Too much like us? Researchers uncover key reason a promising cancer therapy is often unsuccessful

Author: Deanna Csomo Ferrell

Neoantigens, which are molecules found on tumor cells that incorporate mutations, help our immune system fight cancers and could be the most promising components of future cancer vaccines—if only scientists knew with a high degree of certainty which neoantigens work.

A collaboration of scientists led by Brian Baker, Coleman Professor of Life Sciences in the Department of Chemistry and Biochemistry at the University of Notre Dame, successfully identified features of different neoantigens and discovered why most don’t yield anti-cancer immune responses even when they are expected to: They look and act too much like us.

Results from his recent paper were published in the Proceedings of the National Academy of Sciences.

“One of the routes that we've taken this paper, which surprisingly much of the field has ignored, comes back to basic biology: Your immune system is trained to ignore you,” said Baker, who is affiliated with the Harper Cancer Research Institute at the University of Notre Dame. “So what we did is a simple thing: We started with the basic question of, what is a neoantigen’s difference from ‘self’?”

Human T cells are part of our immune system, and attack cells infected by intruders like viruses or bacteria. T cells recognize foreign antigens found on diseased cell surfaces—the “non-self.” They also will attack transplanted organs and tissues because those are also non-self. Cancerous cells are also diseased, but the neoantigens on cancer cells are often too much like ourselves to alert our T cells to take action against them. It’s one reason that, once established, cancer can spread unchecked. However, successful neoantigens can lead to natural anti-cancer immune responses, or serve as the basis for future cancer vaccines.

To determine which neoantigens that T cells might successfully target, clinicians sequence a patient’s tumor genome and compare that to the wild type genome, which is the genome as it appears in nature, Baker said. They might find many mutations and use a variety of tools to predict which neoantigens are the most optimal. Unfortunately, T cells target very few of the many hundreds to thousands of neoantigens that are present. The prediction tools in general perform poorly at identifying these, Baker said, so most neoantigens that are identified don’t lead to tumor killing. 

“That’s where the field has been stuck,” he said. “How do you go from these big lists of all the mutations that are present to find those rare neoantigens that are actually any good at eliciting anti-tumor immune responses?” he said.

To reach a potential solution, Baker’s lab examined different neoantigens from mouse models of cancer. Prior work had identified all neoantigens from the mouse tumors, and the researchers decided to study three in more detail, including an inactive neoantigen, a moderately active neoantigen, and a highly active neoantigen that protected against cancer growth in all mice studied. Previous work had identified these, but researchers didn't know how or why the different antigens acted the way they did.

“We learned a variety of things, but the underlying theme was that neoantigen activity was best understood when we compared each neoantigen to its wild-type, self counterpart. The active neoantigens were different from self in various ways, whereas the inactive neoantigen, despite containing a mutation, was almost indistinguishable from self,” Baker said. “So it's the kind of mutation, and how it distinguishes the neoantigen from its self counterpart that seems to play a large role.”

Additionally, neoantigen activity was associated with differences that are linked to T cell recognition.

“Many current algorithms, or prediction approaches, would have selected the inactive neoantigen,” Baker said. “But again, most approaches are not considering something: These neoantigens are not different enough from self.”

The reason this hadn’t been discovered before is because neoantigens from cancer cells work differently than how antigens from viruses or other pathogens work, Baker said. T cells are not tolerized to viruses like they are to our own tissues and cells, of which tumors are derived from.

“These results will have an immediate impact on building and refining new classes of immunotherapies for cancer patients. In particular, these data will aid in prioritizing which neoantigens to include in personalized cancer vaccines,” said Dr. Christopher A. Klebanoff, associate member and attending physician, Memorial Sloan Kettering Cancer Center in New York City and one of the world’s experts in the field of cancer immunotherapy.

Next steps would be to test the findings in more detail. If supported, the results could be incorporated into new prediction tools, and impact the decisions on which neoantigens that clinicians would use to develop a personalized medicine plan, Baker said.

In addition to Baker, other researchers involved in the work include; postdoctoral fellows Jean Custodio and Chad Brambley; research assistant professor Cory Ayres; graduate students Tatiana Rosales and Grant Keller; and undergraduate students Alyssa Arbuiso and Lauren Landau. The other researcher, Dr. Pramod Srivastava, collaborated from the University of Connecticut, where he is a professor in the Department of Immunology and the Eversource Energy Chair in Experimental Oncology.

Originally published by Deanna Csomo Ferrell at on December 12, 2023.