Predicting survival of advanced laryngeal squamous cell carcinoma: comparison of machine learning models and Cox regression models

Abstract

Laryngeal squamous cell carcinoma (LSCC) is a common tumor type. High recurrence rates remain an important factor affecting the survival and quality of life of advanced LSCC patients. We aimed to build a new nomogram and a random survival forest model using machine learning to predict the risk of LSCC progress. The study included 671 patients with AJCC stages III–IV LSCC. To develop a prognostic model, Cox regression analyses were used to assess the relationship between clinic-pathologic factors and disease-free survival (DFS). RSF analysis was also used to predict the DFS of LSCC patients. The ROC curve revealed that the Cox model exhibited good sensitivity and specificity in predicting DFS in the training and validation cohorts (1 year, validation AUC = 0.679, training AUC = 0.693; 3 years, validation AUC = 0.716, training AUC = 0.655; 5 years, validation AUC = 0.717, training AUC = 0.659). Random survival forest analysis showed that N stage, clinical stage, and postoperative chemoradiotherapy were prognostically significant variables associated with survival. The random forest model exhibited better prediction ability than the Cox regression model in the training cohort; however, the two models showed similar prediction ability in the validation cohort.

Prognosing post-treatment outcomes of head and neck cancer using structured data and machine learning: A systematic review

“…The only study that performed time-to-event analyses did not report a c-index for the RSF model, but it was 0.60 for Cox PH [ 83 ].…”

Identification of crucial anoikis-related genes as novel biomarkers and potential therapeutic targets for lung adenocarcinoma via bioinformatic analysis and experimental verification

“…Then, we performed the least absolute shrinkage and selection operator (LASSO) Cox regression analysis to avoid excessive variables [10], and identified five ARGs (PLK1, SLC2A1, ANGPTL4, CDKN3, HMGA1) (Figure 2B). Meanwhile, we performed the random forest algorithm to rank the importance of ARGs [11], and selected the top five important ARGs (PLK1, SLC2A1, ANGPTL4, CDKN3, PBK) (Figure 2C). Furthermore, four ARGs (PLK1, SLC2A1, ANGPTL4, CDKN3) were obtained by intersections of ARGs screened by the above two machine learning algorithms (Figure 2D), and these four genes have been found to play essential roles in LUAD development [12][13][14][15].…”

Nomogram to predict prognosis of head and neck rhabdomyosarcoma patients in children and adolescents

“…A nomogram is a graphical tool for displaying the predicted value of individual survival based on significant variables produced by multivariate regression analysis briefly and intuitively (11, 12); currently, it has been widely used in disease prediction of malignant tumors such as laryngeal carcinoma and lung cancer etc (13)(14)(15). There are also some nomograms that have been constructed and proven to be beneficial in the management of RMS (16)(17)(18), and they present obvious advantages such as improved predictive accuracy, robustness, and usability, all of which increase their potentials in clinical practices.…”