Prognostic biomarkers for primary Sjögren's syndrome-associated non-Hodgkin's lymphoma.

person Kevin Sheng-Kai Ma Department of Life Science, National Taiwan University, Taipei City, Taiwan info_outline Kevin Sheng-Kai Ma, Angel Alfonso Velarde Lopez

Authors person Kevin Sheng-Kai Ma Department of Life Science, National Taiwan University, Taipei City, Taiwan info_outline Kevin Sheng-Kai Ma, Angel Alfonso Velarde Lopez Organizations Department of Life Science, National Taiwan University, Taipei City, Taiwan, Center of Global Health, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA Abstract Disclosures Research Funding No funding received None Background: Bidirectional relationship between non-Hodgkin’s lymphoma (NHL) and Sjögren’s syndrome (SS) has been recognized in large-scale cohort studies. We identified significant differentially expressed genes for SS-associated NHLs, and evaluated whether these genes could serve as a prognostic biomarker of diffuse large B-cell lymphoma (DLBCL). Methods: RNA-sequencing data from whole-transcriptome expression profiling were compared in a canonical pathway analysis using z-score and p-value visualization to identify significant differentially expressed genes and underlying cellular mechanisms for SS-associated NHLs including Burkitt lymphoma (BL), DLBCL, follicular lymphoma (FL), mantle cell lymphoma (MCL), and marginal zone lymphoma (MZL). Collected biopsies included lymphoma biopsies for BL (n = 59), DLBCL (n = 88), FL (n = 65), MCL (n = 43), and MZL (n = 23), sorted B cells from healthy donors (n = 10), as well as parotid gland tissues for SS (n = 17). RNA-seq data with P values less than 0.05 or -log 10 (P) larger than 1.3 were considered significant, and positive z-scores indicated up-regulation while negative z-scores indicated down-regulation. After screening potential biomarkers through canonical pathway analysis, significant differentially expressed genes were subjected to survival analysis with genomic and clinical data of 420 DLBCL patients from The Cancer Genome Atlas (TCGA) using a log-rank test. Area under the ROC curve (AUC) was derived to validate the predictive ability of the survival model. Results: Among all genes involved in SS-associated NHLs, 16 genes were significantly differentially expressed in all types of NHLs. These genes included CXCL13, SEPP1, COL1A1, RGS13, IGFBP7, COL3A1, SPARCL1, GABBR1, SLC40A1, CXCL14, CXCL12, CXCL9, CCL19, VCAM1, C3, and CLU. The fold changes of the 16 genes for DLBCL were 465.67, 209.68, 436.94, 366.94, 189.59, 456.09, 195.37, 152.05, 90.29, 26.29, 215.18, 72.06, 51.09, 91.91, 74.70, and 20.88, respectively. Among all SS-associated pathways underlying NHL pathogenesis, the endocannabinoid system-driven cancer inhibition pathways were most significantly altered in SS-associated NHLs. The z-scores of the cancer inhibition pathway were -0.14 for BL (-log 10 (P) = 4.54), -1.298 for DLBCL (-log 10 (P) = 3.00), -0.33 for FL (-log 10 (P) = 4.67), -0.87 for MCL (-log 10 (P) = 3.77), and -2.61 for MZL (-log 10 (P) = 4.31). Survival analysis revealed that together the 16 genes may serve as a prognostic biomarker of DLBCL (hazard ratio = 3.57, 95% CI = 2.53–5.03), which was statistically significant (log rank test P < 0.0001) and reliable (AUC = 0.897). Conclusions: Significant differentially expressed genes underlying the pathogenesis for SS-associated NHLs may be used as a biomarker for predicting the survival rate of DLBCL. Further studies are warranted to elucidate the functional association between these differentially expressed genes.