Abstract
CAR T cells directed against CSPG4 limit the growth of brain tumors in cultured neurospheres and glioma xenograft mouse models, with no signs of immune escape owing to loss of antigen expression.
A new target for chimeric antigen receptor (CAR) T-cell therapies for glioblastoma could help overcome some of the problems of tumor escape and heterogeneous expression that have limited the effectiveness of other tumor-associated antigens included in clinical CAR T-cell candidates for this deadly brain cancer.
According to a preclinical study, T cells transduced to express a transmembrane signaling protein called chondroitin sulfate proteoglycan 4 (CSPG4) can successfully blunt the growth of brain tumors in cultured neurospheres and in glioma xenograft models, with no signs of immune evasion owing to loss of antigen expression (Sci Transl Med 2018;10:eaao2731).
Experts who study CAR T-cell therapies for glioblastoma have been buoyed by the findings. “This antigen definitely could have great potential,” says Irina Balyasnikova, PhD, of Northwestern University Feinberg School of Medicine in Chicago, IL. However, as Donald O'Rourke, MD, of the University of Pennsylvania in Philadelphia, points out: “There's just not a lot known about this molecule in glioblastoma.”
Plus, adds O'Rourke, who has tested an EGFRvIII-directed CAR T-cell therapy for brain cancer and is planning another trial of the treatment in combination with a checkpoint inhibitor, “extrapolating from mouse CAR T data to humans is a big leap” (Sci Transl Med 2017;9:eaaa0984).
Gianpietro Dotti, MD, of the University of North Carolina Lineberger Comprehensive Cancer Center in Chapel Hill, chose to study anti-CSPG4 CAR T-cell therapy in glioblastoma because of its established role in arresting the progression of many solid tumors and its restricted distribution in normal tissues. In 46 brain tumor specimens, he and his colleagues found that CSPG4 was highly expressed in 31 and undetectable in only one. Consistent with other studies, the researchers also found that CSPG4 expression provided a prognostic indicator of shorter survival.
Yet even in those glioblastoma cells with low to moderate CSPG4 expression, Dotti's CAR T cells proved effective—seemingly because microglia in the tumor microenvironment secreted TNFα, which induced cancer cells to express more CSPG4 as a defensive strategy. That natural induction of CSPG4 could explain why the CAR T cells effectively controlled tumor growth and promoted survival in the 10 mice tested, six of which were cancer-free more than 6 months later.
According to Dotti, antigen-activated CAR T cells themselves release TNFα in the tumor microenvironment—as shown in February by Balyasnikova's group—which could prompt more antigen expression at the site of CAR T-cell action (Mol Ther 2018 Feb 7 [Epub ahead of print]). “You create a sort of loop,” says Dotti, who aims to test the CSPG4-directed CAR T-cell therapy in patients with glioblastoma next year.
Whether TNFα-mediated induction truly makes CSPG4 expression uniform across the tumor remains to be seen, though. “All will hinge on this claim that you don't have antigen loss,” says Martha Chekenya, PhD, of Bergen University in Norway, and she worries that the model tested—allogenic human CAR T cells in xenograft mice—might not be representative of the autologous strategy that would be deployed clinically.
Balyasnikova also has concerns about off-target effects. Dotti's team found no CSPG4 expression in vessels of normal brain tissue, but Balyasnikova wonders about other healthy cells. “We have to proceed with caution,” she says, “because this protein has been shown to be expressed in certain progenitor cells in the developing and adult brain.” – Elie Dolgin
For more news on cancer research, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/CDNews.