چكيده لاتين
Papillary thyroid carcinoma (PTC) is a malignancy whose etiology is not yet fully understood. The competitive endogenous RNA (ceRNA) hypothesis offers a model to better comprehend the molecular mechanisms driving carcinogenesis. This study aimed to construct a novel ceRNA network to identify reliable diagnostic and prognostic indicators applicable across all stages of PTC. Transcriptome analysis identified differentially expressed genes (DEGs) common to all PTC stages compared to normal tissue. Key genes were subsequently screened using MCC, IVI, and MCODE algorithms to construct a novel five-layer ceRNA network and its associated DE-TF regulatory network. The diagnostic performance of elements within both networks was evaluated using Receiver Operating Characteristic (ROC) curves. Furthermore, a risk assessment model was developed by identifying key genes from the ceRNA and DE-TF components through univariable Cox and LASSO regression analyses. Findings were validated in tissue and peripheral blood samples using RT-qPCR, and correlations between gene expression and blood calcium levels were investigated.
The ceRNA and DE-TF networks comprised 30 and 22 components, respectively. Logistic regression analysis identified a panel of 8 genes (PKMYT1, E2F1, NFATC1, STAT6, E2F3, LINC02910, GAS5, and TK1) as reliable diagnostic markers for PTC, achieving an AUC of 96.9%. Among these, PKMYT1 and GAS5 were identified as stage-specific markers in tumor tissues, showing significant upregulation in High-Aggressive (HA) tumors compared to Low-Aggressive (LA) ones. These two genes demonstrated strong diagnostic value in differentiating HA from LA tumors, with AUCs of 0.81 and 0.87, respectively. circMET, which was upregulated in both groups, showed potential in distinguishing LA tumors from adjacent normal tissues (AUC = 0.81). Non-invasive evaluation in peripheral blood revealed a distinct pattern, while GAS5 was downregulated, miR-3671 and circMET were significantly upregulated in the HA group compared to the LA group, offering a meaningful profile for distinguishing disease progression stages. Additionally, GAS5 expression was associated with blood calcium levels.
The SERN prognostic model, comprising STAT6, E2F1, RMI2, and NR4A1, highlighted the significance of these genes as reliable prognostic markers for overall survival in PTC. In conclusion, PKMYT1, GAS5, and circMET are significantly associated with PTC aggressiveness. PKMYT1 and GAS5 serve as strong indicators for tumor aggression, while circMET is valuable for early detection, and GAS5 correlates with physiological calcium levels.
