In an effort to discover new therapeutically useful neoantigen targets, Khodadoust et al. used new tools and found evidence for an “old” idea. In this study, the team took an integrated genomic and proteomic approach to mantle cell lymphoma from 17 untreated patients to explore the peptides presented by MHC class I and class II molecules, which are highly expressed on the surface of this tumor type. What they uncovered was evidence for renewed hope in the possible efficacy of idiotype vaccination targeting the patient-specific lymphoma immunoglobulin (Ig).
Whole exome sequencing and targeted sequencing of the rearranged lymphoma Ig revealed between 13 and 175 somatic mutations per patient, most of which were unique to each patient. Nearly half (46%) of the expressed genes bearing coding mutations had at least one peptide presented on a MHC class I or II molecule. This suggests active proteasomal processing of these mutated proteins. However, none of these presented peptides were derived from mutated regions of any of these genes, with the exception of Ig neoantigen peptides. Ultimately, 67 neoantigenic epitopes were recovered, all of which were from genes coding for Ig heavy- or light-chain variable regions.
Peptides from the lymphoma Ig were presented on MHC class I and class II molecules in all 17 patients, but the antigen presentation patterns differed between the two MHC classes. Of the more than 200 Ig-derived MHC class I peptides detected by mass spectrometry, the vast majority were from the constant domains, while only 21 were from the heavy- and light-chain variable regions, and only one of them was considered “neoantigenic.” On the other hand, of the 499 Ig-derived MHC class II peptides detected, 147 were found in variable regions and 66 of those were neoantigenic.
As the authors point out, this unexpected bias against MHC class I presentation (even though strong binders from the variable region had been predicted for MHC class I) suggests immunoediting, but the exact mechanism is yet to be determined. Conversely, the rather frequent presentation of variable-region peptides by MHC class II suggests that recognition by CD4+ cells does not inhibit, but may actually promote tumor progression. Supporting this theory, Ig neoantigen-specific T cells have been shown to have a Th2/Th17-like phenotype.
To investigate whether Ig-neoantigen-specific CD4+ T cells can be developed as cytotoxic therapy, the researchers analyzed peripheral blood samples from three patients in search of T cells specific to these Ig-derived neoantigens and found them in one of three samples tested. Autologous tumor cell vaccination further induced the expansion of these neoantigen-specific T cell clones in this patient. Isolated CD4+ T cells from another patient were stimulated and expanded with autologous Ig neoepitopes and demonstrated the capability to specifically recognize and kill the patient’s own lymphoma cells.
This data suggests that immunoglobulin neoantigens could potentially serve as targets for lymphoma immunotherapy. Previous clinical trials had high hopes for the success of idiotype vaccines, but saw disappointing results. These new results suggest that MHC presentation of epitopes derived from variable regions is not universal across individual lymphomas, raising the possibility that this therapy may only be effective on patients that do actually present Ig neoepitopes.
by Lauren Hitchings