机器人辅助胰体尾切除术后胰瘘相关因素分析

Analysis of risk factors associated with postoperative pancreatic fistula after robotic-assisted distal pancreatectomy

目的探讨机器人辅助胰体尾切除术(RDP)术后发生胰瘘的相关因素。方法本研究为回顾性队列研究。回顾性收集2021年1月至2023年12月于上海交通大学医学院附属瑞金医院胰腺外科接受不同方式胰体尾切除术的1 211例胰体尾部肿瘤患者的临床资料。其中机器人辅助组(RDP组)440例, 男性173例, 女性267例, 年龄[M(IQR)]55(29)岁;开腹组720例, 男性390例, 女性330例, 年龄64(15)岁;腹腔镜组51例, 男性17例, 女性34例, 年龄56(25)岁。根据是否发生临床相关胰瘘(B、C级)将440例RDP组患者分为胰瘘组和无胰瘘组, 对围手术期与术后胰瘘发生相关的27个因素进行单因素和多因素分析, 单因素分析采用独立样本t检验、Mann-WhitneyU检验或χ^(2)检验, 多因素分析采用二元Logistic回归。结果按病理学类型分层后, RDP组与开腹组胰瘘发生率的差异无统计学意义(良性:χ^(2)=1.200, P=0.952;恶性:χ^(2)=0.391, P=0.532)。RDP组的手术时间(Z1=15.113, P1<0.01;Z2=4.232, P2<0.01)、术中出血量(Z1=12.530, P1<0.01, Z2=2.550, P2=0.032)均低于开腹组及腹腔镜组, 术后住院时间短于开腹组(Z1=10.947, P1<0.01), 而与腹腔镜组无差异(P2>0.05)。440例RDP组患者均顺利完成手术, 中转开腹1例, 104例(23.6%)患者发生临床相关胰瘘, 无围手术期死亡。单因素分析结果显示, 性别(χ^(2)=12.048, P=0.001)、吸烟史(χ^(2)=6.327, P=0.012)、饮酒史(χ^(2)=17.597, P<0.01)、胰腺离断方法(χ^(2)=9.839, P=0.002)、早期淀粉酶水平(Z=5.187, P<0.01)和排气恢复时间(χ^(2)=4.485, P=0.034)与RDP术后胰瘘发生相关。通过受试者工作特征曲线得到早期引流液淀粉酶的最佳截断值为7 719.5 IU/ml, 曲线下面积为0.676。二元Logistic回归分析结果显示, 有饮酒史(P=0.002, 95%CI:0.112~0.623)、手工法离断胰腺(P=0.001, 95%CI:1.446~4.082)、早期引流液淀粉酶浓度≥7 719.5 IU/ml(P<0.01, 95%CI:0.151~0.438)和延迟排气(P=0.020, 95%CI:1.131~4.233)为RDP术后发生胰瘘的独立危险因素。结论接受RDP治疗的胰体尾部肿瘤患者有饮酒史、手工法离断胰腺、早期引流液淀粉酶浓度≥7 719.5 IU/ml和延迟排气时, 更易发生胰瘘。

Objective:To investigate pertinent risk factors for postoperative pancreatic fistula(POPF)after robotic-assisted distal pancreatectomy(RDP).Methods:This is a retrospective cohort study.Clinical data of 1211 patients who underwent various methods of distal pancreatectomy at the Department of General Surgery,Ruijin Hospital,Shanghai Jiaotong University School of Medicine,between January 2021 and December 2023 were retrospectively collected.Among the 1211 patients,440 cases were in the robot-assisted group(173 males and 267 females),with an age(M(IQR))of 55(29)years;720 cases were in the open surgery group(390 males and 330 females),with an age of 64(15)years;and 51 cases were in the laparoscopic group(17 males and 34 females),with an age of 56(25)years.These 440 patients who underwent RDP were divided into two cohorts based on the presence of clinically relevant pancreatic fistulas(grades B and C).Univariate and multivariate analysis were performed on 27 factors related to POPF.Univariate analysis methods included independent sample t-test,Mann-Whitney U test,andχ^(2) test,while multivariate analysis utilized binary logistic regression.Results:After stratification by pathological type,there was no significant difference in the incidence of pancreatic fistula between the robot-assisted group and the open surgery group(benign tumor:χ^(2)=1.200,P=0.952;malignant tumor:χ^(2)=0.391,P=0.532).The surgical duration of the RDP group(Z 1=15.113,P 1<0.01;Z 2=4.232,P 2<0.01)was significantly shorter than that of the open surgery and laparoscopic groups,so as the intraoperative blood loss(Z 1=12.530,P 1<0.01;Z 2=2.550,P 2=0.032).Postoperative hospital stay in the RDP group was significantly shorter than that in the open surgery group(Z 1=10.947,P 1<0.01),but not different from that in the laparoscopic group(P 2>0.05).All 440 patients underwent successful surgery,of which there was only 1 case who underwent a conversion to open surgery.A total of 104 patients(23.6%)developed clinically relevant pancreatic fistulas,and no perioperative mortality was observed.Univariate analysis revealed that 6 factors were associated with POPF after RDP:gender(χ^(2)=12.048,P=0.001),history of smoking(χ^(2)=6.327,P=0.012),history of alcohol consumption(χ^(2)=17.597,P<0.01),manual pancreas division(χ^(2)=9.839,P=0.002),early elevation of amylase in drainage fluid(Z=5.187,P<0.01),and delayed gastric emptying(χ^(2)=4.485,P=0.034).No statistically significant association with POPF was found for the remaining factors(all P>0.05).The cut-off value for the early amylase level in the drainage fluid was determined to be 7719.5 IU/ml,with an area under curve of 0.676 determined by receiver operating characteristic curve analysis.Binary logistic regression analysis identified a history of alcohol consumption(P=0.002,95%CI:0.112 to 0.623),manual pancreas division(P=0.001,95%CI:1.446 to 4.082),early amylase level of drainage fluid≥7719.5 IU/ml(P<0.01,95%CI:0.151 to 0.438),and delayed gastric emptying(P=0.020,95%CI:1.131 to 4.233)as independent risk factors for POPF of RDP.Conclusion:Patients with pancreatic body and tail tumors who receive RDP therapy are at increased risk of developing a pancreatic fistula if they have a history of alcohol consumption,manual pancreas division,early elevation of amylase in drainage fluid to≥7719.5 IU/ml,or delayed gastric emptying.