Contrary to what might be expected, chimeric antigen receptor (CAR) T cells with stronger signaling capabilities were less effective against lymphoma cells in a mouse model, investigators reported.
Intracellular signaling strength was a key determinant of T cell fate in the study, which was published in the journal Science Signaling.
By contrast, CAR signaling pathways could not be predicted solely by the costimulatory domains used to construct the receptor, investigators said.
Based on those findings, tailoring CAR design based on signal strength might improve the efficacy and reduce the toxicity of CAR T-cell therapy, according to Alexander Salter, an MD/PhD student at Fred Hutchinson Cancer Research Center, Seattle, Wash.
In a press conference, Mr. Salter described results of the study, which used mass spectrometry to evaluate CARs encoding CD28 or 4-1BB costimulatory domains in primary human T cells.
While CARs with CD28 domains elicited more robust intracellular signaling than those with 4-1BB domains, there was considerable overlap in activation of T cell signaling pathways, Mr. Salter said.
That overlap was somewhat surprising, according to Mr. Salter, since researchers have generally assumed that CARs with CD28 and 4-1BB costimulatory domains will primarily signal through those respective pathways.
“No matter what costimulatory domain was encoded by the receptor, both CARs… activated both CD28 and 41BB signaling pathways,” Mr. Salter said.
The major determinant of efficacy in the study turned out to be not the domain used to construct the receptor, but the speed and strength of signaling, he added. In particular, the CARs that evoked stronger signals also had increased T cell dysfunction, decreasing their potency in the mouse lymphoma model.
The T cells with a CD28 CAR had very strong initial antitumor function that quickly waned in the mouse model of lymphoma; by contrast, the “slower burning” 4-1BB CAR signal led to T cells that better retained their function in vivo and were associated with longer median survival in the model, he said.
Those findings suggest tailoring CAR design based on signal strength may improve clinical efficacy and reduce toxicity.
As part of the study, Mr. Salter and his co-investigators were able to modify the CAR CD28 domain to make the signaling of the CD28 CARs less intense. “This is a modification that we think should be considered in future CAR design,” Mr. Salter said.
While the alterations in the CD28 signaling domain were able to reduce levels of cytokines produced by T cells, the study was primarily designed to look at the efficacy, noted Stanley Riddell, MD, scientific director of the Immunotherapy Integrated Research Center at Fred Hutchinson Cancer Research Center.
“Our models were not set up to address the question of toxicity, so we can’t directly say this would translate to what we would see in patients,” Dr. Riddell said during the press conference. “But I think we gleaned a lot of insights as to why cytokines are produced at greater or lesser levels with various CAR designs, and insights as to how to redesign these receptors to lower the levels of cytokines they make without compromising their ability to kill.”
Dr. Riddell is a founder, shareholder, and scientific advisor of Juno Therapeutics, and together with Mr. Salter, he has filed a patent application on the use of mutant CD28 CARs for cellular therapy. Co-author Raphael Gottardo, PhD, also with Fred Hutchinson Cancer Research Center, is a consultant for Juno Therapeutics. No other competing interests were reported.
SOURCE: Salter AI et al., Sci Signal. 2018 Aug 21;11. pii:eaat6753.