1. LASSO-based nomogram predicts the risk factors of low anterior resection syndrome for middle and low rectal cancer underwent robotic surgery.
- Author
-
Houqiong, Ju, Yuli, Yuan, Yahang, Liang, Tao, Li, Yang, Liu, Yaxiong, Liu, Dongning, Liu, and Taiyuan, Li
- Subjects
SURGICAL robots ,RISK assessment ,PREDICTION models ,EARLY medical intervention ,RESEARCH funding ,BODY mass index ,ACADEMIC medical centers ,T-test (Statistics) ,LOGISTIC regression analysis ,FISHER exact test ,TREATMENT effectiveness ,RETROSPECTIVE studies ,DESCRIPTIVE statistics ,MULTIVARIATE analysis ,DECISION making in clinical medicine ,MAGNETIC resonance imaging ,CHI-squared test ,EVALUATION of medical care ,RECTUM tumors ,SURGICAL complications ,ODDS ratio ,QUALITY of life ,STATISTICS ,COMBINED modality therapy ,DATA analysis software ,CONFIDENCE intervals ,TUMOR classification ,SENSITIVITY & specificity (Statistics) ,DISEASE risk factors - Abstract
Background: This study aims to analyze the influencing factors of postoperative Low Anterior Resection Syndrome (LARS) in patients with middle and low rectal cancer who underwent robotic surgery. It also seeks to predict the probability of LARS through a visual, quantitative, and graphical nomogram. This approach is expected to lower the risk of postoperative LARS in these patients and improve their quality of life through effective prevention and early intervention. Patients and methods: This research involved patients with middle and low rectal cancer who underwent robotic surgery in the Department of Gastrointestinal Surgery at the First Affiliated Hospital of Nanchang University from January 2015 to October 2022. A series of intestinal dysfunction symptoms arising from postoperative rectal cancer were diagnosed and graded using LARS scoring criteria. After the initial screening of all variables related to LARS with Lasso regression, they were included in logistic regression for further univariate and multivariate analysis to identify independent risk factors for LARS. A prediction model was then constructed. Results: The study included 358 patients. The parameters identified by Lasso regression included obstruction, BMI, tumor localization, maximum tumor diameter, AJCC stage, stoma, neoadjuvant therapy (NAT), and postoperative adjuvant therapy (AT). Univariate and multivariate analyses indicated that a higher BMI, lower tumor localization, higher AJCC stage, neoadjuvant therapy, and postoperative adjuvant therapy were independent risk factors for total LARS. The AUC of the prediction nomogram was 0.834, with a sensitivity of 0.825 and specificity of 0.741. The calibration curve demonstrated excellent concordance with the nomogram, indicating the prediction curve fit the diagonal well. Conclusion: Higher BMI, lower tumor localization, higher AJCC stage, neoadjuvant therapy, and adjuvant therapy were identified as independent risk factors for total LARS. A new predictive nomogram for postoperative LARS in patients with middle and low rectal cancer undergoing robotic surgery was developed, proving to be stable and reliable. This tool will assist clinicians in managing the postoperative treatment of these patients, facilitating better clinical decision-making and maximizing patient benefits. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF