Johns Hopkins spinout spotlights a showcase animal test for antibody-ligand I/O traps — matching checkpoints with a rein on Tregs
With the blockbuster popularity of PD-1/L1 checkpoints changing the way tumors are treated around the world, the spotlight in research has shifted away from the successful though limited and somewhat crude first generation of these therapies to new ways to amp up their efficacy and durability.
One of the key hurdles, investigators have found, is the generation of regulatory T cells — Tregs — that suppress the immune response to cancer cells. And now a research team at Johns Hopkins led by Atul Bedi, a practicing oncologist and associate professor at Johns Hopkins University School of Medicine, say they have developed an antibody/ligand with a bifurcated warhead that can do a better job on the checkpoint side while blunting the Tregs that both prevent efficacy as well as help trigger a gradual loss of potency, leading to recurrence.
Working with software from the AI/machine learning specialists at Baltimore-based Insilico Medicine, researchers were able to do a significant amount of pathway exploration. They found that activation of the transforming growth factor-β (TGFβ) pathway was linked to the presence of the FOXP3 biomarker, spotlighting its link to Tregs. So the research team under Bedi created a Y-trap: fusing a CTLA-4 antibody as well as a PD-1 with a “TGFβ trap” designed to turn off, redirect and delete Tregs.
“This Y-trap not only disables CTLA-4 function, but disrupts the TGFβ feedback loop that is necessary for induction and maintenance of Tregs in the tumor,” says Bedi, who licensed the tech through Johns Hopkins to a startup called Y-Trap, Inc, which he manages with San Francisco-based CEO Sonia Bhanot.
This first project is still very much at the preclinical phase, with researchers working with mouse models of cancer. And longterm success with mouse models of cancer is rare. But the use of AI combined with better drug design has delivered what the researchers vow is a superior antibody-ligand construct, with animal data to back up its claim that the Y-traps are able to surpass atezo (Roche’s PD-L1 Tecentriq) and avelumab (Bavencio from Pfizer and Merck KGaA) with a promise to be able to break out of the 1 in 5 or so ratio on efficacy with a better prognosis for durability. And they’re focused initially on cancers where there’s a known resistance to checkpoint therapies.
Bedi explained to me that his work on Y-traps has been evolving for the past 6 years. Bootstrapping the work with grants from the NIH as well as regional groups, he says he’s brought in a seed round from private investors and built the framework for an advanced preclinical platform company with a new technology that can branch out into a variety of different fields, including infectious diseases, where the immune system plays a key role. And he has a full set of I/O Y-trap programs that can be brought together to attack different hurdles — like cell exhaustion — facing the first generation of checkpoints, getting straight into the tumor microenvironment.
With AI and machine learning, says Bedi, you can feed in data on responses “that gives you clues as to which Y-traps are best to address them.” Go further down the road, he says, and eventually you’ll be able to personalize these Y-trap constructs to individual patients, rather than just certain subgroups based on key biomarkers.
“My aim in this field is aging research,” notes Alex Zhavoronkov, the CEO at Insilico Medicine, “looking at new ways to make the immune system look younger; immune senescence, waking up the immune system.”
Last summer Insilico partnered with GlaxoSmithKline on their maiden effort on AI and machine learning in drug development. And Zhavoronkov has been backed by UK billionaire Jim Mellon, who helped found the startup Juvenescence, which is making aging research its central focus, at a time most of the players in R&D are just getting their feet wet with this technology.
The next step at Y-Traps Inc is to move past the seed funding and raise a venture round to create a full-fledged platform biotech that can move to file an IND and then start human studies in I/O, where the technology can address the most obvious challenges well recognized in the field. Then they can start branching out, exploring industry partnerships to leverage more work in the field.
That all fits with a key trend in oncology research. As a tsunami of checkpoints promises to commoditize the first generation of checkpoints, the emphasis now is on the next-gen innovations that can maintain a product’s unique status. And Bedi is betting that Y-traps can play a big role in that.
We’ll keep you posted on that.
Illustration: Insilico Medicine