Adaptive Immune Receptor Recovery Distinctions in Burkitt's Lymphoma Correlate with Better Survival

Introduction: The adaptive immune response, as characterized by productive immune receptor (IR) recombination reads from sequencing data, has been shown to represent survival and tumor microenvironment differences in many disease states. Further, matching human leukocyte antigen (HLA) alleles with specific T-cell receptors has been shown to represent outcome distinctions in autoimmune diseases, infectious diseases, and cancer. Burkitt lymphoma (BL) is well-known to be caused by Epstein-Barr virus. However, few immunogenomics analyses of endemic BL have been performed.

Methods: We utilized the Cancer Genome Characterization Initiative - Burkitt Lymphoma Genome Sequencing Project (CGCI-BLGSP) data available at the Genomic Data Commons website, focusing only on endemic BL cases, represented by 105 RNA-seq files taken from primary tumor that represented 105 cases. Variables provided in this study included gender, race, age of diagnosis, days to last follow-up, vital status, and Ann Arbor pathologic stage. These files were mined for HLA alleles utilizing xHLA (Xie et al. 2017). The RNAseq files were also mined for productive adaptive IR recombination reads and associated V and J-IDs, as extensively described in Chobrutskiy et al. 2020, which yielded reads for TRA, TRB, TRD, TRG, IGH, IGK, and IGL.

We then correlated these adaptive immune receptor recombination reads with the phenotypic data noted above. This was first done by identifying whether the presence of at least one recombination read correlated with better survival. Then, the mean recombination read count of each of the seven receptors was assessed across Ann Arbor pathologic stage utilizing the analysis of variance (ANOVA) test. Next, we assessed specific HLA alleles, VJ gene segment usage, and HLA-VJ combinations to determine whether there were survival distinctions represented by these features. These analyses were done with an original python script and confirmed with IBM SPSS v29.

Results: 697,090 adaptive IR recombination reads were recovered across the 99 samples with staging data. Greater than 97% of all samples had at least one TRA, TRB, IGH, IGK, and IGL recombination read. Regarding TRG and TRD, we found that for those cases with a TRG read, overall survival (OS) was higher than those without (OS median with TRG vs. without: 387 days vs. 11 days, respectively, log-rank p = 0.002). Further, we identified decreases in TRB, TRA, and TRG recombination read count as tumor stage progressed (ANOVA p-value = 0.006, 0.008, 0.02, respectively). For example, we found that the mean TRB combination read count for Stage I disease was 248.4 ± 62.2, which fell to 144.1, 110.0, and 105.5 mean TRB recombination read counts for stage II, III, and IV disease, respectively. We also identified HLA-TCR VJ combinations which were associated with survival where the individual HLA allele or V or J gene segment was not significantly associated with survival, i.e., thereby revealing a potential interaction effect for the HLA alleles and TCR V or J gene segments. An example of this was DQB1*05:01+TRBV3-2, where the 23 cases with this HLA-VJ combination did not reach the median OS, whereas the median OS of people without this combination was 215 days (log-rank p = 0.01; Individual component p-values are noted as follows - HLA-DQB1*05:01 log-rank p = 0.17, TRBV3-2 log-rank p = 0.03).

Conclusions: These results point towards the importance of the T-cell response in early control of the tumor and reflects tumor microenvironment changes that occur early in Burkitt's lymphoma which could assist in prognosis. Further, this study has identified specific successful responses in endemic BL disease which one could aim to replicate with tumor vaccines, T-cell receptor mimic antibodies, and other immunotherapy avenues.

No relevant conflicts of interest to declare.