Immunosuppressed Tumor Microenvironment in <i>MYC</i>-Rearranged High-Grade B-Cell Lymphomas Compared to Diffuse Large B-Cell Lymphomas, Not Otherwise Specified

Background The MYC oncogene exerts numerous tumor-promoting functions and may enable anti-tumor immune escape by contributing to the remodeling of the tumor microenvironment (TME, PMID 33081056). In light of the development of novel immunotherapies for lymphomas, we compared the TME composition of MYC-rearranged (MYC-R) high-grade B-cell lymphomas (HGBL) to ‘MYC negative’ diffuse large B-cell lymphomas, not otherwise specified (DLBCL) at DNA, RNA and protein levels. Methods Biopsy specimens from HGBL patients (HOVON-152 trial, NCT03620578) and DLBCL (HOVON-902 cohort, NCT04139252) were selected based on sample availability. MYC-R, BCL2-R and BCL6-R were determined using fluorescence in situ hybridization (FISH), using break-apart probes according to standard diagnostic lymphoma work up. We performed genomic profiling (targeted next-generation sequencing (tNGS), lon Torrent; n=36 HGBL/26 DLBCL) using the validated AmpliSeq BLYMFv2 panel, gene-expression profiling (GEP, NanoString; n=53/80) using the BLYMF777 probe set as previously described (PMID 34478526/35454765), and imaging mass cytometry (IMC, Hyperion; n=40/32) using a 41 antibody-marker panel for phenotyping to define the molecular landscape of tumors and determine the composition of their TME. Statistical differences between abundances were determined using the Fisher's exact test and Student's t-test. Results The majority of HGBL cases exhibited recurrent BLC2 rearrangements: 66% of cases in the genomic profiling and GEP cohorts and 62,5% in the IMC cohort were classified as ‘BCL2 double hit (DH)‘. BCL6 DH was observed in 11-15% of HGBL cases. Additionally, 21-22% of all HGBL cases had MYC, BCL2 and BCL6 rearrangements and were classified as triple hit. MYC-R were absent in all DLBCL cases. DNA: genomic profiling Consistent with previous literature, chromatin modifiers CREBBP and KMT2D were frequently mutated in HGBL as well as in DLBCL. Notably, HGBL had more mutations in TNFRSF14 (HVEM, p=0.007), BCL6 (p=0.01), and B-cell transcription factor IRF8 (p=0.04), while DLBCL showed higher frequencies of mutations in ZEB2 (p=0.01), KLHL6, GRHPR, IRF4 (all p=0.03) and tumor suppressor BTG2 (p=0.05), indicating distinct mutational profiles outside previously defined genetic subgroups (PMID 29713087/29641966). RNA: GEP signatures The majority of both HGBL and DLBCL cases were found to be GCB according to COO (Lymph2Cx). Despite mutational differences between the cohorts, GEP revealed a higher expression of genes related to the dark zone signature (DZsig, PMID 36302166/37552496) and MYC activity (PMID 27923830) in HGBL, while these genes were expressed at low levels in DLBCL. Protein: IMC phenotypes Using IMC we determined the composition of TME immune phenotypes in four cellular compartments: tumor, lymphoid, myeloid and stromal cells. In total, 666.809 cells (range 3-18K per patient) were obtained and 131 unique cellular phenotypes were defined. Clustering IMC phenotypes revealed three TME types: lymphoid immune-rich (mainly DLBCL cases), lymphoid immune-depleted (largely HGBL cases), and intermediate (including both DLBCL and HGBL). In detail, HGBL cases exhibited lower percentages of lymphoid cells (p&amp;lt;0.001) and were depleted for CD8+ T-cells, including naïve, effector memory and central memory CD8+ cytotoxic T-cells (all p&amp;lt;0.001), suggesting an immunosuppressed microenvironment. The amount of FOXP3+ T-cells (p=0.04), γδ-T-cells (p=0.03), and innate (like) lymphocytes (p&amp;lt;0.001) was also decreased in HGBL as compared to DLBCL. The TME of HGBL was significantly depleted in (proinflammatory) monocytes (p&amp;lt;0.04) and PD-L1+ macrophages (p&amp;lt;0.001), while there was a trend toward lower percentages of total myeloid cells in HGBL (p=0.07). Tumor and stromal cell percentages were similar between both groups (p=0.3). Conclusion This large comparison of MYC-R HGBL versus ‘MYC negative’ DLBCL cases using multi-omic analysis revealed a distinct TME in HGBL patients with increased frequencies of mutations in B-cell transcription factors, an abundance of discriminative MYC-related gene-expression signatures and, most strikingly, significantly lower infiltration of immune cells, especially cytotoxic T-cells, as compared to DLBCL. This suggests that MYC is linked to an immune-suppressed TME. These differences should be taken into account when developing T-cell engaging therapies for optimal efficacy.