Gilman’s Next CEO Trick: Taming CAR-T Cells With Obsidian

Xconomy Boston — 

Genetically modified medicine is here. Two CAR-T therapies, made from a patient’s living T cells, are on the market in the U.S., and many more are in various stages of clinical testing.

As remarkable as they might seem, however, these medicines are still fairly crude, with no way to control them once inside a patient’s body. But a host of academics and industry researchers want to add controls to cell therapies and broaden their use. A startup called Obsidian Therapeutics is the latest to emerge.

At the helm of Obsidian, of Cambridge, MA, is veteran biotech entrepreneur Michael Gilman, who is simultaneously running another nearby startup, Arrakis Therapeutics. Obsidian today is announcing it has raised $49.5 million from GV, founding backer Atlas Venture and others to move its work forward.

It is joining a crowded field. Multiple efforts to control cell are advancing. Bellicum Pharmaceuticals (NASDAQ: BLCM) has been developing controllable T cells. Before it was acquired by Gilead Sciences (NASDAQ: GILD), Kite Pharma invested in Cell Design Labs, a spinout from the University of California, San Francisco, that aims to build control switches into CAR-T products. Intrexon (NASDAQ: XON) is developing control switches for cell therapies. “There is a ton of interest in these approaches,” Gilman says.

CAR-T, in which a patient’s modified T cells transform into more efficient cancer killers, has made it to the market in the U.S. with products from Novartis (NYSE: NVS) for pediatric leukemia and Gilead for adult non-Hodgkin lymphoma. More from Juno Therapeutics (NASDAQ: JUNO), Bluebird Bio (NASDAQ: BIIB), Cellectis (NASDAQ: CLLS), and Nanjing Legend Biotech, among others, are in the pipeline. Huge sums of cash are being spent on the collective effort.

But CAR-T remains limited to people with certain leukemias and lymphomas who have run out of options. Other types of blood cancer are being targeted, as well. A farther horizon would be the more common solid-tumor cancers, such as lung, breast, and ovarian.

To reach these goals, CAR-T researchers need to understand and prevent some of the ferocious immune responses and other side effects, such as brain swelling, that CAR-T can unleash. How can the cells be tamed to be both effective and safe?

Obsidian’s solution originated in the labs of Harvard University and Stanford University in the 1990s and has found new life with the advancements in cell therapy.

Obsidian is working on several ways to control the behavior of modified T cells.

—First, it wants to make changes to the piece that defines a CAR-T cell: the chimeric antigen receptor. Current versions are modified to express a surface protein—a CAR—that is tuned into a specific tumor marker, or antigen, so it knows what to attack. It’s like training a police dog to sniff out a particular scent. Obsidian wants to shut down the CAR’s ability to receive that signal when CAR-T cells get too excited and multiply too fast. When they get excited by an antigen, they release more toxic chemicals called cytokines and kill more tumor cells. Too much cytokine release can be dangerous to the patient, however. An on-off CAR switch might reduce the T cell frenzy and keep side effects in check. It might also give the T cells a break. There’s an emerging argument that overstimulated CAR-T cells can exhaust themselves, allowing tumors to bounce back.

—Second, Obsidian wants to give CAR-T cells a genetically engineered boost of a particularly potent cytokine called IL-12, but only in specific circumstances to break down a tumor’s defenses. IL-12 is extremely dangerous to humans, so Obsidian has come up with a check: IL-12 that has an extra tag on it. The tag would signal to the body’s natural waste-disposal system that the IL-12 should immediately be trashed. The “trash” tag could be blocked, however, by the patient swallowing a pill. The pill’s chemistry would, in essence, jam up the trash signal and let IL-12 attack the tumor. Depending on how strong the pill is, the level could be dialed up and down, not just switched on and off. This could be a strategy for CAR-T attacks on solid tumors, which have more sophisticated layers of defense than blood-borne cancers.

—Third, Obsidian is working on a way to refresh CAR-T cells that have begun to fade.

Gilman acknowledges that the big CAR-T companies have already built the clinical and commercial infrastructure for cell products. Obsidian’s short-term prospects might lean on partnerships with them. Obsidian’s “long game,” says Gilman, is to make its own products with the “next level of cellular therapies—the cells with interesting biology but haven’t been domesticated yet.” Natural killer cells, another type of immune cell, could be fruitful for Obsidian, as could tumor infiltrating lymphocytes, which are T cells that have naturally figured out how to crack the defenses of a tumor.

Gilman won’t say which specific pills Obsidian intends to use as an on-switch, except that they are FDA-approved, safe, and generic. Preclinical studies are probably two years away, he says.

Between Obsidian and Arrakis, Gilman—who has run and sold two biotech startups since 2012—has a lot on his plate these days. Mondays and Tuesdays are spent at Arrakis, Wednesdays and Thursdays at Obsidian. “Fridays are up for grabs,” he says. “Whoever needs me, gets me.”

ShangPharma Investment Group, Vertex Ventures, and the venture arms of Takeda, Amgen, and Alexandria Real Estate Equities all took part in the financing.