Could CAR-T treatments work in glioblastoma?

The news last week that Arizona Senator John McCain has been diagnosed with glioblastoma brought stark reminders of the challenges presented by this particular type of brain tumor: The five-year survival rate for glioblastoma patients over age 55 is just 4%, according to the American Cancer Society. Now researchers at the University of Pennsylvania have generated evidence that combining different types of immuno-oncology treatments be key for engaging the immune system to fight glioblastoma—data they outlined one day before McCain’s diagnosis was revealed.

Penn scientists have spent the last two years exploring the potential of chimeric antigen receptor T (CAR-T) cell therapy in glioblastoma. The technology, which is similar to the CAR-T treatment for blood cancer that Novartis has been developing with Penn’s Perelman School of Medicine, involves taking T cells from patients and engineering them to recognize a protein on tumor cells. In the case of glioblastoma, that protein is EGFRvIII, an antigen that has been shown to drive tumor growth in about one-third of patients.

RELATED: Racing toward a historic CAR-T approval, Novartis wins unanimous FDA panel backing

In a newly published study testing the technology in 10 patients, the researchers report that the CAR-T therapy seems to cause the tumor’s microenvironment to be more immunosuppressive, meaning it prompts the tumor to put up even more barriers that don’t allow the immune system to fight the cancer. This is a major obstacle, to be sure, but it’s one that may be able to be overcome with existing drugs, such as “checkpoint inhibitors” that block immune suppressors like PD-1, they suggest in a press release. The study was published in the journal Science Translational Medicine.

The trial involved seven patients who underwent surgery and three who did not. Within two weeks of receiving the engineered CAR-T cells, the researchers saw signs that they were active in the tumors of four patients who underwent surgery immediately after the treatment.

But there were also signs of immunosuppression in those patients. What’s more, the researchers observed that EGFRvIII expression was widely variable across all the patients in the study. And even though the engineered T cells were detectable in the blood of all 10 patients in the first month after they received them, the number of cells began to decline after two weeks, according to the release.

"This trial showed that there is a need to target additional antigens in glioblastoma, as well as overcome the immunosuppressive environment that the CAR T cells encountered in the tumor," said co-author Marcela Maus, M.D., Ph.D., a former faculty member at Penn who is now director of cellular immunotherapy at the Massachusetts General Hospital Cancer Center. Combining the cells—which were well tolerated in the trial—with checkpoint inhibitors may be an option worth exploring in the future, the researchers believe.

CAR-T technology has generated all of its promising results to date in blood cancers. Just two weeks ago, Novartis got a unanimous thumbs-up from an advisory panel to the FDA on tisagenlecleucel-T, its CAR-T treatment for relapsed/refractory B-cell acute lymphoblastic leukemia. The FDA is set to make its final ruling in October. Kite and Juno are working on similar treatments in blood cancers.

Solid tumors have been more difficult to tackle with the new technology, however. Novartis has expressed its support for moving the technology into solid tumors; the company told FierceBiotech in the spring that it’s moving quickly on a CAR-T to treat ovarian cancer. Penn's glioblastoma trial was sponsored by Novartis.

RELATED: Novartis drug development chief outlines CAR-T research commitment

Other companies are also looking at CAR-T options for solid tumors. They include Belgian biotech Celyad, which is assessing its CAR-T treatment, NKR-2, in five solid tumor types.

The Penn researchers were disappointed that their CAR-T treatment didn’t produce meaningful tumor shrinkage in the glioblastoma patients in their trial, although one patient did show stable disease at 18 months and has maintained that response. Still, they believe their evidence shows there may be potential for perfecting CAR-T technology in glioblastoma. The trial “demonstrates the limitations: the antigen heterogeneity and the [tumor microenvironment] fight back,” said lead author Donald M. O'Rourke, M.D., associate professor of neurosurgery at Penn. “This points us and the rest of the field in the right direction."