Integrated imaging model for predicting lateral lymph node metastasis in patients undergoing lateral lymph node dissection for rectal cancer

Purpose Lateral lymph node metastasis (LLNM) is an important cause of local recurrence in middle and lower rectal cancer. This study aimed to develop an integrated imaging-based prediction model to improve risk stratification for LLNM. Methods This retrospective study included 288 patients with stage II–III rectal cancer located below the peritoneal reflection who underwent curative resection with LLND. Preoperative magnetic resonance imaging (MRI) and computed tomography colonography (CTC) were used to evaluate extramural venous invasion (EMVI) on MRI, mesorectal lymph node size (MLN), lateral lymph node size (LLN), and tumor distance from the anal verge measured on CTC (AV distance). Multivariable logistic regression was performed to identify independent predictors and develop a prediction model. Model discrimination was assessed using receiver operating characteristic curves and internally validated using bootstrap resampling. Risk stratification was performed based on predicted probabilities. Results LLNM was observed in 33 patients (11.5%). Multivariable analysis identified EMVI positivity on MRI, LLN ≥ 7 mm, MLN ≥ 9 mm, and shorter AV distance (modeled as a continuous variable) as independent predictors. The model showed good discrimination (AUC 0.87, 95% CI 0.79–0.95), with improved performance compared with conventional imaging-based models. A low-risk group (< 5%) had an LLNM rate of approximately 3%, whereas a high-risk group (≥ 20%) showed an LLNM rate exceeding 50%. Conclusion This integrated imaging-based model demonstrated good discrimination for predicting LLNM among patients with middle and lower rectal cancer undergoing LLND and may support preoperative risk stratification.