Predictive Model For Papillary Thyroid Cancer Recurrence Risk Based On Glycolysis-related Genes

Objective: Surgical resection is the treatment of choice for papillary thyroid carcinoma(PTC).After standardized treatment,patients with PTC generally have a good prognosis with a ten-year survival of up to 90%,but about 15% of patients are still at risk of recurrence.Glycolysis is one of the distinguishing features of malignancy.This study intends to construct a glycolysis-related risk prediction model to provide ideas for the management of postoperative recurrence of PTC.Methods: PTC-related sequencing data and clinical data were downloaded from TCGA database and GEO database.Recurrence-related glycolysis-related genes were screened by differential analysis,survival analysis and LASSO algorithm,and cellular experiments were used for expression and functional validation.Glycolysis-related risk score(GRS)was then developed by Cox regression algorithm,and a comprehensive decision tree based on GRS model was constructed by decision tree algorithm to help clinical decision making.Various statistical methods were used to assess the model efficacy.In addition,this study explored the potential mechanisms affecting the risk stratification of PTC by pathway enrichment analysis and immune infiltration analysis.All analyses were performed in R statistical software.Results: High glycolysis scores were strongly associated with poor prognosis.Glycolysis-related genes(ADM,CD44,MKI67 and TYMS)were upregulated in PTC expression and closely associated with PTC prognosis.The GRS score constructed based on these four genes could effectively differentiate the risk of recurrence after PTC(AUC = 0.767).The results of multifactorial Cox analysis further suggested that both GRS and lymph node metastasis status were independent risk factors for PTC prognosis(HR > 1,p < 0.05).Then,a decision system developed based on the 4-gene model and lymph node metastasis status could classify patients into three low,medium and high risk groups,with significant differences between groups.In addition,mechanistic exploration revealed that ADM,CD44,MKI67 and TYMS may influence the prognosis of PTC patients by modulating the immune response.Conclusion: The model constructed based on 4 glycolysis-related genes(ADM,CD44,MKI67 and TYMS)can effectively assess the risk of recurrence after PTC and has some clinical value.