Endovascular management of delayed post-pancreatectomy haemorrhage.

OBJECTIVES: To assess the patient outcome after endovascular treatment of delayed post-pancreatectomy haemorrhage (PPH) as first-line treatment. METHODS: Between January 2005 and November 2013, all consecutive patients referred for endovascular treatment of PPH were included. Active bleeding, pseudoaneurysms, collections and the involved artery were recorded on pretreatment CT. Endovascular procedures were classified as technical success (source of bleeding identified on angiogram and treated), technical failure (source of bleeding identified but incompletely treated) and abstention (no abnormality identified, no treatment performed). Factors associated with rebleeding were analysed. RESULTS: Sixty-nine patients (53 men) were included (mean 59 years old (32-75)). Pretreatment CT showed 27 (39 %) active bleeding. In 22 (32 %) cases, no involved artery was identified. Technical success, failure and abstention were observed in 48 (70 %), 9 (13 %) and 12 patients (17 %), respectively. Thirty patients (43 %) experienced rebleeding. Rebleeding rates were 29 %, 58 % and 100 % in case of success, abstention and failure (p < 0.001). Treatment failure/abstention was the only factor associated with rebleeding. Overall, 74 % of the patients were successfully treated by endovascular procedure(s) alone. CONCLUSION: After a first endovascular procedure for PPH, the rebleeding rate is high and depends upon the success of the procedure. Most patients are successfully treated by endovascular approach(es) alone. KEY POINTS: • After a first endovascular procedure for PPH the rebleeding rate is high • The rebleeding rate is significantly associated with initial technical success • Three-quarters of the patients are successfully treated by endovascular procedure(s) alone.