Analysis of factors associated with the prognosis of papillary thyroid cancer and the construction of a survival model

ObjectiveTo study the survival prediction value of lymph node ratio (LNR) and preoperative thyroglobulin (Tg) in the prognosis of thyroid papillary carcinoma (PTC). MethodsA total of 495 patients with PTC and lymph node metastasis treated at the Cancer Hospital of Xinjiang Medical University were selected for a retrospective study. The disease-free survival (DFS) of patients was the follow-up endpoint. DFS was calculated for all patients. The Cox proportional risk regression model and nomogram were used to predict the survival prognosis of PTC with lymph node metastasis by index. LNR and preoperative Tg level cutoff values were obtained using ROC curves. To express DFS, Kaplan-Meier survival curves were created. Using 3- and 5-year calibration curves and AUC values, the prognostic models' precision and discrimination were assessed. Clinical decision curve analysis was used to forecast clinical benefitability. Finally, the results were validated using internal cross-validation. ResultsThe cutoff values of LNR and preoperative Tg level were 0.295 and 50.24, respectively, and they were divided into two groups according to the cutoff values. Multifactorial Cox regression models showed that NLNM, LNR, and preoperative Tg level (all p < 0.05) were independent risk factors affecting the prognosis of PTC with lymph node metastasis. Kaplan-Meier curves showed higher DFS rates in the group with low NLNM (<10), LNR (<0.295), and preoperative Tg level (<50.24) groups. The 3-year and 5-year calibration curves showed good agreement. A ROC curve analysis was performed on the nomogram model, and its AUC values at 3 and 5 years were, respectively, 0.805 and 0.793. Clinical decision curves indicate good clinical benefit. Finally, internal cross-validation demonstrated the legitimacy of the prognostic model. ConclusionThe LNR and preoperative Tg levels, in combination with other independent factors, were effective in predicting the survival prognosis for patients with PTC.