Primary Sjögren’s syndrome (pSS) is a chronic systemic autoimmune disease characterized by exocrine gland damage and extraglandular involvements. To identify potential biomarkers for the early detection of pSS and to further investigate the potential roles of the biomarkers in the progression of pSS, our previous RNA sequencing data and four microarray data of salivary glands (SGs) were combined for integrative transcriptome analysis between pSS and non-pSS. Differential gene expression analysis, gene co-expression network analysis, and pathway analysis were conducted to detect hub genes, which were subsequently investigated in peripheral blood mononuclear cell (PBMC) and plasma. Correlation analysis, single-gene Gene Set Enrichment Analysis, and receiver operating characteristic (ROC) curve were applied to investigate the potential function of the hub genes and their classification capacity for pSS. A total of 51 common up-regulated genes were identified among different pSS cohorts. A key module was found to be the most closely linked to pSS, which was significantly associated with inflammation-related pathways. Seven overlapped hub genes (ICOS, SELL, CR2, BANK1, MS4A1, ZC3H12D, and CCR7) were identified, among which ICOS was demonstrated to be involved in most crucial immune pathways. ICOS was up-regulated not only in SGs but also in PBMC and plasma in pSS, and the expression of ICOS was closely associated with lymphocytic infiltration in SGs and disease activity of pSS patients. It showed a strong classification capacity with classic clinical index in SGs (ROC curve 0.9821) and significant distinct discrimination in PBMC (ROC curve 0.9107). These findings are expected to gain a further insight into the pathogenesis of pSS and provide a promising candidate for the early detection of pSS.