BACKGROUND: Postoperative pancreatic fistula (POPF) after pancreatoduodenectomy (PD) leads to prolonged hospitalization and potentially fatal complications. We sought to determine whether estimated pancreatic parenchymal remnant volume (EPPRV) on preoperative computed tomography (CT) predicts clinically relevant POPF. METHODS: This retrospective study included 246 patients who underwent PD between 2008 and 2013. Pancreatic thickness, pancreatic width, and main pancreatic duct (MPD) diameter at the estimated transection line in addition to estimated whole pancreatic remnant volume (EWPRV) were measured on preoperative CT images. MPD volume was subtracted from EWPRV to determine EPPRV. The predictive ability of preoperative CT parameters for POPF was evaluated. RESULTS: EPPRV was an independent predictor of POPF and had a stronger association with POPF than EWPRV. Receiver operating characteristic curve analysis showed that EPPRV had the greatest area under the curve (0.885) for predicting POPF. EPPRV ≥25.5 cm(3) was the best cutoff value for predicting POPF, with a high negative predictive value (0.934) and low likelihood ratio of a negative result (0.235). Multivariate analysis including the preoperative CT parameters and well-known risk factors for POPF showed that EPPRV ≥25.5 cm(3) had the greatest odds ratio for POPF. EPPRV was correlated with pancreatic juice volume. Patients with EPPRV ≥25.5 cm(3) had a greater drainage fluid amylase concentration and greater duration of drainage tube placement than those with EPPRV <25.5 cm(3). CONCLUSION: EPPRV from preoperative CT was highly predictive of POPF and may help in development of management for POPF after PD.
BACKGROUND:Postoperative pancreatic fistula (POPF) after pancreatoduodenectomy (PD) leads to prolonged hospitalization and potentially fatal complications. We sought to determine whether estimated pancreatic parenchymal remnant volume (EPPRV) on preoperative computed tomography (CT) predicts clinically relevant POPF. METHODS: This retrospective study included 246 patients who underwent PD between 2008 and 2013. Pancreatic thickness, pancreatic width, and main pancreatic duct (MPD) diameter at the estimated transection line in addition to estimated whole pancreatic remnant volume (EWPRV) were measured on preoperative CT images. MPD volume was subtracted from EWPRV to determine EPPRV. The predictive ability of preoperative CT parameters for POPF was evaluated. RESULTS: EPPRV was an independent predictor of POPF and had a stronger association with POPF than EWPRV. Receiver operating characteristic curve analysis showed that EPPRV had the greatest area under the curve (0.885) for predicting POPF. EPPRV ≥25.5 cm(3) was the best cutoff value for predicting POPF, with a high negative predictive value (0.934) and low likelihood ratio of a negative result (0.235). Multivariate analysis including the preoperative CT parameters and well-known risk factors for POPF showed that EPPRV ≥25.5 cm(3) had the greatest odds ratio for POPF. EPPRV was correlated with pancreatic juice volume. Patients with EPPRV ≥25.5 cm(3) had a greater drainage fluid amylase concentration and greater duration of drainage tube placement than those with EPPRV <25.5 cm(3). CONCLUSION: EPPRV from preoperative CT was highly predictive of POPF and may help in development of management for POPF after PD.
Authors: Ye Rim Chang; Jae Seung Kang; Jin-Young Jang; Woo Hyun Jung; Mee Joo Kang; Kyung Bun Lee; Sun-Whe Kim Journal: World J Surg Date: 2017-06 Impact factor: 3.352