Machine learning model for prediction of low anterior resection syndrome following laparoscopic anterior resection of rectal cancer: A multicenter study.

BACKGROUND: Low anterior resection syndrome (LARS) severely impairs patient postoperative quality of life, especially major LARS. However, there are few tools that can accurately predict major LARS in clinical practice. AIM: To develop a machine learning model using preoperative and intraoperative factors for predicting major LARS following laparoscopic surgery of rectal cancer in Chinese populations. METHODS: Clinical data and follow-up information of patients who received laparoscopic anterior resection for rectal cancer from two medical centers (one discovery cohort and one external validation cohort) were included in this retrospective study. For the discovery cohort, the machine learning prediction algorithms were developed and internally validated. In the external validation cohort, we evaluated the trained model using various performance metrics. Further, the clinical utility of the model was tested by decision curve analysis. RESULTS: Overall, 1651 patients were included in the present study. Anastomotic height, neoadjuvant therapy, diverting stoma, body mass index, clinical stage, specimen length, tumor size, and age were the risk factors associated with major LARS. They were used to construct the machine learning model to predict major LARS. The trained random forest (RF) model performed with an area under the curve of 0.852 and a sensitivity of 0.795 (95%CI: 0.681-0.877), a specificity of 0.758 (95%CI: 0.671-0.828), and Brier score of 0.166 in the external validation set. Compared to the previous preoperative LARS score model, the current model exhibited superior predictive performance in predicting major LARS in our cohort (accuracy of 0.772 for the RF model vs 0.355 for the preoperative LARS score model). CONCLUSION: We developed and validated a robust tool for predicting major LARS. This model could potentially be used in the clinic to identify patients with a high risk of developing major LARS and then improve the quality of life.

Comprehensive abdominal composition evaluation of rectal cancer patients with anastomotic leakage compared with body mass index-matched controls.

BACKGROUND: Anastomotic leakage (AL) is a fatal complication in patients with rectal cancer after undergoing anterior resection. However, the role of abdominal composition in the development of AL has not been studied. AIM: To investigate the relationship between abdominal composition and AL in rectal cancer patients after undergoing anterior resection. METHODS: A retrospective case-matched cohort study was conducted. Complete data for 78 patients with AL were acquired and this cohort was defined as the AL group. The controls were matched for the same sex and body mass index (± 1 kg/m2). Parameters related to abdominal composition including visceral fat area (VFA), subcutaneous fat area (SFA), subcutaneous fat thickness (SFT), skeletal muscle area (SMA), skeletal muscle index (SMI), abdominal circumference (AC), anterior to posterior diameter of abdominal cavity (APD), and transverse diameter of abdominal cavity (TD) were evaluated based on computed tomography (CT) images using the following Hounsfield Unit (HU) thresholds: SFA: -190 to -30, SMA: -29 to 150, and VFA: -150 to -20. The significance of abdominal composition-related parameters was quantified using feature importance analysis; an artificial intelligence method was used to evaluate the contribution of each included variable. RESULTS: Two thousand two hundred and thirty-eight rectal cancer patients who underwent anterior resection from 2010 to 2020 in a large academic hospital were investigated. Finally, 156 cases were enrolled in the study. Patients in the AL group showed longer operative time (225.03 ± 55.29 vs 207.17 ± 40.80, P = 0.023), lower levels of preoperative hemoglobin (123.32 ± 21.17 vs 132.60 ±1 6.31, P = 0.003) and albumin (38.34 ± 4.01 vs 40.52 ± 3.97, P = 0.001), larger tumor size (4.07 ± 1.36 vs 2.76 ± 1.28, P < 0.001), and later cancer stage (P < 0.001) compared to the controls. Patients who developed AL exhibited a larger VFA (125.68 ± 73.59 vs 97.03 ± 57.66, P = 0.008) and a smaller APD (77.30 ± 23.23 vs 92.09 ± 26.40, P < 0.001) and TD (22.90 ± 2.23 vs 24.21 ± 2.90, P = 0.002) compared to their matched controls. Feature importance analysis revealed that TD, APD, and VFA were the three most important abdominal composition-related features. CONCLUSION: AL patients have a higher visceral fat content and a narrower abdominal structure compared to matched controls.