Risk factors and predictive nomograms for early mortality in patients with thyroid cancer lung metastasis based on the SEER database and a Chinese population study.

Background: The lung is the most vulnerable site for distant thyroid cancer (TC) metastasis, and individuals who have TC lung metastases (TCLMs) succumb to the illness shortly after diagnosis. This study aims to identify the risk factors of early mortality in TCLM patients and develop a reliable and accurate prediction model. An accurate nomogram for predicting early mortality (survival time ≤3 months) in TCLM patients is necessary. Methods: Between 2010 and 2015, information gathered from TCLM patients in the Surveillance, Epidemiology, and End Results (SEER) database was used to develop and internally evaluate a prediction model. External validation was performed using data acquired from a Chinese population. All-cause early death (ACED) encompassed mortality from any cause within this period, whereas cancer-specific early death (CSED) specifically referred to deaths explicitly attributed to TC or its complications on the death certificate. The risk factors for CSED and ACED were identified independently using univariate and multivariable logistic regressions. The nomogram's accuracy was confirmed via receiver operating characteristic (ROC) curve analysis, and calibration curves were used to evaluate the consistency between the model predictions and the actual outcomes. Decision curve analysis (DCA) was performed to assess the model's clinical applicability. Results: This study included 945 patients, 636 (67.3%) of whom died shortly after diagnosis and 335 (35.4%) of whom died from TCLM-related complications. Multivariable logistic regression analyses independently identified six predictors for ACED and seven predictors for CSED. The areas under the curve (AUCs) of the nomogram for predicting ACED and CSED were 0.912 [95% confidence interval (CI): 0.889-0.931] and 0.732 (95% CI: 0.691-0.776), respectively. Combined with the results of the calibration curve analysis, these findings demonstrated that the nomograms effectively predicted the risk of early death in both the internal and external sets. DCA revealed that the nomograms provide considerable clinical advantages. Conclusions: In the present study, nomograms were developed to reliably predict the risk of early mortality in individuals with TCLM. These tools can assist physicians in identifying high-risk patients and implementing tailored treatment plans as soon as possible.