Despite the hype around adoptive T-cell therapies, it’s still a challenge to find therapeutic targets that are consistent across tumor types and aren’t expressed in healthy cells. Now, a team of scientists from the University of Pennsylvania working with immunotherapy drug development biotech Immatics has identified an antigen component that checks both of those boxes.
In a study published Wednesday in Science Translational Medicine, the researchers reported their finding of a protein that when targeted through adoptive T-cell receptor therapy slowed cancer growth in mice. The protein was highly prevalent in tumor samples from patients with 11 different types of solid tumors but was rarely present in healthy cells, making it a prime potential target for T-cell therapies. Now, Immatics plans to pursue an investigational new drug application with the FDA in hope of launching a phase I clinical trial.
The team’s goal for its research was to find antigens that were present in different tumor types and could safely be targeted in T-cell therapies without harming other cells. They started by analyzing sets of normal and tumor tissue samples with Immatics’ XPRESIDENT discovery platform, which uses mass spectrometry to examine and compare hundreds of peptides across the immunopeptidome, or the ensembles of peptides associated with human leukocyte antigens. The XPRESIDENT analysis was paired with the company’s XCUBE bioinformatics approach, using artificial intelligence to quantify peptide copy numbers in each cell.
“We detect and quantify what’s there and what makes the difference between cancer and normal,” Toni Weinschenk, co-corresponding author and Chief Innovation Officer at Immatics, explained in an email. In other studies, immune system peptides have been analyzed through binding predictions. But with Immatics' discovery platform and bioinformatics tools at their disposal, the team could take a quantitative approach. “We don’t have to predict,” Weinschenk said.
The scientists pinned down an antigen component that was highly expressed in cancer cells from multiple types of tumors but rarely in human tissue: an epitope encoded by exon 6 of the gene collagen type VI α-3, or COL6A3, which is produced from a tumor-specific alternative splicing event that rarely happens in healthy cells. The COL6A3 protein is a component of type VI collagen, which is present in connective tissue throughout the body, but exon 6 is expressed only in the stroma cells of the tumor microenvironment.
“Protein collagen type VI alpha-3 (COL6A3) is abundantly expressed and therefore not selective enough to be targeted by adoptive cellular therapy,” Weinschenk said. “The tumor selectivity is based on a tumor-associated splicing event, which retains the target-encoding exon 6.”
With a target in their sights, the researchers set out to develop T-cell receptors, or TCRs, that would recognize the epitope without attacking other peptides. Again using data from Immatics' discovery platform to compare reactivity between tumor and normal cells, the team developed high-affinity TCR-T cells and injected them into mice that had been grafted with human leukemia cells. The cancer growth slowed, and the mice didn’t experience serious side effects.
To Weinschenk, the results show how the company's XPRESIDENT approach can be used not only to identify potential therapeutic targets but also to develop TCRs with strong safety profiles.
Next, Immatics plans to conduct more studies on the COL6A3 epitope to support its IND application with the FDA, though the company didn’t offer a timeline. If it were to move forward with a phase I trial, the new epitope would be the ninth product in the company’s pipeline on top of its programs with Genmab, GSK and Bristol Myers Squibb.