Exploration Of Nomogram For Predicting Difficulty Of Laparoscopic Surgery In Rectal Cancer

Purpose:This study aimed to identify risk factors associated with difficulty of laparoscopic radical resection of rectal cancer,and to establish a predictive nomogram and decision curve to help communication between clinicians and patients and make individual clinical treatment decisions.Methods:A total of 977 patients with rectal cancer receiving laparoscopic radical resection between January 2014 and December 2016 were enrolled in this study.Clinical information including gender,age,body mass index(BMI),distance between tumor and the anal verge(DTA),maximum transverse diameter of tumor,interischial distance(IS),intertuberous distance(IT),history of abdominal surgery(HAS),preoperative intestinal obstruction,preoperative chemoradiotherapy(CRT),operation time,estimated blood loss,conversion to open surgery,transabdominal surgery,postoperative hospitalization days,and postoperative complications(Clavien-Dingo classification).Difficulty of the operation was defined according to the scoring criteria of Escal et al as follows:1)operation time>300mins(3 points),2)conversion to open surgery(3 points),3)postoperative stay>7 days(2 points),4)transabdominal surgery(2 points),5)blood loss>200ml(1 points),and 6)postoperative complications(1 points).The total score was 0-12,and a grade beyond 3 was considered to indicate high risk of surgical difficulty,whereas a grade of 3 or less indicated low risk of difficulty.In total,126 patients(12.9 per cent)were in the difficult operation group and 851(87.1 per cent)in the non-difficult operation group.Logistic regression analysis was performed to identify predictive factors associated with surgical difficulty,and to build a difficult operation risk nomogram.Discrimination and calibration of the nomogram model were assessed by using the C-index and calibration plot,and internal validation was assessed using the bootstrapping validation.In addition,decision curve was plotted to help clinicians to make clinical decisions.Results:Multivariate analysis showed that BMI>28kg/m~2,distance between tumor and the anal margin≤5 cm,the maximum transverse diameter of tumor>3cm tumor,interischial distance<10cm,history of abdominal surgery and preoperative radiotherapy were independent risk factors for surgical difficulty in rectal cancer patients.Then,a nomogram model was built using R software to predict the risk of difficulty operation in laparoscopic rectal cancer surgery.The model displayed good discrimination with a C-index of 0.744(95%CI:0.690-0.797)and a high C-index value of 0.738 in the interval validation,indicating that the nomogram model had good predictive power.In addition,decision curve analysis indicated that the nomogram was clinically useful when intervention was from 0.1 to 0.85,suggesting that rectal cancer patients can get more clinical benefits.Conclusion:We first defined difficult laparoscopic surgery for rectal cancer,and identified independent risk factors for difficult operation,including BMI>28kg/m~2,distance between tumor and the anal margin≤5cm,the maximum transverse diameter of tumor>3cm tumor,interischial distance<10cm,history of abdominal surgery and preoperative radiotherapy.Then,a nomogram model was established to predict the risk of difficult operation in rectal cancer patients.Finally,decision curve was plotted to help clinicians to develop treatment strategies.Internal validation of the nomogram model showed that it was a good tool for clinical decision-making,including supporting clinicians’surgical choices,facilitating communication between doctors and patients,and adjusting patient’s surgical expectations.