[Interview] Cracking cancer’s cold cases: the startup that may have found immunotherapy’s blind spot

Cellus is betting that reshaping the terrain around a tumor -- not just going after the tumor itself -- can flip the switch on immunotherapy resistance.

The young Korean startup’s lead candidate, CLS-A1, blocks CXCL16, a chemokine secreted by tumors to summon immune-suppressing macrophages and myeloid cells. It also keeps CD8+ T cells, the immune system’s primary enforcers, from breaching the tumor.

The approach sounds roundabout, but the company insists the results are anything but.

“In cancer models where immunotherapy alone couldn’t kill any tumor cells, combining it with our drug killed up to 87 percent,” CEO Kim Seong-keun said in an interview with Korea Biomedical Review. 

That result came in triple-negative breast cancer (TNBC), one of the most notoriously unresponsive tumor types. CLS-A1 not only paired well with PD-1 inhibitors but also shrank tumors by up to 57.5 percent on its own, Kim said, including those barren of T cell infiltration.

He called TNBC the “prototypical non-responsive tumor model” and the clearest place to show that CLS-A1 could do what others couldn’t.

Where most immune-oncology drugs try to boost the T cells directly, CLS-A1 takes a different path. It doesn’t destroy the suppressive cells outright. Instead, it cuts the line by intercepting CXCL16’s signal without wiping out the macrophages that send it. 

That design, Kim said, “minimizes off-target effects,” and so far, preclinical data show no major safety concerns.

“Checkpoint inhibitors are powerful weapons,” he said. “But turning the tide of battle requires strategic coordination.” He paused for effect, then added, “T cells can’t complete their mission if other immune cells are actively helping the enemy. Some are even traitors.”

Cellus isn’t alone in trying to improve checkpoint efficacy, but it’s aiming earlier in the immune cascade, before the battle even begins.

The company was founded in 2020, but the science behind it stretches back years. Kim, an immunologist at the Catholic University of Korea, had spent half a decade studying how macrophages shaped tumor behavior. He kept circling one target.

“I couldn’t move forward until I believed in it,” he said. “Just identifying the right target took five years. We built the company after that.”

His co-founder, Seoul National University endocrinologist Cho Sun-wook, helped zero in on CXCL16. Together, they filed Korea’s first antibody patent against the target in 2021, followed by filings in the United States, Europe, and Japan.

While others explored small molecules and peptides, Cellus took the antibody route. “We’re probably the first,” Kim said. “And globally, we’re the fastest.”

The patent outlines potential use in a wide range of cancers, including prostate, anaplastic thyroid, and non-small cell lung cancer. It also details plans to potentially expand the platform into bispecifics and antibody-drug conjugates.

The therapy is geared toward tumors that overexpress CXCL16, a biomarker that can be measured using standard tools like flow cytometry or immunohistochemistry. And while most companies enter the clinic through traditional phase 1 trials, Cellus is eyeing a more nimble approach.

It plans to generate early human data through investigator-led studies and, according to Kim, has already held regulatory meetings with several global players. “We’ll likely need to co-develop with checkpoint inhibitor companies,” he said, framing CLS-A1 as an immune system upgrade rather than a standalone fix.

The company has raised 4.4 billion won ($3.22 million) to date and is seeking another 12 billion to complete studies required before entering human trials. To keep operations moving, it’s also commercializing its AI-powered platform for modeling the tumor microenvironment and predicting toxicity.

Cellus is positioning CLS-A1 as a foundational technology, one that addresses the broader problem with immunotherapy. Research suggests that only 20 to 40 percent of patients respond to checkpoint inhibitors, depending on tumor type and immune status. And for cold tumors that resist T cell infiltration altogether, that number drops even lower. By some estimates, non-responders account for 70 percent of all solid tumors.

“Most cancer drugs aim at tumor cells,” Kim said. “We’re aiming at the environment that protects them.”

CLS-A1 may be Cellus’s first act, but it’s not its last. The company is developing CLS-A2, a bispecific antibody that hits both CXCL16 and RANKL, the latter a key driver of bone resorption and the target behind Amgen’s bone blockbuster Prolia (denosumab).

“Prolia blocks RANKL,” Kim said. “But CXCL16 stimulates RANKL in the first place. So by blocking both, we expect a stronger effect.”

That’s the kind of logic Cellus is betting on. Fix the environment, and the rest might follow.