Each year, Nature Biotechnology highlights companies that have received sizeable early-stage funding in the previous year. A rescued chimeric antigen receptor from Cargo Therapeutics for hard-to-treat lymphomas shows promise in patients while a trispecific advances toward human testing.

Treating childhood leukemia was the original concept behind Cargo Therapeutics. “Agents developed for pediatric diseases are not moving forward in the corporate pipelines because of the small market,” says Crystal Mackall, who founded Cargo with fellow Stanford pediatric hematologist Robbie Majzner and lawyer Nancy Goodman, executive director of Kids v Cancer, a nonprofit dedicated to improving cancer treatments for children. In the early 2010s, Mackall’s lab at the National Cancer Institute (NCI) developed a chimeric antigen receptor1 (CAR) for T cell targeting of the CD22 antigen on B cells. CARs couple extracellular antibody antigen-binding domains to transmembrane T cell signaling domains in a single synthetic receptor, which is transfected ex vivo into patient T cells for expansion and then reinfusion. This allows T cell targeting of tumor surface antigens. The US Food and Drug Administration has approved six such therapies for various blood cancers.

Cargo CEO Gina Chapman Credit: Cargo Therapeutics

Almost a decade ago, the NCI launched a phase 1 study2 of its CD22 CAR in B cell acute lymphocytic leukemia (B-ALL) and then licensed the technology to Juno Therapeutics, which was then developing the CD19-targeted CAR-T cell therapy later approved as Breyanzi (lisocabtagene maraleucel). But Juno did not move forward with the CD22 CAR-T, not in B-ALL and not in large B cell lymphoma (LBCL), despite Mackall’s urging. So Mackall, who moved to Stanford in 2016, started an academically funded LBCL trial there, and a second trial in B-ALL.

The NCI, with Goodman’s help, eventually recovered the license from Juno. Goodman, Mackall, Majzner and Stanford postdoc Louai Labanieh founded Syncopation Life Sciences in December, 2019, relicensing the CD22 CAR-T therapy (renamed firicabtagene autoleucel, or firi-cel) from the NCI, for treating pediatric B-ALL. “The founding of the company was very much driven with that goal in mind,” says Mackall. “While we were doing that, we saw the activity at Stanford in lymphoma.” The first three LBCL patients treated3 achieved a complete response. They remain in remission today. Almost overnight LBCL, which mostly affects older adults, became the first priority of the company, later renamed Cargo.

The medical need is there. Three different CD19 CAR-T cell therapies are approved for treating relapsed or refractory LBCL, and they induce durable long-term remission in 40 percent of patients. But “for the patients who relapse after CD19 CAR, the median survival is only about six months,” says MD Anderson Cancer Center hematologist Sattva Neelapu. Current CD19 failure treatment options put 20–37% of patient cancers into remission, but most relapse later.

In phase 1, firi-cel achieved a complete remission rate of 53% (ref. 4). Notably, 73% of those showing a complete response (at the phase 2 dose) remained in remission after a median follow-up of 30 months, “remarkable for the post-CD19 CAR space,” says Neelapu. Because relapses in LBCL after two years are rare, says Neelapu, most are likely cured. “The limitation, of course, is it’s a single institution study, relatively small sample size, about 38 patients,” he says. “So we’ll have to see what the multicenter study will show.… We also want to see long-term tolerability and safety.” That phase 2 study, which could lead to Food and Drug Administration approval, is “progressing very well,” says Cargo CEO Gina Chapman, with interim results scheduled for release in the first half of next year.

Firi-cel’s efficacy stands out among CD22 CAR-T cell therapies, which have generally been disappointing. Firi-cel employs a distinctive CAR. Even though affinity for CD22 is suboptimal, the receptor binds a membrane-proximal CD22 epitope, explains Mackall, making it more potent. Also there is a little bit of ‘tonic signaling’ — T cell activation in absence of antigen. Too much tonic signaling is bad because it induces T cell exhaustion and apoptosis, but these T cells “seem to be primed and ready to go as soon as they are infused into patients,” says Neelapu. “That’s the unique biology that seems to make this CD22 CAR more effective.”

Firi-cel is Cargo’s lead product, but the CAR-T cell field is evolving toward multispecific CARs because tumors develop resistance by downregulating any single target antigen. Chapman calls multispecifics “the next generation of potentially curative CAR-T cell therapies,” and for more than two years Cargo has been developing a trispecific CD19/CD20/CD22 CAR. Unfortunately, bispecific CARs have proven little better than monospecifics in clinical trials. For example, Mackall, at Stanford, tested a bivalent CD19–CD22 tandem CAR5, with two distinct antigen-binding sites engineered into a single chimeric receptor, but anti-CD22 activity fell off.

Cargo’s molecule, by contrast, is a tricistronic CAR, with the three separate CARs engineered into a single lentiviral vector. “I think the reason our multivalent didn’t work is just that the CD22 binding portion was likely sterically inhibited,” says Mackall. “That would not happen if we made it a separate CAR on the surface.”

Cargo screened thousands of molecules before settling on its trispecific CAR, CRG-023. It has novel binders for CD19 and CD20: “lower affinity binders that deliver really interesting potency across antigen densities,” says Cargo chief scientific officer Michael Ports. That should prevent tumor antigen escape. It also includes a CD2 costimulatory signaling domain, in addition to the standard 4-1BB. Many tumors downregulate CD58, the ligand for CD2, which promotes resistance to CAR-T therapy. Now, even in the absence of CD58 on tumors, CRG-023 is “firing the CD2 signal,” says Ports. “It delivers a very robust, beautiful activation response in our product.” Preclinical data are unpublished, but the company is moving toward clinical trials “as quickly as we can,” says Chapman.

Bicistronic and tricistronic CARs are hard to make, for two reasons. First, the vectors require long sequences, which can impair viral packaging, disrupt transduction and reduce CAR expression. Ports says Cargo carefully screened for transduction efficiency and high-quality lentiviral titers before selecting CRG-023. The other challenging issue is T cell exhaustion. “You put two or three CARs in the same cell, there’s concern that the cell may get overactivated and undergo exhaustion and activation-induced cell death,” says Neelapu. While a legitimate concern, says Ports, “we’re simply just not seeing it. We see that the total level of CAR receptor expressed is comparable to monospecific CARs.”

Cargo also hopes to employ Stash, a proprietary platform for homogeneous multiple transgene delivery. Applications include complex tumors requiring more CAR engineering, and improved manufacturing. Mackall’s Stanford lab has pioneered several novel CAR approaches, including regulatable CARs6 and logic-gated CARs7, but these are not active Cargo projects. “I’ve been told by seasoned entrepreneurs that a company rarely dies of starvation — mostly they die of indigestion,” Mackall says. “Would I like to do all those other things? Yes. Maybe Cargo would do it in the future, who knows, depending on the opportunities. But we thought coming up with a way to provide an opportunity for the 60% of lymphoma patients who fail CAR-19s was a great way to get a company going.”

Cargo has not forgotten its original raison d’être, pediatric leukemia. A new, potentially pivotal CD22 CAR-T trial in B-ALL is in the works. “That is really gratifying to me,” says Mackall. “Hopefully we’ll be able to get an indication in peds at some point in the future. Yes, it’s taken a long time.”