Percutaneous transhepatic cholangial drainage (PTCD) has been proven to be one of the most important treatments for benign and malignant obstructive jaundice. However, severe hemobilia can occur after PTCD, consequently causing life-threatening hemorrhagic shock, which has a high mortality rate (nearly 50%).¹ In a retrospective study performed by Rivera-Sanfeliz et al, the incidence of severe hemobilia was 2.31% in 346 patients receiving PTCD.² Damages or abnormal connections of intrahepatic vessels resulting from PTCD, namely, arterio-biliary fistula, biliary venous fistula, or pseudoaneurysm are the major causes of this hemobilia.³ Additionally, coagulation disorder in patients with obstructive jaundice is also a recognized cause.⁴

For clinicians, the difficulties in diagnosis and treatment of severe hemobilia following PTCD lie in detecting the specific site of bleeding and choosing a correct surgical method. In recent years, superselective angiography has been developed as an effective method for the diagnosis of hemobilia, which is able to detect significant hemobilia in over 90% of patients.⁵ Surgery, the predominant solution for the treatment of this bleeding, can effectively treat the primary lesions and completely remove the blood clots from the biliary in order to avoid obstruction or stone formation. Most patients with severe hemobilia after PTCD, however, have poor general health and unstable vital signs, greatly impairing their tolerance to anesthesia and surgery for the treatment of the hemobilia.^5,6 Thus, surgery may lead to exacerbation of the disease or even death of the patient during the hemostatic treatment. Therefore, an alternative solution to effectively and safely stop severe hemobilia following PTCD must be available to clinicians.

Interventional therapy has been tested as an effective approach for treating bleeding associated with severe hemobilia.^7,8 Nevertheless, few reports about the efficacy and safety of this therapy exist in the literature. In our hospital, from May 2003 to May 2010, 9 patients with severe hemobilia following PTCD received interventional therapy. The clinical results in these patients are presented in this report.

Materials and Methods

Patient population

Nine patients were included. Five were male (55.6%) and 4 were female (44.4%). The median age was 46 years (range, 22–79 years). All patients had obstructive jaundice and received PTCD. The etiology of obstructive jaundice was as follows: cholangiocarcinoma (n = 3), pancreatic carcinoma (n = 2), metastatic gastric carcinoma (n = 2), and choledocholithiasis (n = 2). Common clinical presentations of hemobilia after PTCD included persistent or recurrent abdominal pain, upper gastrointestinal bleeding, intermittent massive hemorrhagic bile from the drainage tube, no alleviation of jaundice and possibly even aggravation, hemorrhagic shock, fever, and melena (Table 1). Additional signs of severe biliary tract infection included recurrent high fever (>103 F°) and chills, observed in 1 patient. No patients were able to tolerate surgery owing to poor general conditions.

Table 1  The outcomes of interventional therapy for the treatment of severe hemobilia after PTCD in all patients

View Fullscreen

Results

The outcomes of all patients after interventional therapy are summarized in Table 1. The angiographies of pre-intervention and post-intervention in the representative patients with hepatic arterio-biliary fistula, pseudoaneurysm, and biliary-portal vein fistula are shown in Fig. 1. Three- and 6-month follow-up examinations showed that absolute hemostasis was achieved in all patients, and no severe complications were observed during the follow-up period.

View Figure

• Download as Powerpoint
• Download Figure

Discussion

Selective hepatic arteriography is considered the preferred method for diagnosis of severe hemobilia because of its high sensitivity, being capable of immediately and accurately detecting bleeding at such low rates as approximately 0.5 to 2 mL/min.⁹ The diagnostic accuracy has been reported to be approximately 85% to 95%. Cholangiography, performed by introducing angiographic media through the biliary drainage tube, is mainly used to detect biliary-portal vein fistula but is another way to diagnose hemobilia. Therefore, for the patients with severe hemobilia after PTCD, in parallel to the immediate supportive treatments such as blood transfusion, hepatic arteriography and cholangiography should be conducted as early as possible to detect the primary hemorrhage site. Ultrasonography, computed tomography (CT), and duodenoscopy are potential alternatives in cases where the selective hepatic arteriography and cholangiography are not available, but the bleeding lesion is usually difficult to position accurately with these methods.¹⁰

In recent years, superselective hepatic artery embolization has been accepted as an effective interventional therapy for the treatment of severe hemobilia.⁵ Its efficacy and safety have been evaluated by several studies,^8,9 indicating it can offer the same advantages as minimally invasive therapy such as a smaller incision, fewer complications, and a faster recovery; thus it is more appropriate for patients with poor general health and unstable vital signs, specifically patients with hemorrhagic shock. Embolizing both the distal and proximal parts of the bleeding artery simultaneously is currently recognized as the standard procedure by most researchers and clinicians because it effectively prevents recurrent hemorrhage caused by the countercurrent flow of blood through a branch to the bleeding artery as well as controls the initial hemorrhage.⁵

The materials used for embolization include gelatin sponge, absolute ethanol, n-butyl-2-cyanoacrylate (NBCA), detachable balloon, and coils. Gelatin sponge and coils, the most commonly used materials, usually are applied in combination because embolization with only gelatin sponge rarely achieves absolute hemostasis and easily induces recurrent hemorrhage.^11,12 Therefore, for the patients with hepatic arterio-biliary fistula and pseudoaneurysm in our study, gelatin sponge and platinum coil were combined for hepatic artery embolization, and absolute hemostasis was achieved in all patients after 3- and 6-month follow-up. However, it should be noted that Glubran 2 acrylic glue was also used in a patient with pseudoaneurysm in addition to the combination of gelatin sponge and platinum coil. Glubran 2 acrylic glue, as a liquid nonirritating embolization agent with low viscosity, is able to enter and embolize the small and circuitous bleeding artery branch in the cavity of pseudoaneurysm, effectively preventing compensatory blood supply and thus resulting in absolute hemostasis. In this patient, another small branch adjacent to the pseudoaneurysm was detected to be responsible for the recurrent hemobilia after the embolization with gelatin sponge and coil. Considering the thin diameter and circuitous nature of this branch, Glubran 2 acrylic glue was applied in order to achieve absolute hemostasis.

Only one patient was diagnosed as having biliary-portal vein fistula, which has a very low incidence rate in patients receiving PTCD. Surgery is the generally accepted treatment for this complication, though there are few reports on its efficacy. In our study, this patient received the embolization with a biliary fully covered stent, and absolute hemostasis was obtained after the embolization without any recurrent hemobilia at 3- and 6-month follow-up. According to the authors' experience, biliary-portal vein fistula often occurs between the choledochus and the main portal vein, thus allowing the placement of a biliary stent for embolizing orificium fistulae. Therefore, considering its efficacy in controlling hemorrhage, embolization with a biliary stent may be recognized as a new and effective method for the treatment of biliary-portal vein fistula. Nevertheless, further studies and more cases are needed to investigate and confirm the clinical efficacy of this approach.

In our study, the complications associated with embolization included transient abdominal pain, fever, and elevation of liver aminotransferases, which could be alleviated by symptomatic treatment. More serious complications, including necrosis in the embolization area, hepatic failure, ectopic embolization, gastrointestinal hemorrhage, and septicemia, have been reported by previous studies but were not observed in patients of our study.¹³ In conclusion, our results indicate that interventional therapy is an effective and safe method for the treatment of severe hemobilia after PTCD because of the advantages including a smaller incision, fewer complications, and faster recovery. We conclude that it should be used as a preferable approach for the patients with poor general health conditions.