Human macrophages equipped with chimeric antigen receptor constructs infiltrate solid tumors, ingest malignant tissue, and stimulate adaptive immunity in mouse models. Several new biotech companies are racing to bring the technology into clinical trials.

First, there were chimeric antigen receptor (CAR) T cells. Then came CAR natural killer cells. Now, scientists at the University of Pennsylvania (Penn) Perelman School of Medicine in Philadelphia have outfitted human macrophages with CAR constructs, creating cancer-homing Pac-Man–like cells with the ability to infiltrate solid tumors, ingest malignant tissue, and stimulate adaptive immunity in mouse models (Cancer Discov 2020;10:484).

In mice implanted with HER2-expressing ovarian cancer cells, a team led by Saar Gill, MD, PhD, and Michael Klichinsky, PharmD, PhD, showed that macrophages transfected with a HER2-targeting CAR could decrease tumor burden and prolong survival (Nat Biotechnol 2020 Mar 23 [Epub ahead of print]). The therapeutic benefit derived not only from direct antigen-specific phagocytosis, but also from indirect proinflammatory effects, including enhanced antigen processing and cross-presentation to tumor-specific T cells.

“They have the ability to flip a cold tumor into a warm tumor if they are properly engineered,” Klichinsky says.

Getting the engineering to work was not trivial. Because the lentiviral and retroviral vectors typically used for CAR transduction don't work well with primary macrophages, the researchers developed an adenoviral system that allowed them to deliver the CAR transgene while inducing the cells to adopt the “M1” phenotype associated with producing high levels of proinflammatory cytokines. Even in the face of cytokines that normally skew macrophages toward the “M2” suppressive state, the CAR-transfected cells maintain their antitumor activity. “They seem to be locked into this phenotype,” Klichinsky says.

M2-type macrophages are often one of the most abundant cells in tumor infiltrates, “so the idea that you would want to add more macrophages to the tumor microenvironment isn't immediately obvious,” says Meghan Morrissey, PhD, of the University of California, San Francisco (UCSF), who was not involved in the research. As such, this “suggests that macrophages can be reprogrammed to go from being a negative all the way to being a positive.”

Saar and Klichinsky's work underpins Carisma Therapeutics, a company they founded in 2016 that has raised approximately $60 million to date. According to Klichinsky, who serves as vice president of discovery research, Carisma's lead CAR-macrophage candidate, an autologous HER2-targeted therapy called CT-0508, is slated to enter human testing before the end of the year; programs centered on targeting mesothelin and PSMA remain in earlier stages of discovery.

Two competing firms have similar technologies in development. Thunder Biotech spun out of the laboratory of Kim O'Neill, DPhil, of Brigham Young University in Provo, UT, who has shown that mesothelin-targeted CAR macrophages can shrink tumors in a mouse model of triple-negative breast cancer. Meanwhile, Myeloid Therapeutics—whose scientific founders include UCSF's Ron Vale, PhD, and Siddhartha Mukherjee, MD, DPhil, of Columbia University in New York, NY—is engineering myeloid cells for vaccine and tumor-killing purposes.

According to Myeloid's Chief Scientific Officer Daniel Getts, PhD, these undifferentiated myeloid cells traffic more effectively to tumors than terminally differentiated macrophages. In contrast to Carisma, both Myeloid and Thunder are using nonadenovirus transduction systems.

Yet, perhaps the most notable difference between the three companies' technologies lies in the intracellular signaling domain chosen for each CAR construct: CT-0508 takes advantage of CD3ζ, part of the T-cell antigen receptor and the same domain used in first-generation CAR T-cell therapies, whereas Thunder's and Myeloid's products incorporate signaling domains from Toll-like receptors and phagocytotic receptors, respectively.

Judging by the Penn team's results, CD3ζ, despite its usual role in T-cell stimulation, seems to prime macrophages into an M1 phenotype, O'Neill says, “but it's probably not as good as the inherent mechanism” of macrophage activation.

Carisma is now developing novel co-stimulation domains that are “rationally designed” for macrophage activation, Klichinsky says. “If we find significant improvement, those will be implemented in future iterations” of the therapy. –Elie Dolgin

For more news on cancer research, visit Cancer Discovery online at