Biliary complications after liver transplantation (LT) have an incidence of 10% to 30%1; reduced-size LT, split-liver transplantation (SLT), and living-donor LT are associated with increased rates of biliary complications.2 A lower rate of complications directly related to hepaticojejunostomy (HJA) versus duct-to-duct anastomosis has been described.3–5 Surgical repair and endoscopic management have been reported to successfully treat post-LT biliary fistulas from duct-to-duct anastomosis, but when the biliary leaks complicate HJA or they rise from the cut surface of the liver, an endoscopic approach is not possible. The aim of this study was to evaluate the safety and efficacy of percutaneous treatment of post-LT biliary leakages occurring in HJA or in split-liver recipients in the absence of biliary tree dilatation.
Biliary leaks complicating hepaticojejunostomy (HJA) or fistulas from cut surface are severe complications after liver transplantation (LT) and split-liver transplantation (SLT). The aim of the study was to describe our experience about the safety and efficacy of radiological percutaneous treatment without dilatation of intrahepatic biliary ducts. From 1990 to 2006, 1595 LTs in 1463 patients were performed in our center. In 1199 LTs (75.2%), a duct-to-duct anastomosis was performed, and in 396 (24.8%), an HJA was performed. One hundred twenty-nine anastomotic or cut-surface bile leakages occurred in 115 patients. Sixty-two biliary leaks occurred in 54 patients with HJA; in 48 cases, an anastomotic fistula was found. Cut-surface fistulas occurred in 14 cases: 5 in right SLTs and 5 in left SLTs. Twenty-two patients were treated with 23 percutaneous approaches for 17 HJA fistulas and 6 cut-surface leaks without intrahepatic bile duct dilatation. Two percutaneous therapeutic approaches were used: percutaneous transhepatic biliary drainage (PTBD) for fistula alone and PTBD with percutaneous drainage of biliary collection in patients with both complications. PTBD was successful in 21 cases (91.3%); the median delay from catheter insertion and leak resolution was 10.3 days (range: 7–41). The median maintenance of drainage was 14.8 days. In 1 patient, fistula recurrence after PTBD needed a surgical approach; after that, an anastomotic fistula was still found, and a new PTBD was successfully performed. In another patient, PTBD was immediately followed by retransplantation for portal vein thrombosis. There were no complications related to the interventional procedure. In conclusion, biliary fistulas after HJA in LT or after SLT can be successfully treated by PTBD. The absence of enlarged intrahepatic biliary ducts should not be a contraindication for percutaneous treatment. Liver Transpl 14:611–615, 2008. © 2008 AASLD.
PATIENTS AND METHODS
From 1990 to 2006, 1595 LTs in 1463 patients were performed in our center; in 1199 LTs (75%), a duct-to-duct anastomosis was performed, whereas in 396 (25%), an HJA was preferred.
One hundred twenty-nine (8.1%) anastomotic or cut-surface bile leakages occurred in 115 patients; of 396 HJAs, 62 (15.6%) biliary leaks occurred in 54 patients. In 48 cases (77.4%), an anastomotic fistula was found, and in 14 cases (22.6%), cut-surface fistulas were found. Cut-surface fistulas occurred after right SLT (5 cases), after left SLT (5 cases), after living-donor LT (2 cases), after reduced-size LT (1 case), and after hepatectomy following LT (1 case). The complication was associated with a percutaneous biliary leak in 2 patients and with abdominal biloma in 4 patients.
Of 62 biliary leaks in HJA, 22 cases were treated with 23 percutaneous approaches for 17 HJA fistulas and 6 cut-surface leaks or abscess, despite the absence of biliary tree dilatation (we considered intrahepatic ducts with a diameter over 1.7 mm to be “enlarged”).6 These 22 patients represent all cases of not enlarged biliary ducts, whereas the remaining 40 patients had enlarged intrahepatic ducts; however, in all patients with post-LT biliary leaks (with enlarged or not enlarged ducts), we attempted percutaneous treatment because in our institution surgery represents the second therapeutic choice.
Biliary fistula was clinically diagnosed on the basis of the following:
Clinical findings (fever, leukocytosis, increased cholestasis indexes, abdominal pain, peritonitis, sepsis, or recurrent cholangitis).
Evidence of biliary leakage from surgical drain.
Evidence of perianastomotic fluid collection at ultrasound evaluation or magnetic resonance cholangiopancreatography.
The suspicion was radiologically confirmed by percutaneous cholangiography.
All patients followed the same diagnostic protocol, and the interventional procedures were performed by the same radiologist.
Two radiological percutaneous therapeutic approaches were used: percutaneous transhepatic biliary drainage (PTBD) for fistula alone and PTBD with percutaneous drainage of biliary collection in patients with both fistula and biliary collection.
All patients underwent percutaneous transhepatic biliary cholangiography after mild sedation (for example, n-butyl-bromide and ketorolac trometamine) and local anesthesia (2% lidocaine chloridrate). We always used a 22-gauge Chiba needle for sterile cannulation of a peripheral biliary radicle. After the peripheral bile duct was punctured, a guide was advanced through the anastomosis into the jejunal limb and was followed by insertion of an 8-French external-internal biliary drainage catheter (Meditech Soft Drain, Boston Scientific, Natick, MA). The catheter, with 2 series of side holes separated by a blind segment, was positioned within the bile duct above the fistula and the distal hole in the jejunal limb in order to obtain a complete exclusion of the bile leak. The percutaneous biliary fistulas were treated by embolization with fibrin sponge (Spongostan), whereas subphrenic or subhepatic collections were drained by percutaneous drainage. We performed a cholangiography at the seventh day after the drainage insertion and every 7 days afterward. The catheter was removed when cholangiography showed the complete resolution of the fistula.
PTBD was successfully performed in all patients at the first procedural sitting; therefore, sometimes several attempts were required to successfully locate and cannulate the biliary tree.
The median length of biliary drainage maintenance was 14.8 days (range: 7–74).
In 21 cases (91.3%), leakage resolved after biliary drainage (median: 10.3 days, range: 7–41 days). In 1 patient, fistula recurrence after PTBD needed a surgical approach, and HJA was redone 10 days after the percutaneous drainage because the patient became symptomatic (fever and abdominal pain) and radiological control evidenced a high-output leak; after the second HJA, an anastomotic fistula was still found, and a new biliary drainage was successfully performed. Another patient, despite the complete resolution of the biliary leakage after 7 days, underwent surgical drainage of a subhepatic infected biloma. In a last case, PTBD was immediately followed by retransplantation for portal vein and hepatic artery thrombosis associated with biliary anastomotic fistula.
In 2 cases of fistula associated with biliary abscess, the drainage was maintained for 29 and 74 days.
Bile leaks evidenced by minimal injection pressure of contrast were considered high-output leaks; the majority of the patient in our series had high-output leaks (Fig. 1A,B). In 1 patient, the anastomosis was almost completely disjointed; because of the critical general condition of the patient, PTBD was preferred as a first approach, and it obtained the complete resolution of the fistula (Fig. 2A,B).
Biliary injuries at cut surfaces varied from minimal leaks to high-output fistulas (Fig. 3A,B).
No complications related to percutaneous procedures required either surgical or radiological correction. The group of patients had a median follow-up of 1110 days after catheter removal; in only 1 case we found evidence of a biliary stricture, 6 years after percutaneous biliary drainage.
Biliary leakages complicate the early postoperative period of LT. HJA fistulas are frequently related to bile duct necrosis or ischemia7–11; uncommon sites of early bile leaks are the cut surfaces of split or reduced-size livers.
Biliary fistulas are detected from clinical signs such as leukocytosis, fever, abdominal pain, bile from surgical drain, and ultrasound or computed tomography scan evidence of abdominal collections.12–16
Traditionally, biliary leakages have been surgically corrected,17–19 but more recently, nonsurgical procedures, endoscopic and/or radiological techniques, have been suggested20–22; the different possibilities are generally related to the local availability of endoscopic and radiological expertise.
No prospective randomized trials comparing the results of surgical and nonsurgical management of post–orthotopic LT biliary complications have been reported, and actually there is still a debate regarding the best treatment. In a comparative study of surgical and nonsurgical treatments, Kuo et al.23 found that endoscopy and interventional radiology had worse results than surgery; although the difference was significant, it was a retrospective uncontrolled and nonrandomized study.
Biliary leakage from HJA cannot be treated with an endoscopic approach; the only therapeutic approaches for such complications are surgery and interventional radiological treatment,24 although they are performed according to the local opportunities rather than codified results. In several centers, surgical intervention is the treatment of choice for biliary fistulas at an HJA site, but the ischemia of the biliary tract must be accurately evaluated; to redo a biliary anastomosis without an adequate arterial blood supply can lead to a bad outcome.
Nevertheless, surgical re-exploration is often difficult because of the adhesions in the periportal area; another relevant limit of surgery is the anesthesiological management of these patients, who are sometimes clinically unstable.
Therefore, many authors have suggested the need for an alternative therapeutic approach.3, 25–27 Percutaneous transhepatic radiological treatment, performed under local anesthesia, can be a low-risk option to avoid unnecessary surgery, and it is better tolerated by transplanted patients.
Post-LT biliary fistulas are usually associated with not enlarged intrahepatic biliary ducts, and performing PTBD may be sometimes technically difficult; the last consideration makes it mandatory that the percutaneous approach should be performed only by experienced radiologists.28, 29
In major hepatic transplant centers worldwide, the treatment of choice (surgical or percutaneous) of post-LT biliary fistulas from HJA or a cut surface in split-liver recipients is based on the experience of each center rather than codified guidelines; several authors have explained that a therapeutic percutaneous transhepatic cholangiographic approach to HJA or cut-surface leaks, with placement of an internal-external drain accompanied by percutaneous drainage of fluid collections, is recommendable,2, 22, 30, 31 and it appears to be a safer approach than further surgery.32, 33
Potential risks of percutaneous treatment include minor complications (biloma, transient hemobilia, puncture-site hematoma, drainage obstruction, dislocation, and rupture) and major complications (sepsis, hemorrhage, fistulas, subcapsular hematoma, abscess, peritonitis, pancreatitis, cholangitis, and pseudoaneurysm). Sepsis and hemorrhage represent the most important complications, and the reported rates are 2.5%.34 The recommended overall procedure threshold for all major complications is 10%.
Actually, we think that percutaneous treatment of biliary leakages from HJA or a cut liver surface should also always precede the surgical approach in the absence of intrahepatic biliary duct dilatation. In our experience, although long-term maintenance of PTBD resulted in patient discomfort, the PTBD-related procedures were not associated with complications; moreover, fistulas healed in more than 90% of patients without need of surgery. The group of patients described in the study had a median follow-up of 1110 days after catheter removal; in only 1 case we found evidence of a biliary stricture, 6 years after percutaneous cholangiography. We do not think that the stricture was directly correlated to the procedure performed 6 years before. We did not find any report of large series of liver-transplanted patients defining the risk of biliary stricture following percutaneous treatment of bile leakage, but we believe that the key point is to avoid long retention of the drainage after the resolution of the fistula.
In conclusion, our results indicate that PTBD is a relevant technique in the treatment of biliary leakage from HJA post-LT or from a cut liver surface after SLT. The procedure allows us to obtain the complete healing of the leakage, and we suggest considering PTBD always before any surgical approach; the absence of intrahepatic biliary duct dilatation should not be a contraindication for the treatment.