Summary
A bioRxiv preprint reports that plasma mass spectrometry could detect some tumor-associated peptides, but neoepitope searches were constrained by limited overlap between cfDNA and tumor mutations and by MS sensitivity. Tissue remained the stronger reference for identifying tumor mutations and neoepitopes.
A bioRxiv preprint reports that searching blood plasma for cancer neoepitopes is possible in principle, but the approach was constrained by which tumor mutations could be detected in circulating DNA and by the sensitivity of mass spectrometry (MS). In the study, plasma yielded tumor-associated peptides, while tissue provided a more complete reference for identifying mutation-derived peptides.
Neoepitopes are peptide fragments produced by tumor-specific genetic changes. Cells display many such fragments bound to human leukocyte antigen (HLA) molecules, which help immune cells recognize them. The researchers tested whether those HLA-bound peptides could be found in plasma and matched to each patient’s mutations.
How the researchers searched plasma
The team analysed plasma HLA ligands from four cancer patients using MS, a technique that identifies molecules by measuring their mass and related properties. They compared the peptide results with patient-specific mutation sets, or mutanomes, derived either from cell-free DNA (cfDNA) in plasma or from tumor genomic DNA (tgDNA) in tissue. Matched tumor tissue and healthy-donor plasma served as controls. Candidate neoepitopes were checked with synthetic peptides, and immune responses were assessed using IFN-γ ELISpot and intracellular cytokine staining.
The cfDNA-based search did not identify plasma neoepitopes. The authors report only 15% overlap between the cfDNA and tumor-tissue mutanomes, limiting how many tumor mutations were available to guide the peptide search. When the tissue-derived mutanome was used as a complementary reference, the researchers found two neoepitopes in one patient’s plasma. Both were also found in that patient’s tumor tissue, but their plasma signals were near the MS detection limit.
What the results show
The plasma analysis was more successful for HLA ligands not identified as mutation-derived. The wild-type peptide profiles in plasma and tissue overlapped by 58%, and 91% of plasma HLA ligands were also found in tissue. Of 15 tumor-associated HLA ligands detected in plasma, 13 were recovered from the matching tissue.
Tissue analysis also identified three neoepitopes that were not found in plasma. Two of these prompted antigen-specific T-cell responses when tested with peripheral blood cells from healthy donors. That result describes responses in the laboratory assay, not a response observed in the patients.
Taken together, the findings suggest plasma can support profiling of some HLA-bound peptides, including tumor-associated ones. For neoepitope searches, however, the plasma route depends on detecting the relevant tumor mutations in cfDNA and then measuring the corresponding peptides. In this small study, tissue remained the more useful reference for both steps. The authors propose that improved detection methods could make minimally invasive plasma analysis more useful in future; the reported results are from a preprint.