Integrated analysis of hub gene expression in multiple myeloma.

Purpose To explore the expression and clinical significance of factors associated with multiple myeloma (MM) and identify new diagnostic markers. Methods Two gene expression array data sets (GSE6477 and GSE5900) were downloaded and differentially expressed genes (DEGs) in bone marrow from patients with MM and healthy donors analyzed. Kyoto Encyclopedia of Genes and Genomes pathway enrichment and Gene Ontology annotation of DEGs was conducted and a protein-protein interaction network generated. Plasma and bone marrow samples from patients with MM were analyzed for cytokine expression by ELISA and correlations between cytokine levels and clinical indicators evaluated. Results Of 908 DEGs, 416 were up-regulated and 492 down-regulated. Further, 161 proteins pairs and 21 nodes were detected, and eight hub genes (CXCL2, CXCL8, CXCL12, ELANE, LCN2, CX3CL1, CCL13, and CCL27) screened out. Expression levels of CXCL8, CXCL2, CXCL12, LCN2, and CCL13 were low in CD138+ plasma cells, and expression levels of the eight cytokines differed significantly in peripheral blood plasma from patients with MM and healthy controls. ROC curve analysis determined optimal diagnostic thresholds determined for: CCL27 (189 ng/mL), CXCL2 (313 ng/L), CX3CL1 (132 ng/L), CCL13 (235 pg/mL), CXCL8 (884 ng/L), ELANE (50 µg/L), LCN2 (8 µg/L), and CXCL12 (2525 pg/mL). Conclusions CX3CL1, CCL13, CXCL8, and CXCL12 levels were positively correlated with those of hemoglobin and β2 microglobulin (β2-MG); CCL27 and CXCL2 with β2-MG; and CCL13 and ELANE with white blood cell count and age, respectively. CCL27, CXCL2, and β2-MG levels were associated with MM incidence.