There were differentially expressed genes identified
There were 3151 differentially expressed genes identified from the transcriptome profiling data of 497 patients (Fig. 2). Among these differentially expressed genes, 2190 genes were found to be upregulated in the group with infiltrating lymphocytes ≥1%. As shown in Table 2, these genes were enriched in pathways related to immune response, hematopoiesis, cytokine production, and cell adhesion molecules. We found that several genes of interest showed upregulated expression, including PDCD1 (Spearman\'s rho = 0.229, P < 0.0001), CTLA4 (rho = 0.328, P < 0.0001), IDO1 (rho = 0.246, P < 0.0001), IDO2 (rho = 0.197, P < 0.0001), and FOXP3 (rho = 0.287, P < 0.0001). There was no difference in expression of CD274 (PD-L1) between the two groups.
In addition, we found that 961 genes were downregulated in the group with infiltrating lymphocytes ≥1%. Enriched pathways are listed in Table 3. Initial TCGA identified several thyroid-related genes constituting the thyroid differentiation score (TDS). Among the 13 mRNA expression levels included in the TDS, three were downregulated: DUOX1 (Spearman\'s rho = −0.177, P = 0.0001), DUOX2 (rho = −0.179, P = 0.0001), and TG (rho = −0.208, P < 0.0001). Notably, PVRL4 expression was upregulated (rho = 0.130, P = 0.004). The remaining 9 genes in the TDS showed no significant difference between the two groups, including SLC5A5 (sodium-iodide symporter) and TPO. For other thyroid-related genes, TSHR expression was significantly downregulated (rho = −0.176, P = 0.0001). The distributions are depicted in Fig. 3.
Discussion Chronic lymphocytic thyroiditis is the most common autoimmune inflammatory thyroid disease. Interestingly, the prevalence of chronic lymphocytic thyroiditis has increased over time. It is not uncommon to find thyroid cancer in patients who had a thyroidectomy for thyroiditis even when malignancy was not suspected preoperatively. However, the association between thyroiditis and thyroid cancer remains controversial. We and other groups have shown that lymphocytic thyroiditis is associated with an increased risk of thyroid cancer. Still others have argued that the presence of thyroiditis has a higher frequency of suspicious bch by way of fine-needle aspiration cytology, which results in a selection bias. In this study, we found that papillary thyroid cancer with infiltrating lymphocytes was associated with a better overall survival. Our finding is consistent with previous reports, suggesting that differentiated thyroid cancer with coexisting thyroiditis has a better long-term outcome. We demonstrated a higher frequency of lymph node metastasis in lymphocyte-infiltrating thyroid cancers. Although similar findings have been reported in other studies, a meta-analysis suggests that papillary cancer with coexisting thyroiditis was associated with a lower rate of lymph node metastasis (P = 0.041). The discrepancy may lie in the fact that prophylactic lymph node dissection is not routinely performed in differentiated thyroid cancer (9.9% Nx status in the TCGA data). In the meta-analysis, there was significant statistical heterogeneity among the studies (P = 0.012). Moreover, we noted that lymphocyte-infiltrating cancer was more likely to have multifocal involvement. This is in keeping with the experience of other researchers. As for the clonality of multifocal thyroid cancer, an independent origin of the different tumor has been demonstrated. In this regard, our observation may partially be explained by the active role of inflammation in thyroid tumorigenesis. Papillary thyroid cancer with lymphocytic infiltrates was associated with classical histologic features, but not with the status of BRAF mutation in our study. BRAF mutation is the most common genetic alteration in papillary thyroid cancer, and is typically associated with classical subtype or tall-cell variant. Studies have yielded discrepant results concerning the impact of inflammation on BRAF mutation in thyroid cancer. Some studies have shown that papillary thyroid cancer with lymphocytic thyroiditis or with greater lymphocyte infiltration is more likely to be BRAF wild-type. Nonetheless, there are reports that BRAF mutation was associated with a higher frequency of tumor-associated lymphocytes. Consistent with our findings, Huang and colleagues found that tall-cell and classical papillary thyroid cancer possesses more immune activities under miRNA regulation than the follicular variant of papillary cancer. Our pathway analysis also confirmed an upregulation of immune response and cytokine production in lymphocyte-infiltrating thyroid cancer. However, this observation requires further validation whether an upregulation in these pathways accounts for a better overall survival in patients with a lymphocyte-infiltrating thyroid cancer because the expression of some immunosuppressive molecules including PD-1 (programmed death 1) and IDO1 (indoleamine 2,3-dioxygenase 1) was also upregulated.