For years, CAR-T cell therapy has stood as one of cancer medicine’s most striking successes. By removing a patient’s T cells, genetically engineering them to recognize a cancer-associated molecule and returning them to the body, physicians have produced durable remissions in some leukemias, lymphomas and other blood cancers. Yet the same strategy has struggled against solid tumors, including cancers of the brain, liver, lung, breast and pancreas. Researchers at the USC Viterbi School of Engineering now say they have developed a way to give CAR-T cells a temporary target inside solid tumors, potentially overcoming one of the central barriers that has limited the treatment’s reach.
The approach, described in a study published in Science Advances, is called SHIFTERS. Rather than searching for a naturally occurring antigen that appears exclusively on cancer cells, the system is designed to make tumor cells display one. Its target is CD19, a surface protein already recognized by many clinically developed CAR-T cells. CD19 is normally associated with B cells and is not broadly displayed by solid tumors. SHIFTERS uses a genetic program, combined with the low-oxygen environment characteristic of many tumors and externally applied focused ultrasound, to prompt selected cancer cells to temporarily present CD19 on their surfaces.
The system is built around a two-signal logic gate. The first signal is hypoxia, or low oxygen concentration, a hallmark of rapidly growing solid tumors. As malignant tissue expands, its blood vessels often fail to deliver enough oxygen, creating oxygen-deprived regions within the tumor mass. SHIFTERS is engineered to respond to molecular conditions associated with this hypoxic environment. The second signal is supplied by a physician through focused ultrasound. Because ultrasound can be directed toward tissue beneath the skin without an incision, it provides a spatial control mechanism: the genetic circuit is intended to become active only where tumor biology and the physician’s acoustic instruction coincide.
When both conditions are present, the engineered program activates CD19 production at the tumor-cell surface. This converts previously invisible cancer cells into temporary beacons for CAR-T cells. The T cells do not need to recognize the original biology of the tumor, which may vary widely among patients and cancer types. Instead, they use their existing CD19-specific receptor to identify the primed cells, form an immunological synapse and release cytotoxic molecules such as perforin and granzymes. These molecules damage the target cell and initiate its death. According to the USC team, the induced marker can remain detectable for approximately one week, creating a treatment window during which the location and timing of immune-cell activation can be controlled.
That design addresses a fundamental difficulty in solid-tumor immunotherapy. Blood cancers often carry relatively uniform surface markers that can be targeted throughout the malignant cell population. Solid tumors, by contrast, develop from normal tissues and frequently share many molecular features with healthy cells. Even when a candidate antigen is abundant in a tumor, it may also appear in essential organs, raising the risk of dangerous off-target damage. Tumors are also heterogeneous: different regions, and sometimes different cells within the same tumor, may carry different mutations and surface proteins. A temporary, externally controlled marker could offer a way to separate target recognition from the tumor’s naturally inconsistent antigen landscape.
The researchers evaluated SHIFTERS through a series of preclinical experiments. They first studied the system in cultured cancer cells, where they could measure genetic activation, CD19 display and CAR-T-mediated killing under controlled oxygen and ultrasound conditions. They then moved to three-dimensional tumor models, which reproduce some of the physical barriers found in actual tumors, including dense cellular organization and limited diffusion. Finally, they tested the strategy in animal models carrying human brain and liver tumors. Across these stages, the team reported that ultrasound treatment increased CD19 expression and enabled CAR-T cells to attack tumor tissue more effectively than they did without the priming step.
In animal experiments, tumors exposed to the ultrasound-guided system shrank substantially, while tumors that did not receive the same activation continued to grow. The findings do not yet establish that SHIFTERS will work in patients, but they suggest that the approach can translate a physical treatment signal into a molecular recognition signal. Focused ultrasound is already used in medicine for several purposes, including imaging and selected therapeutic applications, although the acoustic parameters and safety requirements for this system would need to be carefully defined for each tumor type and anatomical location. Treating a brain tumor, for example, introduces additional challenges because the skull can distort and weaken ultrasound waves, while liver and pancreatic tumors may move with respiration.
One of the study’s most notable observations was that not every cancer cell needed to display CD19 for the treatment to produce a broader antitumor response. The team reported that activating the marker on roughly 10% to 25% of tumor cells was sufficient to drive substantial killing in laboratory models. These CD19-positive cells appeared to act as “training centers” or initiating targets for the immune response. After recognizing and destroying them, CAR-T cells and other immune mechanisms may contribute to wider damage across neighboring cancer cells, including cells that never displayed the engineered marker. This effect could be especially important in heterogeneous tumors, where reaching every malignant cell with a genetic therapy may be unrealistic.
The researchers describe this phenomenon as a form of bystander or collateral immune killing, although its exact biological basis will require further investigation. Direct CAR-T recognition of CD19-positive cells may release inflammatory signals, alter the tumor microenvironment and expose additional tumor antigens to the immune system. Dying cancer cells can also release intracellular proteins that are processed and presented to other immune cells, potentially broadening the response beyond the synthetic CD19 target. Whether this amplification remains effective in the immunosuppressive environment of human solid tumors is unknown. Many such tumors contain regulatory immune cells, abnormal blood vessels, fibrotic tissue and metabolic conditions that can restrict T-cell entry and function.
Delivery is currently the largest practical obstacle. SHIFTERS requires tumor cells to receive genetic instructions encoding the hypoxia- and ultrasound-responsive program, and those instructions must reach enough of the tumor without spreading dangerously to healthy tissue. The USC team is comparing lipid nanoparticles with modified viral vectors. Lipid nanoparticles can carry nucleic acids without using a replicating virus and have become important in several biomedical applications, but their distribution and persistence in solid tumors can be uneven. Viral vectors may deliver genes more efficiently to certain cells, yet they introduce additional questions involving immune reactions, dose control, manufacturing and the possibility of unintended expression outside the treatment zone.
The study was conducted in animals and has not yet been tested in people. Before clinical trials could begin, investigators would need to establish reliable delivery, determine how long CD19 expression lasts, define safe ultrasound intensities and evaluate the risk of activating CAR-T cells in healthy tissue. They would also need to study whether repeated treatment is possible, how the therapy behaves in tumors with different oxygen patterns and whether engineered cells can penetrate the tumor at clinically meaningful levels. The work was funded in part by the National Institutes of Health. The authors disclosed that Yi Wang is a scientific co-founder and consultant of Cell E&G Inc. and Acoustic Cell Therapy Inc., while the other authors reported no competing interests.
Despite these limitations, SHIFTERS represents a shift in the way researchers are approaching the antigen problem. Instead of waiting for solid tumors to reveal a perfect natural marker, the strategy seeks to install a temporary one under the direction of tumor physiology and focused ultrasound. Peter Yingxiao Wang, who led the work at USC, said the broader goal is to “rewire” both the tumor and the T cell so they can recognize and destroy one another more effectively. The technology remains years from clinical use and will require larger animal studies, improved delivery systems and rigorous safety testing. But if those hurdles can be overcome, an ultrasound-controlled, temporary antigen could give CAR-T therapy a new route into cancers that have so far remained largely beyond its reach.
Subject of Research: Animals
Article Title: Ultrasound priming gated by solid tumor hallmarks to guide CAR-T therapy
News Publication Date: 10-Jun-2026
Web References: https://www.science.org/doi/10.1126/sciadv.aed0666; https://viterbischool.usc.edu/; https://www.cancer.gov/about-cancer/treatment/research/car-t-cells
References: Science Advances, DOI: 10.1126/sciadv.aed0666
Keywords
CAR-T cell therapy, solid tumors, cancer immunotherapy, focused ultrasound, SHIFTERS, CD19, hypoxia, genetic engineering, glioblastoma, liver cancer, tumor targeting, immune-cell therapy
