Jie Shan1, Rui Jiang2, Xin Chen1, Yi Zhong3, Wei Zhang4, Lizhe Xie5, Jie Cheng6, Hongbing Jiang7. 1. MSc Student, Jiangsu Key Laboratory of Oral Diseases, Nanjing Medical University, Nanjing, China; and Resident, Department of Oral and Maxillofacial Surgery, The Affiliated Stomatological Hospital of Nanjing Medical University, Nanjing, China. 2. BSc Student, College of Communication and Information Engineering, Nanjing University of Posts and Telecommunications, Nanjing, China. 3. Resident, Department of Oral Pathology, The Affiliated Stomatological Hospital of Nanjing Medical University, Nanjing, China. 4. Associated Department Head, Department of Oral Pathology, The Affiliated Stomatological Hospital of Nanjing Medical University, Nanjing, China. 5. Associated Professor, Jiangsu Key Laboratory of Oral Diseases, Nanjing Medical University, Nanjing, China. 6. Professor, Jiangsu Key Laboratory of Oral Diseases, Nanjing Medical University, Nanjing, China; and Associated Department Head, Department of Oral and Maxillofacial Surgery, The Affiliated Stomatological Hospital of Nanjing Medical University, Nanjing, China. 7. Professor, Jiangsu Key Laboratory of Oral Diseases, Nanjing Medical University, Nanjing, China; and Department Head, Department of Oral and Maxillofacial Surgery, The Affiliated Stomatological Hospital of Nanjing Medical University, Nanjing, China. Electronic address: jhb@njmu.edu.cn.
Abstract
PURPOSE: Early-stage oral tongue squamous cell cancer (OTSCC) has a rate of metastasis to the cervical lymph nodes of 20 to 50%. This study aimed to build and validate 4 machine learning (ML) models to predict the occurrence of lymph node metastasis before and after surgery for early-stage (cT1N0 to cT2N0) OTSCC. MATERIALS AND METHODS: We designed a retrospective cross-sectional study and reviewed the clinical and pathologic records of patients with early-stage OTSCC. The sample was composed of 2 groups with different node status (positive or negative) and was randomly split into training (70%) and testing (30%) sets. Four common ML algorithms-logistic regression, random forest, support vector machine, and naive Bayes-were used to predict pathologic nodal metastasis of early-stage OTSCC. The area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were used to assess the performance of these models and conventional methods including depth of invasion (DOI), neutrophil-to-lymphocyte ratio (NLR), and tumor budding. RESULTS: A total of 145 patients (56 with positive and 89 with negative lymph nodes) were included in this study. The performance of ML models was significantly superior to that of conventional prediction methods. The random forest model performed best (AUC, 0.786; sensitivity, 85%; specificity, 75%) and exceeded the performance of NLR (AUC, 0.539; sensitivity, 53.6%; specificity, 53.9%; P = .003). When DOI, worst pattern of invasion, lymphocytic host response, and tumor budding were added to model analysis according to patients' postoperative pathologic records, the support vector machine model performed best (AUC, 0.956; sensitivity, 100%; specificity, 87.5%) and was superior to univariate assessment of tumor budding (AUC, 0.830; sensitivity, 80.9%; specificity, 87.5%, P = .002), DOI (AUC, 0.613; sensitivity, 91.1%; specificity, 31.5%; P < .001), and NLR. CONCLUSIONS: ML shows a better performance in predicting lymph node metastasis of early-stage OTSCC than conventional prediction methods of DOI, NLR, or tumor budding.
PURPOSE: Early-stage oral tongue squamous cell cancer (OTSCC) has a rate of metastasis to the cervical lymph nodes of 20 to 50%. This study aimed to build and validate 4 machine learning (ML) models to predict the occurrence of lymph node metastasis before and after surgery for early-stage (cT1N0 to cT2N0) OTSCC. MATERIALS AND METHODS: We designed a retrospective cross-sectional study and reviewed the clinical and pathologic records of patients with early-stage OTSCC. The sample was composed of 2 groups with different node status (positive or negative) and was randomly split into training (70%) and testing (30%) sets. Four common ML algorithms-logistic regression, random forest, support vector machine, and naive Bayes-were used to predict pathologic nodal metastasis of early-stage OTSCC. The area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were used to assess the performance of these models and conventional methods including depth of invasion (DOI), neutrophil-to-lymphocyte ratio (NLR), and tumor budding. RESULTS: A total of 145 patients (56 with positive and 89 with negative lymph nodes) were included in this study. The performance of ML models was significantly superior to that of conventional prediction methods. The random forest model performed best (AUC, 0.786; sensitivity, 85%; specificity, 75%) and exceeded the performance of NLR (AUC, 0.539; sensitivity, 53.6%; specificity, 53.9%; P = .003). When DOI, worst pattern of invasion, lymphocytic host response, and tumor budding were added to model analysis according to patients' postoperative pathologic records, the support vector machine model performed best (AUC, 0.956; sensitivity, 100%; specificity, 87.5%) and was superior to univariate assessment of tumor budding (AUC, 0.830; sensitivity, 80.9%; specificity, 87.5%, P = .002), DOI (AUC, 0.613; sensitivity, 91.1%; specificity, 31.5%; P < .001), and NLR. CONCLUSIONS: ML shows a better performance in predicting lymph node metastasis of early-stage OTSCC than conventional prediction methods of DOI, NLR, or tumor budding.
Authors: Venkatesan Chandran; M G Sumithra; Alagar Karthick; Tony George; M Deivakani; Balan Elakkiya; Umashankar Subramaniam; S Manoharan Journal: Biomed Res Int Date: 2021-05-04 Impact factor: 3.411