also a co-founder of Myelogene Inc., the spinout company that holds the exclusive license to the patent portfolio covering the USC-Stanford methods.
Related: Energy Hog Alert: Senators Probe Data Center’s Hidden Power Consumption
The Road Ahead
The signal from the mouse model is encouraging, but the clinical path is long. Myelogene has not yet filed an investigational new drug application with the FDA, and no human clinical trial for the GMP platform has been announced. The FDA raised the regulatory bar for cell therapies in May 2026, now requiring future CAR-T approvals to demonstrate superiority over standard of care.
Myelogene’s founders also provided partial funding for the research through a sponsored research agreement. This means none of the paper’s authors is a fully independent evaluator of the platform’s prospects. The human version of this platform will need to show that the bone marrow engraftment, progenitor identity maintenance, and anti-tumor activity observed in mice hold up in people.
Ying stated that the future of immunotherapy may depend not only on designing better CAR receptors, but also on choosing the right developmental stage of the cell.