|
Back to 2014 Annual Meeting Posters
Techniques of Pancreatic Reconstruction After Pancreaticoduodenectomy: Analysis of Pancreatic Fistula Rate and Severity, Risk Factors and Impact of Omental Wrap
Stefano D'Ugo*, John M. O'Callaghan, Zahir Soonawalla, Srikanth Reddy, Michael Silva, Gabriele Spoletini, Peter Friend
Hepatobiliary and Pancreatic Surgery, Oxford University Hospitals, Oxford, United Kingdom
Background:
Postoperative pancreatic fistula (PF) is the leading cause of death and morbidity after pancreaticoduodenectomy (PD). Different techniques of pancreatic reconstruction have been described; however, the best method to reduce occurrence of PF is debated. Recently omental wrap has been proposed to decrease the rate of severe PF. Primary aim of the study is to compare incidence of PF among different techniques and investigate the role of omental wrap as protection against severe PF.
Methods:
Patients undergoing PD between October 2009 and August 2013 in a single Centre were retrospectively analyzed. Based on the reconstruction, they were divided in: pancreaticojejunostomy (PJ) vs pancreaticogastrostomy (PG); PJ single- vs double-layer (SL/DL); invaginated vs duct to mucosa anastomosis. Use of omental wrap was recorded. After surgery all the patients were managed homogeneously according to a specific Unit protocol. Drain amylase was tested on day 1, 3 and 5 if drain still in place; diagnosis of PF was done according to the International Study Group on Pancreatic Fistula (ISGPF) definition and grading. Univariate and multivariate analysis was done to identify risk factors for PF.
Results:
163 patients were included in the analysis (M/F:95/68), with mean age of 64.3±11 years. Pylorus preserving technique was done in 92% (n=150), and vascular resection in 14% (n=23). Based on the reconstruction 16% (n=26) had PJ/SL, 33% (n=54) PJ/DL, 18% (n=30) PJ dunking, 32% (n=53) PG. In 38% (n=62) of them omental wrap was used. Firm pancreas was recognised in 51% of cases and dilated duct in 55%. Overall PF rate was 36.8% (n=60); clinically relevant PF (grade B/C) were 31 (19%). Risk factor for PF at univariate analysis were: higher Body Mass Index (BMI) (27.8±0.8 vs 25.0±0.7;p=0.01); Cardiovascular (CV)-disease (30% vs 12%;p=0.008); small duct (55% vs 26%;p<0.0001); soft gland (61% vs 31%; p=0.006). Independent risk factors at multivariate were drain amylase on day 1 (p=0.016), BMI (p=0.007) and soft gland (p=0.02). No differences among the anastomotic techniques were detected. Omental wrap reduced the grade B/C PF rate close to statistic significance (13% vs 26%;p=0.06). Drain amylase on day 1 and 3 was a positive predictive factor for PF (p<0.00001); a cut-off value of 300 UI/L on day 1 can identify patients at significant risk for PF (p<0.001).
Conclusions:
PF rate and severity are not impacted by the reconstruction technique. BMI, CV-disease, pancreatic duct size and gland texture are significant risk factors. Day 1 and 3 amylase can predict the occurrence of PF, and a cut-off of 300 UI/L on day 1 can be used in clinical practice. Omental wrap might play a role in reducing clinically relevant PF.
Back to 2014 Annual Meeting Posters
|
