Mustang Bio, City of Hope CAR-T therapy extends survival in phase 1 glioblastoma trial

Mustang Bio and City of Hope’s chimeric antigen receptor T-cell (CAR-T) therapy for recurrent glioblastoma resulted in stable disease or better in half of the subjects in a phase 1 clinical trial, including one patient whose cancer has not recurred for more than five years. 

The trial, described March 7 in Nature Medicine, is the largest reported so far on a CAR-T therapy in glioblastoma. It also builds the case for the feasibility and possible advantages of injecting CAR-T cells directly into a brain tumor and the cerebrospinal fluid, a method that could ultimately change how the disease is treated.

“I think a lot of our findings are reshaping the field and how we’re thinking about applying cell therapy to brain tumors,” Christine Brown, Ph.D., deputy director of the T Cell Therapeutics Research Laboratories at City of Hope and creator of the therapy tested in the trial, told Fierce Biotech Research in an interview. “This paper both sets the foundation clinically, but also tries to get at some of the features of glioblastoma that make it difficult to treat, including with immunotherapy.”

Into the brain 

Glioblastoma, also known as glioblastoma multiforme, is the most common and most aggressive primary brain tumor. For the first occurrence, the median survival without treatment is around nine months. But that figure grows only slightly with therapy to about 12 to 14 months, as the tumors return more resistant than before.

All 58 patients in the phase 1 study had experienced at least one recurrence prior to enrolling in the trial. The majority were being treated for their third, Behnam Badie, M.D., chief of neurosurgery at City of Hope and senior author on the study, told Fierce. They had previously received some combination of surgery, chemotherapy or radiation, the current standard of care for the disease.

“The gliomas become more aggressive when they recur,” Badie said. “So basically, [the patients] had no other treatment options when they saw us.” 

Each subject in the trial received a weekly dose of CAR-T cells aimed at the antigen interleukin-13 receptor alpha 2, or IL13Rα2, a long-established target in glioblastoma that has previously been explored preclinically and clinically for CAR-T therapy, including in a smaller phase 1 trial by Brown and Badie’s team at City of Hope. The therapy has been licensed by Mustang Bio, which is now developing the treatment.

The patients received increasingly higher doses as the trial progressed, administered via different routes: direct injection into the tumor, infusion into the cerebrospinal fluid or, for the final cohort, both. The patients could continue to receive CAR-T cell infusions as long as they still met the enrollment criteria and had CAR-Ts left to use.


Given that systemic administration of CAR-T therapy is linked to dangerous side effects in the brain—specifically, a set of symptoms called immune effector cell-associated neurotoxicity syndrome, or ICANS—the researchers initially worried that injecting CAR-Ts directly into the brain would go very wrong. 

“When we started this study in 2015, neurotoxicity and ICANS with CAR-T cell therapy was a really hot topic,” Brown said. For that reason, the first infusion was a “very low mouse dose,” she said. “We stayed up all night, because we weren’t sure what we would see,” Brown recalled. 

Surprisingly, there was no neurotoxicity in any of the patients. Some developed headaches and fevers for a few days, but that was the extent of their side effects, Badie said. That’s likely because, unlike the CD19 antigen targeted by other types of CAR-Ts, the IL13Rα2 receptor is expressed almost exclusively by the tumor and some small populations of inflammatory cells—not on neurons or cranial nerves, Badie explained. On top of that, research by Stanford and City of Hope has suggested that the adverse effects of CAR-T therapy for brain tumors might be slightly different than treating other types of tumors, Brown noted.

“I don’t think we completely understand the similarities and differences,” she said. “That’s still an area to be explored.”

Encouraging signs

The overall median survival duration for all patients in the trial was eight months. Twenty-nine had stable disease for at least two months after the therapy was administered. For patients who received CAR-Ts both in their tumors and their cerebrospinal fluid at a predetermined maximum feasible dose, the median survival was 10.2 months, higher than the expected six-month survival rate for patients treated for recurrent glioblastoma.

The CAR-Ts given to this cohort were also manufactured with an optimized manufacturing process, Brown told Fierce in a follow-up email.

One of the patients, a 58-year-old man, remains disease-free five and a half years after his treatment began. Though he has some residual complications as a result of the tumor and prior treatment—he’s paralyzed on his left side—he is able to spend time with his children and grandchild.

Even for the rest of the subjects in the trial, all of whom ultimately passed away, “it was encouraging to see that [recurrence] was delayed” in so many of them, Badie said, adding that the tumor usually comes back within a month after surgery. “That means there was really good local control initially.” 


Insights from outliers

Brown and Badie’s team is working to figure out what exactly made those outliers respond relatively well to treatment, a project made more robust by their ability to take samples of the tumor and its surrounding microenvironment in patients whose treatment was infused directly into the cancer. One element that stands out is that the patients who fared best had a different immune cell profile than the others, suggesting there’s something important about the interaction between host immunity and CAR-T cell therapy, Brown said. 

“When we looked at their [pretreatment] tumors, we found that the immune contexture actually correlated with a better response,” she explained. Specifically, patients with higher numbers of a certain type of T cells—CD3 T cells—had better responses. 

“It’s really driving this working hypothesis that in solid tumors, the design of the cell therapy and targets are important, but it’s also important to look at how it engages with host immunity,” Brown said. “We think this study provides initial insights into that.”

The team has data from early-stage trials on single-target CAR-T therapies for glioblastoma that act on other antigens, too, though details on those studies haven’t yet been published. Meanwhile, based on the results of the newly published trial, the team is building bispecific CARs that go after more than one antigen at a time as well as designing a study that combines multiple CARs in a single treatment. 

“We need to make an impact on this disease for these patients,” Brown said. “So we hope that this is one step in the right direction.”