A preoperative MRI pelvic multiparameter nomogram for predicting sphincter preservation and operative difficulty in laparoscopic total mesorectal excision for ultra-low rectal cancer

术前盆腔MRI多参数列线图用于预测超低位直肠癌腹腔镜全直肠系膜切除术中括约肌保留情况和手术难度

阅读:1

Abstract

BACKGROUND: Sphincter preservation (SP) during laparoscopic total mesorectal excision (LaTME) for low rectal cancer is challenged by a deep, narrow pelvis. Robust, MRI-based predictors for both SP success and operative difficulty are lacking. METHODS: We conducted a retrospective study of 275 consecutive patients undergoing LaTME (development cohort, Jan 2022-Feb 2024). An independent, temporal validation cohort (n = 118; Mar-Dec 2024) from the same institution was used. Eight predefined pelvic MRI parameters were measured. The primary outcome was SP failure (conversion to abdominoperineal resection). Surgical difficulty was defined as operative time ≥ 245 min. Independent predictors from multivariable analysis were used to construct a predictive nomogram. Model performance was evaluated by discrimination (AUC), calibration, and clinical utility (decision curve analysis). RESULTS: Shorter tumor distance from the anal verge (OR, 0.26), greater tumor diameter (OR, 1.45), and narrower intertuberous distance (OR, 0.75) were independent predictors of SP failure. The nomogram showed good discrimination (AUC: 0.85 development, 0.84 validation) and calibration. Among patients with successful SP (n = 278), neoadjuvant radiotherapy (OR, 2.11), shorter tumor distance (OR, 0.65), and larger mesorectal anteroposterior diameter (OR, 1.73) were independently associated with surgical difficulty. CONCLUSION: We developed and validated an MRI-based nomogram that quantifies the likelihood of SP and identifies factors associated with surgical difficulty in LaTME for low rectal cancer. This tool facilitates preoperative risk stratification and personalized surgical planning.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。