RSNA 2014 

Abstract Archives of the RSNA, 2014


SSK06-07

Multi-detector CT Findings of Pancreatic Fistula after Pancreaticoduodenectomy: Correlation with Surgical Grading

Scientific Papers

Presented on December 3, 2014
Presented as part of SSK06: Gastrointestinal (Pancreas Benign Disease)

Participants

Jin Woong Kim MD, Abstract Co-Author: Nothing to Disclose
Sang Soo Shin MD, Abstract Co-Author: Nothing to Disclose
Suk Hee Heo MD, Presenter: Nothing to Disclose
Hyo Soon Lim MD, Abstract Co-Author: Nothing to Disclose
Yong-Yeon Jeong MD, Abstract Co-Author: Nothing to Disclose
Heoung-Keun Kang MD, Abstract Co-Author: Nothing to Disclose

PURPOSE

To correlate multi-detector CT (MDCT) findings of pancreatic fistula after pancreaticoduodenectomy with surgical grading based on International StudyGroup of Pancreatic Fistula (ISGPF) classification scheme.

METHOD AND MATERIALS

A total of 142 consecutive patients (86 men, 56 women; mean age, 65.8 years),who underwent pancreaticoduodenectomy (pylorus preserving pancreaticoduodenectomy, n=114; Whipple’s operation, n=28) due to suspected periamupullary tumor, and postoperative MDCT, were included in this study. Patients were classified as four groups (no fistula,grade A fistula, grade Bfistula and grade Cfistula) according to ISGPF classification. Pancreatic fistula was defined to be present when there was a gap greater than 2 mm between pancreas and jejunum at pancreaticojejunostomy (PJ). In cases without pancreatic fistula on MDCT images, diameter of fistula was considered to be 0. Among 4 groups, MDCT images were compared in consensus by two radiologists regarding presence of pancreatic fistula at PJ, diameter of fistula, presence of pancreatic parenchymal defect around PJ, postoperative complications (including regional, vascular, intestinal, and biliary complications) using Chi-square test and one-way ANOVA statistics.

RESULTS

Regarding surgical grading, 142 patients were classified as no fistula (n=34), grade A fistula (n=66), grade B fistula (n=20) and grade C fistula (n=22) group. There was no significant difference in patients’ demographics among 4 groups (P>0.05). MDCT findings were significantly different among 4 groups regarding presence of pancreaticfistula at PJ (P< 0.001), presence of pancreatic parenchymal defect around PJ (P< 0.001), incidence oftotal complications (P< 0.001), vascular complications (P = 0.0038), and regional complications (P = 0.004). The diameter of fistula at PJshowed the trend to significantly increase from no fistula group(mean: 0.29 mm) to grade C fistula group (mean: 4.27 mm) (P< 0.001).

CONCLUSION

In regard to the presence of pancreatic fistula, postoperative MDCT findings were well correlated with surgical grading based on ISGPF classification scheme.

CLINICAL RELEVANCE/APPLICATION

MDCT could provide reliable information to suggest the presence of pancreatic fistula after pancreaticoduodenectomy, which is widely regarded as the most ominous of complications following pancreatic resection.

Cite This Abstract

Kim, J, Shin, S, Heo, S, Lim, H, Jeong, Y, Kang, H, Multi-detector CT Findings of Pancreatic Fistula after Pancreaticoduodenectomy: Correlation with Surgical Grading.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14003261.html