A team of researchers from academia and industry has developed a method of engineering T cells to target a mutant protein found in a particularly aggressive form of glioblastoma, paving the way for a phase I clinical trial.

A team of researchers successfully engineered T cells to express a chimeric antigen receptor (CAR) that targets a mutation associated with a particularly aggressive form of glioblastoma, laying the foundation for a phase I clinical trial.

In a preclinical study, researchers redirected humanized CAR T cells to bind to the EGFR variant III mutation (EGFRvIII), a common variant of EGFR in human tumors that occurs in about 30% of patients with glioblastoma and is associated with poorer prognosis. Investigators designed CAR T cells using humanized single-chain variable fragments (scFv) that showed specificity for EGFRvIII over wild-type EGFR. In mice, the EGFRvIII CAR T cells successfully controlled tumor growth, and deeper regression was observed when they were combined with temozolomide chemotherapy, a standard treatment for glioblastoma. The results are published in Science Translational Medicine.

“Studying genetically engineered T cells is a hot area because of the tremendous success there has been with leukemia,” says the study's senior author Marcela Maus, MD, PhD, assistant professor of hematology/oncology at the University of Pennsylvania Abramson Cancer Center in Philadelphia. “This phase I trial will be the first time that we are trying this particular method of engineering T cells to target glioblastoma in humans.”

Although CAR T-cell therapy has shown promise for treating blood cancers, solid tumors have been more challenging because many targetable surface antigens on solid tumors are also expressed in normal cells. However, the authors noted that EGFRvIII might be an ideal CAR target because it is specific to malignant cells and plays a critical role in maintaining oncogenesis.

The researchers, including scientists from Novartis Institutes for BioMedical Research, tested a panel of scFvs in silico and in vitro in order to determine the degree of specificity to EGFRvIII and cross specificity with wild-type EGFR. They then further tested the lead humanized CAR for safety in mouse models grafted with normal EGFR-expressing human skin.

The investigators also tested the lead humanized CAR for efficacy in three mouse models implanted with human glioblastoma cell lines, says Maus. The CAR T cells controlled or shrank most tumors whether the tumors were intracranial or under the skin. Administered intravenously, the engineered T cells controlled the tumor most effectively when given in combination with temozolomide.

The study is the basis for a new phase I trial, based at Penn and the University of California, San Francisco, that is currently enrolling up to 12 adults with EGFRvIII-positive glioblastoma who have either relapsed after standard therapy or have residual disease following surgery, says Maus. Investigators will remove patients' own T cells and reprogram them to target EGFRvIII-expressing tumor cells when they are returned to the patients via intravenous injection.