50 - Communication libre
3 juin 2021, 13:50 - 15:20, Stream 2: SSCViscérale
Individual patient data meta-analysis of delayed gastric emptying after pylorus-preserving versus pylorus-resecting pancreatoduodenectomy
P. Probst1, U. Klaiber1, S. Seide1, M. Kawai2, I. Matsumoto3, O. Strobel1, T. Hackert1, M. W. Büchler1, M. K. Diener1, Presenter: P. Probst1 (1Heidelberg, 2Wakayama, 3Osaka)
Some studies have indicated that resecting the pylorus during partial pancreatoduodenectomy (PD) may lead to reduced delayed gastric emptying (DGE). Randomized controlled trials (RCTs) showed conflicting results regarding superiority of pylorus-resecting PD (prPD) compared to the pylorus-preserving procedure (ppPD). The aim of this individual patient data meta-analysis was to investigate risk factors on an individual patient level which may explain the observed differences between the existing RCTs.
RCTs comparing ppPD and prPD were searched systematically in MEDLINE, Web of Science and CENTRAL. Individual patient data (IPD) from existing RCTs were included. The primary endpoint was DGE according to the International Study Group of Pancreatic Surgery (ISGPS) adjusted for age, sex and body-mass-index (BMI). The meta-regression model was applied to the IPD of the RCTs. Mixed effects models were applied to perform meta-analyses.
IPD from 418 patients (three RCTs) were used for quantitative synthesis. There was no significant statistical difference between ppPD and prPD regarding DGE adjusted for age, sex and BMI (OR 0.72; 95%-CI: 0.41 to 1.22) and DGE grade (RR 1.01; 95%-CI: 0.64 to 1.57). Regarding other relevant perioperative and postoperative outcome parameters, there were also no significant differences among the two techniques.
This IPD meta-analysis comparing preservation and resection of the pylorus during PD confirmed that the resection of the pylorus is not superior to the pylorus-preserving procedure regarding DGE. The pylorus should therefore be preserved whenever possible. Further RCT are futile, because their results are unlikely to change the pooled estimate for DGE.