RESUMO
Background: Postoperative hemorrhage is a potentially lethal complication of pancreatoduodenectomy. This study reports on the use of endovascular hepatic artery stents in the management of postpancreatectomy hemorrhage. Methods: This is a retrospective analysis of a prospectively maintained, consecutive dataset of 440 patients undergoing pancreatoduodenectomy over 68 months. Data are presented on bleeding events and outcomes, and contextualized by the clinical course of the denominator population. International Study Group of Pancreatic Surgery terminology was used to define postpancreatectomy hemorrhage. Results: Sixty-seven (15%) had postoperative hemorrhage. Fifty (75%) were male and this gender difference was significant (P = 0.001; 2 proportions test). Postoperative pancreatic fistulas were more frequent in the postoperative hemorrhage group (P = 0.029; 2 proportions test). The median (interquartile range [IQR]) delay between surgery and postoperative hemorrhage was 5 days (2-14 days). Twenty-six (39%) required intervention comprising reoperation alone in 12, embolization alone in 5, and endovascular hepatic artery stent deployment in 5. Four further patients underwent more than 1 intervention with 2 of these having stents. Endovascular stent placement achieved initial hemostasis in 5 of 7 (72%). Follow-up was for a median (IQR) of 199 days (145-400 days) poststent placement. In 2 patients, the stent remained patent at last follow-up. The remaining 5 stents occluded with a median (IQR) period of proven patency of 10 days (8-22 days). Conclusions: This study shows that in the specific setting of postpancreatoduodenectomy hemorrhage with either a short remnant gastroduodenal artery bleed or a direct bleed from the hepatic artery, where embolization risks occlusion with compromise of liver arterial inflow, endovascular hepatic artery stent is an important hemostatic option but is associated with a high risk of subsequent graft occlusion.
RESUMO
PURPOSE: To assess T2 HASTE MR in acute abdominal imaging and ascertain if it is a reliable alternative to CT in patients under 60. METHOD AND MATERIALS: In a prospective diagnostic performance study from January 2009 to December 2013, patients under 60 presenting with acute abdominal pain, that required imaging following surgical review, were imaged with T2 HASTE MR. Rapid acquisition HASTE (Half Fourier Acquisition Single Shot Turbo Spin Echo) coronal and axial sequences were obtained, without intravenous contrast. Patients were followed up clinically for a minimum of 3 months. RESULTS: 468 cases included in the study. 349 were negative for acute abdominal pathology, 116 positive for acute abdominal pathology and 3 were indeterminate. In the MR positive group (n=116), 64 had surgery confirming findings (34 appendicitis, 14 SBO, 3 ovarian torsion, 3 LBO, intussusception, ovarian carcinoma, ovarian dermoid, 2 pelvic inflammatory disease, diverticular abscess, crohns, 4 endoscopy for acute bowel pathology) while 51 were managed conservatively with concordant follow up (4 SBO, 11 diverticulitis, 6 pelvic inflammatory disease, 7 inflammatory bowel disease, 7 colitis, 6 pyelonephritis, 2 cholecystitis, renal abscess, pseudomembranous colitis, splenic haematoma, mesenteric adenitis, 2 pancreatitis, lymphoma, epiploic appendagitis). 1 patient had an MR diagnosis of appendicitis but at laparoscopy a sigmoid diverticular perforation was diagnosed and the appendix was normal. In the MR negative group (n=349), 324 had uneventful follow-up, 22 had negative laparoscopies, while 3 had subsequent appendectomies, with appendicitis on histology (3 days, 10 days and 2 months post scan). In the MR indeterminate group (n=3), one was treated conservatively with uneventful follow up, one had laparoscopic appendectomy with normal appendix on histology, one had laparoscopic appendectomy with acute appendicitis on histology. When MR correlated with clinical follow up (n=468), overall diagnostic accuracy is 99% (463/468). When MR findings correlated with direct visualisation at surgery/endoscopy (n=90), sensitivity is 98% (95% CI) and specificity is 92% (95% CI). CONCLUSION: This study demonstrates that rapid acquisition axial and coronal T2 HASTE MR is a practical, safe and effective method in the diagnosis of acute abdominal pain. MR is the preferred option to CT in patients of an age prone to radiation with a potential surgical diagnosis. CLINICAL RELEVANCE/APPLICATION: MRI in acute abdominal imaging is both effective and practical and is the preferred imaging option in patients of an age prone to radiation with a potential surgical diagnosis.
