Comprehensive Summary
Xu et al developed and validated a support vector machine (SVM)-based model to look for patterns that predict recurrence risk and associated factors for esophageal squamous cell carcinoma (ESCC) patients after surgery. They used data from 311 patients who had surgery for ESCC between June 2014 and November 2016, and followed them for about 8 years. The SVM model separated patients into high and low recurrence risk groups. It considered factors like tumor stage (TMN), tumor size, cell differentiation, adjuvant therapy, and complications after surgery. The results showed that the most accurate model, SVM6+8, successfully distinguished between high and low recurrence patients. Additionally, it achieved high predictive performance with sensitivities up to 94% and specificities up to 98.11%. The model showed that patients classified as low-risk by the model had significantly longer disease-free survival than those in the high-risk group.
Outcomes and Implications
Many models already exist to predict overall survival in ESCC post-operative patients; very few have focused on post-operative recurrence. The proposed model in this study accurately predicts post-operative recurrence in ESCC patients with high sensitivity and specificity. This approach could help doctors personalize follow-up care and provide patients who need it with closer monitoring or additional treatment after surgery.