Columbia researchers have created a next-generation cell therapy, called HIT cells, that can detect and destroy solid tumors in mice by homing in on a hard-to-spot cancer marker. The work could open a new front in the fight against some of the deadliest cancers.
A new kind of immune cell therapy has wiped out aggressive solid tumors in mice, raising hopes for better treatments against cancers that have long resisted today’s most advanced drugs.
Researchers at Columbia University’s Initiative in Cell Engineering and Therapy (CICET) have developed ultra-sensitive “HIT” cells, a close cousin of CAR-T cell therapy, that can hunt down and destroy kidney, pancreatic and ovarian cancers in mouse models. The work, published Feb. 26 in the journal Science, tackles one of the biggest reasons cell therapies have struggled against solid tumors.
CICET director Michel Sadelain, who helped pioneer modern CAR-T therapies for blood cancers, noted the new approach addresses a key barrier that has kept similar treatments from working in most solid tumors.
“Curing solid tumors is not easy, but this work solves one piece of the puzzle,” Sadelain, who is also the Herbert and Florence Irving Professor of Medicine at Columbia University Vagelos College of Physicians and Surgeons, said in a news release.
CAR-T cell therapy has transformed care for certain leukemias and lymphomas by reprogramming a patient’s own T cells to recognize a specific molecule on cancer cells and attack. Those blood cancers are relatively straightforward targets because nearly every malignant cell is coated with the same marker, such as CD19, making them easy for engineered T cells to find.
Solid tumors are far more complicated. Cancer cells within the same tumor can look very different from one another, and they rarely share a single, abundant marker that can serve as a reliable bullseye for therapy.
“Some molecules have been identified that are found in 25%, 50%, or 75% of tumor cells,” Sadelain added.
If a therapy only recognizes cells with high levels of a target, many cancer cells can slip through, survive treatment and eventually drive the tumor’s return. To truly cure a solid tumor, doctors need a way to track down even the rarest, hardest-to-spot malignant cells.
The Columbia team focused on a molecule called CD70. Earlier studies suggested that CD70 appears on some cells in a range of solid tumors, but at inconsistent levels. That made it seem like a poor candidate for cell therapy, which typically needs a strong, clear signal to lock onto.
Study lead author Sophie Hanina, a research associate scientist at CICET, suspected that earlier work had underestimated how widely CD70 appears on tumor cells. She developed more sensitive methods to detect the molecule and found that while CD70 levels varied widely from cell to cell, every cancer cell in the tumors she studied carried at least a few copies on its surface.
That discovery reframed CD70 as a potentially powerful “homing beacon” for therapy — if the immune cells could be made sensitive enough to see it when it was present only in tiny amounts.
Conventional CAR-T cells are not built for that level of subtlety. They typically need a high density of their target on a cell’s surface to latch on and activate. When Hanina tested CD70-targeted CAR-T cells against solid tumors in the lab, they only recognized and killed cells with abundant CD70, leaving behind those with lower levels.
To get around that limitation, Hanina turned to HIT cells, a next-generation platform under development in Sadelain’s lab. HIT cells are designed to combine the programmable targeting of CAR-T cells with the natural sensitivity of unmodified T cells, allowing them to respond to vanishingly small amounts of a target molecule.
In mouse experiments, HIT cells engineered to recognize CD70 completely eradicated pancreatic, kidney and ovarian tumors. In contrast, traditional CAR-T cells aimed at CD70 only cleared part of the cancer, mirroring the incomplete responses seen in earlier attempts to use CD70 as a target in patients.
The HIT cells also appeared to spare healthy tissue in the mice. Most normal cells in the body do not express CD70, so the engineered cells had little reason to attack noncancerous targets, an important safety consideration for any future human therapy.
The results suggest that the problem with earlier CD70-directed approaches was not the target itself, but the inability of standard CAR-T cells to detect it when it was present at low levels. By boosting sensitivity, HIT cells effectively redefine which tumors can be considered “CD70-positive” and therefore targetable.
The work also speaks to a broader challenge in the field. Many solid tumors likely carry useful markers at low density, but current tools cannot see them well enough to exploit them. More sensitive cell therapies could open up a wider range of targets across many cancer types.
Hanina and Sadelain are now preparing to move their CD70 HIT cells into early-stage clinical trials at Columbia University Irving Medical Center, starting with patients who have ovarian and other CD70-expressing cancers. The team noted that CD70 is present, to some degree, in nearly 20 types of cancer, including notoriously hard-to-treat tumors such as glioblastoma and pancreatic adenocarcinoma.
If the approach proves safe and effective in people, it could offer a new option for patients whose tumors have few other vulnerabilities. It could also serve as a template for designing HIT cells against additional low-level targets in other cancers.
“Studies suggest that the escape of undetected cancer cells is the key impediment to therapeutic success with conventional CAR-T therapy,” added Hanina. “We hope our CD70-directed HIT cells help us find a way to eradicate the entire tumor.”
For now, the findings are limited to animal models, and many questions remain about how HIT cells will behave in the complex environment of human tumors. Solid cancers build physical and chemical barriers that can exhaust or block immune cells, even when they can recognize their targets.
Still, by showing that it is possible to engineer T cells sensitive enough to track a subtle but widespread marker like CD70 and clear tumors in mice, the Columbia team has taken an important step toward making cell therapy a realistic weapon against solid cancers, not just blood cancers.