Empiric vs. Targeted Embolization in Postpancreatectomy Hemorrhage

Transcatheter Arterial Embolization for Delayed Postpancreatectomy Hemorrhage: A Retrospective Study of 312 Patients

Clinical Question 

Is empiric embolization as effective and safe as targeted embolization for treating delayed postpancreatectomy hemorrhage?

Take Away Point 

Empiric embolization is safe and equally effective as targeted embolization in managing delayed postpancreatectomy hemorrhage, showing no significant difference in clinical success and 30-day mortality rates between the groups.

Reference 

Tan, Wenle, Yuan, Kai, Ji, Kan, Xiang, Tao, Xin, Hainan, Li, Xiaohui, Zhang, Wenhe, Song, Zhenfei, Wang, Maoqiang, & Duan, Feng. (2024). Transcatheter Arterial Embolization for Delayed Postpancreatectomy Hemorrhage: A Retrospective Study of 312 Patients. Journal of Vascular and Interventional Radiology, 35, 241–250. Click here for abstract

Study Design 

Retrospective observational cohort study

Funding Source 

Supported by the National Natural Science Foundation of China (No. 82172037)

Setting 

Single-center study at the Chinese PLA General Hospital, Beijing, People’s Republic of China.

Figure 

Figure 2A 66-year-old woman with adenocarcinoma of the duodenal papilla underwent pancreaticoduodenectomy. The patient developed severe abdominal bleeding 46 days after surgery with the hemoglobin concentration decreasing ≥3 g/dL. (a) Contrast-enhanced computed tomography showed a pseudoaneurysm of the hepatic artery. (b) Angiography showed a pseudoaneurysm of the hepatic artery. (c, d) The common hepatic artery was embolized with microcoils and gelatin sponge. Hemostasis was successful after targeted embolization.

Summary 

Conducted at the Chinese PLA General Hospital, this study retrospectively analyzed 312 patients with delayed postpancreatectomy hemorrhage treated from January 2012 to August 2022, comparing the outcomes of empiric embolization (EE) versus targeted embolization (TE).

Of the 312 patients, 185 patients were found to have positive digital subtraction angiography findings and 175 of them underwent targeted embolization based on the findings. Specifically, the common hepatic artery (30.3%), branches of the superior mesenteric artery (21.7%), splenic artery (20.0%) and left gastric artery (19.4%) were found to be the most frequently embolized arteries in the targeted embolization group.

No extravasation or any vascular abnormalities were detected in 137 (43.9%) patients by digital subtraction angiography. Of these patients, 68 patients were subjected to empiric embolization of the assumed ruptured arteries (EE group), and 69 patients received conservative treatment (NE group). In the cases with negative-result digital subtraction angiography, the principles for empirical embolization were as follows:

(a) computed tomography angiography before angiography highly suspected arterial bleeding;

(b) hemodynamic instability due to acute blood loss;

(c) patients with high suspicion of arterial bleeding according to the character and speed of bleeding;

(d) according to the intraoperative conditions of pancreatectomy, empirical embolization should be performed for bleeding arteries highly suspected by pancreatic surgeons without affecting the main functions of the organs.

In the empiric embolization group, the left gastric artery (42.6%), common hepatic artery (19.1%), splenic artery (14.7%), and stump of gastroduodenal artery (11.8%) were found to be the most frequently embolized arteries. In the absence of clear indications of hemorrhage, it was generally not recommended to perform empirical embolization on branches of the superior mesenteric artery for the purpose of preventing intestinal necrosis.

The clinical failure rates of targeted embolization, empiric embolization, and no embolization groups were 29.7%, 30.9%, and 49.3%, respectively. 243 cases in targeted embolization and empiric embolization groups were included in the follow-up prognostic analysis of this study. The 30-day mortality rates were 14.9% (26/ 175) and 10.3% (7/68) in the targeted embolization and empiric embolization groups, respectively. Embolization-related adverse events occurred in 19 patients. Six patients experienced liver abscesses (targeted embolization group, n = 5; empiric embolization group, n = 1), 1 of whom died of septic shock. Five patients in the targeted embolization group experienced splenic abscesses, and 1 patient in the empiric embolization group showed an abscess in the tail of the pancreas. A patient in the targeted embolization group who presented with ischemic necrosis of the bowel after embolization of the superior mesenteric artery branch underwent laparotomy.

The study found no significant differences in clinical success or 30-day mortality rates between empiric embolization and targeted embolization, demonstrating that empiric embolization can be a viable option in cases with angiographically negative findings. Factors such as malignant disease, Grade C pancreatic fistula, intra-abdominal infection, and concurrent extraluminal and intraluminal hemorrhage were identified as risk indicators for clinical failure. Advanced age and intra-abdominal infection were risk factors for 30-day mortality.

Commentary 

The findings underscore the viability of empiric embolization as an effective treatment strategy for delayed postpancreatectomy hemorrhage, particularly in challenging cases where targeted embolization is not feasible due to negative angiographic findings. This study contributes valuable insights into the management of a complex and high-risk patient population with low clinical success rate of 70% and mortality of 13.6%, suggesting that treatment approach can be tailored based on specific clinical scenarios without compromising patient outcomes. Future prospective studies could further validate these findings and refine treatment protocols