Surgical Methods and Social Factors Are Associated With Long-Term Survival in Follicular Thyroid Carcinoma: Construction and Validation of a Prognostic Model Based on Machine Learning Algorithms

Abstract

BackgroundThis study aimed to establish and verify an effective machine learning (ML) model to predict the prognosis of follicular thyroid cancer (FTC), and compare it with the eighth edition of the American Joint Committee on Cancer (AJCC) model.MethodsKaplan-Meier method and Cox regression model were used to analyze the risk factors of cancer-specific survival (CSS). Propensity-score matching (PSM) was used to adjust the confounding factors of different surgeries. Nine different ML algorithms,including eXtreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), Random Forests (RF), Logistic Regression (LR), Adaptive Boosting (AdaBoost), Gaussian Naive Bayes (GaussianNB), K-Nearest Neighbor (KNN), Support Vector Machine (SVM) and Multi-Layer Perceptron (MLP),were used to build prognostic models of FTC.10-fold cross-validation and SHapley Additive exPlanations were used to train and visualize the optimal ML model.The AJCC model was built by multivariate Cox regression and visualized through nomogram. The performance of the XGBoost model and AJCC model was mainly assessed using the area under the receiver operating characteristic (AUROC).ResultsMultivariate Cox regression showed that age, surgical methods, marital status, T classification, N classification and M classification were independent risk factors of CSS. Among different surgeries, the prognosis of one-sided thyroid lobectomy plus isthmectomy (LO plus IO) was the best, followed by total thyroidectomy (hazard ratios: One-sided thyroid LO plus IO, 0.086[95% confidence interval (CI),0.025-0.290], P<0.001; total thyroidectomy (TT), 0.490[95%CI,0.295-0.814], P=0.006). PSM analysis proved that one-sided thyroid LO plus IO, TT, and partial thyroidectomy had no significant differences in long-term prognosis. Our study also revealed that married patients had better prognosis than single, widowed and separated patients (hazard ratios: single, 1.686[95%CI,1.146-2.479], P=0.008; widowed, 1.671[95%CI,1.163-2.402], P=0.006; separated, 4.306[95%CI,2.039-9.093], P<0.001). Among different ML algorithms, the XGBoost model had the best performance, followed by Gaussian NB, RF, LR, MLP, LightGBM, AdaBoost, KNN and SVM. In predicting FTC prognosis, the predictive performance of the XGBoost model was relatively better than the AJCC model (AUROC: 0.886 vs. 0.814).ConclusionFor high-risk groups, effective surgical methods and well marital status can improve the prognosis of FTC. Compared with the traditional AJCC model, the XGBoost model has relatively better prediction accuracy and clinical usage.

Establishment and verification of the first prognostic nomograms in locally advanced thyroid cancer based on the analysis of clinical and follow-up information on 2396 patients

“…It is noteworthy that our study concluded that patients with single, widowed and separated status had worse prognosis and marital status as one of the predictive factors affecting OS in LATC patients, while Li et al [ 26 ] also concluded that marital status in patients with TC was significantly associated with both CSS and OS. Previous studies have also concluded that good marital status has a positive effect on the prognosis of a variety of malignant tumors [ 27 ], and since TC, as an endocrine system disease, emotional changes and psychological well-being may be closely related to its prognosis, we believe that providing effective psychological counseling and social support to unmarried, widowed and separated patients may have a positive effect on the patients’ prognosis. Similarly, we only found that males had a worse prognosis than females in the predictive analyses of OS in the LATC patients, which is very similar to the findings of Li et al [ 28 ] However, Zhou et al [ 29 ] found that the prognosis of male patients with DTC did not differ from that of females after matching by propensity score.…”

Enhancing clinical decision-making: A novel nomogram for stratifying cancer-specific survival in middle-aged individuals with follicular thyroid carcinoma utilizing SEER data

“…In the present study, we found that married middle-aged FTC patients reflected a longer survival time and better prognosis, consistent with previous studies related to thyroid cancer. Exploring the underlying reasons, in terms of psychosocial, economic or treatment adherence, we found that married patients may have a better prognosis due to better social support and mental health status [ 33 ]. Unmarried or divorced patients may face more psychological stress and less social support, which may affect their adherence to treatment and, thus, overall health [ 34 ].…”

Machine learning-based models for the prediction of breast cancer recurrence risk

“…In the test set, the AUC, accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and F1 score were estimated. The best prediction model was selected by evaluating the largest AUC [ 47 , 48 ]. We applied the Shapley Additive Explanation (SHAP) to explain the best-performing predictive model.…”

“…The features were ordered by the mean absolute value of the SHAP values for each feature [ 44 ]. Combining ML with SHAP could provide an explicit explanation of the efficacy prediction [ 44 , 47 , 49 ]. The process is shown in Fig.…”