The attenuation value of preoperative computed tomography as a novel predictor for pancreatic fistula after pancreaticoduodenectomy

Abstract

Purpose

Pancreatic fistula (PF) is the most serious complication following pancreaticoduodenectomy (PD). This study was performed to identify new clinical factors that may predict the development of PF after PD to improve perioperative management.

Methods

Seventy-five consecutive patients who underwent PD from 2012 to 2015 were evaluated. The patients' perioperative data including the computed tomography (CT) parameters were collected. The minimum, maximum, and mean CT attenuation values (HU_min, HU_max, and HU_mean, respectively) were extracted from the pancreatic parenchyma (≥ 100 pixels), and the standard deviation of these values (HU_SD) was determined from the slice in which the superior mesenteric and splenic veins were merged. PF was defined as grade B or C according to the International Study Group for Pancreatic Fistula criteria.

Results

The PF occurrence rate (grade B or C) was 25.3% in 75 patients. A multivariate analysis identified a larger HU_SD (odds ratio 3.092; 95% CI 1.018–9.394) and higher amylase concentration in drainage fluid on postoperative day 1 (odds ratio 1.0001; 95% CI 1.00001–1.00022) as significant risk factors for PF.

Conclusions

The HU_SD of preoperative CT attenuation values in the pancreatic parenchyma was found to be an independent predictor for PF after PD and it might therefore positively contribute to the perioperative management of PD.

https://ift.tt/2rke4ZB