| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Report |
1 Institut de Diagnòstic per la Imatge, Hospital Duran i Reynals, Autovia
de Castelldefels, Km. 2,7, 
Hospitalet de Llobregat 08907, Spain.
2 Liver Transplantation Unit, Hospital Universitari de Bellvitge, Barcelona,
Spain.
3 Department of Pathology, Hospital Universitari de Bellvitge, Barcelona,
Spain.
Received March 5, 2004;
accepted after revision June 30, 2004.
Address correspondence to C. Valls
(carlovalls{at}csub.scs.es).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. MRCP is a useful imaging procedure in the assessment of biliary complications after OLT.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Despite recent advances in surgical and preservation techniques, biliary complications after OLT remain a serious clinical problem that result in increased morbidity, liver dysfunction, and graft loss. Biliary complications are one of the leading causes of liver failure and occur in 10–30% of patients who undergo transplantation [1]. In addition, clinical and biologic signs are nonspecific, and imaging techniques are generally required to establish the diagnosis. Direct cholangiographic techniques such as ERCP or percutaneous transhepatic cholangiography (PTHC) are invasive and carry a relatively high rate of procedure-related complications. ERCP with sphincterotomy has a reported morbidity rate of 9.8% and a mortality rate of 2.3%; therefore it should not be considered a diagnostic tool but rather a therapeutic technique [2]. On the other hand, reported complications of PTHC include perihepatic hematoma, hemobilia, infection, and pancreatitis [3, 4]. Recently, the advent of MR cholangiopancreatography (MRCP) has significantly changed our approach to diagnosis and management of biliary complications after OLT because it allows noninvasive visualization of the biliary tree with exquisite anatomic detail.
The aim of our study was to evaluate the results of MRCP as the only imaging procedure in the diagnosis and management of biliary complications after OLT.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our final study group comprised 63 patients— 48 men and 15 women with a mean age of 53 years (range, 22 to 70 years). Liver transplantation procedures were performed between July 1992 and January 2003. Bile duct continuity was established in all cases by a primary duct-to-duct anastomosis (hepaticohepaticostomy) without T-tube splintage. The interval between OLT and the clinical onset of biliary complications ranged from 1 to 113 months (mean, 30 months).
Indications for transplantation were as follows: hepatitis C liver cirrhosis (n = 18), hepatitis B liver cirrhosis (n = 2), primary biliary cholangitis (n = 3), hepatocellular carcinoma (n = 17), sclerosing cholangitis (n = 1), alcoholic cirrhosis (n = 16), idiopathic cirrhosis (n = 3), Budd-Chiari syndrome (n = 1), familial amyloidotic polyneuropathy (n = 1), and cholangiocarcinoma (n = 1).
MRCP Technique
All MRCP images were acquired with a 1.5-T unit (Gyroscan Intera, Philips)
with a dedicated phased-array coil and high-performance gradients. Two
different MRCP snapshot techniques were applied: thick-slab single-shot turbo
spin-echo T2-weighted sequences and multisection thin-slab, single-shot turbo
spin-echo T2-weighted sequences.
Thick-slab sequences were acquired as a single 20–45 mm slice in coronal and oblique coronal orientations (echo spacing, 8.3 msec; effective TE, 1,000 msec; image matrix, 512 x 512; field of view, 350 mm). Thin-slab MRCP sequences were acquired with sequential 3-mm slices in the axial and coronal planes (echo spacing, 4.2 msec; effective TE, 183 msec; image matrix, 272 x 512; field of view, 385 mm). No patient preparation or sedation was required.
Image Interpretation and Analysis
MRCP images were prospectively assessed by three abdominal radiologists in
consensus. Images were evaluated to detect the presence of strictures of the
biliary tree, biliary dilatation, or intraluminal filling defects.
Intrahepatic bile ducts were considered dilated if the maximum diameter was
more than 3 mm. Extrahepatic bile ducts (common hepatic duct and common bile
duct) were considered dilated if measuring more than 8 mm. For biliary
strictures, the site (anastomotic or nonanastomotic) and length were
evaluated.
On the basis of MRCP findings, biliary complications were classified as one of the following: anastomotic stricture (focal stricture at the site of the biliary anastomosis), nonanastomotic stricture–ischemic cholangitis (long hilar stricture or signs of biliary necrosis and biloma), biliary stone or sludge (endoluminal hypointense filling defects), donor-to-recipient common bile duct disproportion (the diameter of the donor common bile duct is greater than the diameter of the recipient common bile duct; this postoperative finding may mimic anastomotic stricture, but for this study, it was considered a biliary complication), and other complications.
Diagnostic confirmation was obtained with PTHC in seven patients, ERCP in six patients, surgery and histologic study in 21 patients (primary retransplantation in 11 patients and hepaticojejunostomy in 10 patients), liver biopsy in 15 patients, and clinical follow-up in 14 patients in whom no treatment was performed. Direct cholangiography was performed after a positive result on MRCP except in two cases.
The 15 patients with normal MRCP findings and the group of patients with donor-to-recipient common bile duct disproportion (n = 7) had clinical follow-up that ranged between 2 and 48 months. An MRCP diagnosis of biliary complications in a patient with a final diagnosis of normal bile duct, including patients with donor-to-recipient common bile duct disproportion, was considered to be a false-positive result. An initial MRCP diagnosis of normal findings or a misdiagnosis in a patient with a final diagnosis of biliary complications was defined as a false-negative result. MRCP findings of a specific type of biliary complication that was confirmed either at surgery or on direct cholangiography were considered to be a true-positive finding.
The sensitivity, positive predictive value, and accuracy of MRCP for detection of biliary complications were calculated. Sensitivity was defined as the number of cases with MRCP-detected biliary complications divided by the number of final diagnoses of biliary complications. The positive predictive value was defined as the number of biliary complications correctly depicted on imaging divided by the total number of cases considered as a biliary complication on MRCP imaging. The MRCP accuracy was calculated as the number of patients correctly diagnosed on MRCP divided by the total number of patients in the study.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Anastomotic stricture.—This was the most frequent finding in our series accounting for 36.4% (16/44) of all biliary complications. Anastomotic strictures were detected on MRCP as short and very tight strictures at the level of the anastomosis with marked suprastenotic dilatation (Fig. 1). In 56.3% (9/16) of the cases, biliary stones or sludge were found concomitantly (Figs. 2A and 2B). MRCP findings led to the correct diagnosis for all anastomotic stenosis (16/16), with a sensitivity of 100%. There was one false-positive result in a patient with donor-to-recipient common bile duct disproportion (Figs. 3A, 3B, and 3C).
|
|
|
|
|
|
Anastomotic strictures were treated surgically with hepaticojejunostomy in nine cases and with interventional PTHC in seven cases: biliary stent (n = 6) or percutaneous transhepatic balloon dilatation (n = 1).
Nonanastomotic stricture due to ischemic cholangitis.—This condition was found in 25% (11/44) of all biliary complications. We actually found two patterns of this complication: the first, hilar stricture and the second, biliary necrosis and biloma. Hilar stricture (n = 7) was defined as a long stenosis including the right and left hepatic ducts, the biliary bifurcation, and occasionally the common hepatic duct (Figs. 4A and 4B). In 57.1% (4/7) of these patients, no arterial thrombosis was found, whereas in 42.9% (3/7) arterial thrombosis was present.
|
|
Biliary necrosis and biloma (n = 4) consisted of destruction of the bile ducts with formation of biliary lakes (biloma) (Figs. 5A and 5B). All patients with biliary necrosis and biloma had concomitant arterial thrombosis (Fig. 6). A correct diagnosis was possible with MRCP in 90.9% (10/11) of the cases. One false-negative finding occurred in a patient with a biliary cast involving the biliary bifurcation and the common hepatic duct without a definite stenosis on MRCP. The patient was misdiagnosed as having biliary stone or sludge without stenosis (Fig. 7A) and underwent a hepaticojejunostomy without a good clinical outcome. Because of the persistent impaired liver biology, MRCP was repeated and showed hilar stenosis and biloma consistent with ischemic cholangitis (Fig. 7B).
|
|
|
|
|
Patients with nonanastomotic stricture due to ischemic cholangitis were treated with retransplantation in nine cases. In two patients, retransplantation was not possible due to other concomitant medical problems (plastic peritonitis in one patient and serious psychiatric problems in another patient), and only follow-up was performed.
Biliary stones or sludge.—Biliary stones or sludge was found either as an isolated finding (n = 9) or, more frequently, a finding associated with other types of complications (n = 16) (Figs. 2A and 2B). MRCP accurately revealed 88.9% (8/9) of isolated biliary stones. A false-negative finding in a patient was eventually diagnosed with ERCP. Isolated biliary stones were treated either by ERCP (n = 6), surgery (n = 1), or medical treatment (n = 2).
Donor-to-recipient common bile duct disproportion.—Common bile duct disproportion was also a relatively frequent finding in our series, accounting for 18.2% (8/44) of all biliary complications. Although not strictly a biliary complication, common bile duct disproportion can have misleading appearances, mimicking biliary stenosis. The donor common bile duct is thin and smooth, and the recipient common bile duct is slightly enlarged (Fig. 8A). Comparison with intraoperative cholangiographic images is helpful to confirm stability (Fig. 8B). MRCP revealed 87.5% (7/8) of the common bile duct disproportions. One false-negative finding occurred in a patient who was misdiagnosed as having anastomotic stenosis on MRCP but who had normal findings on PTHC (Figs. 3A, 3B, and 3C). This patient had a final diagnosis of recurrent hepatitis C virus at liver biopsy.
|
|
Global Results
MRCP findings were correlated with direct cholangiography, surgery, liver
biopsy, or follow-up. Twenty-two patients had an MRCP examination that was
prospectively considered to show normal findings: 15 patients had completely
normal findings and seven had donor-to-recipient common bile duct
disproportion (one of these was a false-positive finding of biliary
complication that was misdiagnosed as anastomotic stenosis). In the 15
patients in whom MRCP did not disclose any abnormality, final diagnoses were
absence of any abnormality confirmed by clinical follow-up (n = 3),
rejection of the transplanted liver or recurrent hepatitis virus C infection
diagnosed by biopsy (n = 11), and one false-negative finding that was
eventually confirmed as common bile duct stone on ERCP.
There were 43 diagnoses of biliary complication on the basis of MRCP, 41 of which were eventually confirmed as true-positive diagnoses and one as a false-positive diagnosis of biliary complication. In one patient, MRCP allowed correct diagnosis of biliary complication but not the correct type of complication (ischemic cholangitis misdiagnosed as biliary cast).
The one false-positive finding on MRCP was a case of donor-to-recipient common bile duct disproportion that was considered to be an anastomotic stricture on MRCP but showed normal bile ducts on transhepatic cholangiography. A final diagnosis of recurrent hepatitis C virus was determined at biopsy.
We found two cases with false-negative findings on MRCP. One patient was considered to have normal findings on MRCP, but the persistence of symptoms led to performance of ERCP, which disclosed a distal common bile duct stone. The other false-negative case involved the previously described patient in whom ischemic cholangitis was initially misdiagnosed as biliary cast.
All in all, MRCP had a sensitivity of 95.3% (41/43) in the diagnosis of biliary complications after OLT. The positive predictive value was 97.6% (41/42), and the false-negative ratio was 4.54% (2/44). There was only one false-positive, accounting for a false-positive ratio of 2.27% (1/44), and the overall diagnostic accuracy was 95.2% (60/63).
MRCP imaging alone was able to provide a specific diagnosis in 96.8% (61/63) of the patients and ERCP and PTHC were required in only two patients (3.2%, 2/63). Direct cholangiography was required as a therapeutic procedure in 22.2% (14/63) of the patients.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Diagnosis of late biliary complications after liver transplantation is challenging because clinical and biologic findings are subtle or even absent in the early stages of the disease. In addition, evaluation of the bile ducts with conventional direct cholangiographic techniques has an unacceptably high complication rate for use in patients with low clinical suspicion [2]. The advent of new technologic advances in MRI has dramatically changed the diagnosis of biliary complications. New generation MRI units with high-performance gradients and phased-array coils allow high-quality heavily weighted T2 images to be acquired with increased spatial resolution in less than 4 sec, therefore eliminating most motion-related artifacts [9]. MRCP is therefore especially suited for imaging biliary tree complications without the inherent risks related to direct cholangiography [10, 11]. Due to its lack of invasiveness and side effects, in our institution, MRCP is performed almost routinely in the transplant recipients with abnormal results on liver function tests to rule out or confirm a biliary complication.
According to previous surgical reports, late biliary complications after liver transplantation include strictures, obstruction (stones, debris or sludge formation, or cystic duct mucoceles), and ampullary dysfunction [5]. Strictures are the most important late biliary complications after transplantation. In the series of Greif et al. [5], biliary strictures represented 46% of biliary complications. Most biliary strictures result in obstructive jaundice or altered liver function.
Biliary strictures can be subdivided in anastomotic and nonanastomotic strictures. Anastomotic strictures seem to be related to technical factors such as postoperative biliary fistula and fibrosis at the site of the duct-to-duct anastomosis, although ischemia may also be a contributing factor [12]. Fibrotic changes in the anastomosis lead to scar formation, retraction, and marked narrowing of the lumen, which is usually short. Anastomotic strictures are the most frequent type of strictures in transplant recipients accounting for 82–87% of all biliary stenosis [5, 11]. In our series, biliary stenosis was found in 42.9% (27/63) of the patients, but the incidence of anastomotic strictures was only 59.3% (16/27). The remaining eleven patients had intrahepatic nonanastomotic strictures. On MRCP, anastomotic strictures tend to be focal and appear as an abrupt luminal narrowing with proximal duct dilatation. In the series of Laghi et al. [11], MRCP with maximum-intensity-projection technique correctly depicted all anastomotic strictures and the dilated bile ducts above the stricture, which is usually difficult to assess with direct cholangiography. In our series, anastomotic strictures accounted for 36.4% (16/44) of all biliary complications. MRCP was used to correctly diagnose all anastomotic stenosis (16/16) with a sensitivity of 100%. We only had one false-positive finding in a patient with donor-to-recipient common bile duct disproportion.
In contradistinction to extrahepatic anastomotic stenoses, intrahepatic nonanastomotic stenoses are less well understood, and imaging features have not been adequately described. Nonanastomotic, or ischemic-type, biliary strictures usually present as intrahepatic strictures, implying loss of graft if the lesion continues [13, 14]. Hepatic arterial thrombosis, long ischemic time, and inadequate exposure of the biliary epithelium to the preservation solution have been associated with these lesions. However, although hepatic thrombosis has been directly related to some cases of nonanastomotic stricture, patients with a normal arterial vessel may show similar imaging features in the biliary tree. In the series of 21 patients with nonanastomotic strictures studied by Zajko et al. [12], only 57% had thrombosis of the hepatic artery. In a more recent series from the same group, only 27% of the patients with nonanastomotic strictures had associated hepatic artery occlusion [15].
In our series, 63% (7/11) of all nonanastomotic strictures had associated arterial thrombosis. However, we actually found two different patterns of nonanastomotic stricture: hilar stricture and biliary necrosis and biloma. Hilar stricture was found in 63.6% (7/11) of the cases and consisted of a long stenosis including the right and left hepatic ducts and the biliary bifurcation. Only 42% of these patients had concomitant hepatic artery occlusion. Biliary necrosis and biloma accounted for the remaining 37% (4/11) of the cases and consisted of destruction of the bile ducts with formation of biliary lakes (biloma). All patients with biliary necrosis and biloma had concomitant arterial thrombosis. Our results with MRCP are not substantially different from the imaging findings reported by Zajko et al. in 1987 [12] with direct cholangiography, In this series, nonanastomotic leakage consistent with biloma was found in 52% of the patients, whereas isolated nonanastomotic stricture was found in 38.7% of the cases. However, in that study, only patients with known arterial occlusion were included. In the series of Orons et al. [16], the prevalence of biliary complications was markedly increased in patients with hepatic artery stenosis. Interestingly almost 50% of the patients in that series had nonanastomotic stenosis, but no cases of biloma were reported. It seems, therefore, that there is a wide spectrum of clinical and cholangiographic features in patients with nonanastomotic strictures due to ischemic cholangitis. Clinical presentation of hepatic artery occlusion may range from fulminant hepatic necrosis to mild relapsing cholangitis. These different clinical situations may be related to the degree of decrease in the arterial blood flow and the possibilities of receiving blood flow from other collateral sources. Similarly, cholangiographic features in cases of ischemic cholangitis may range between massive biloma with necrosis of the bile ducts in cases of acute arterial occlusion to mild nonanastomotic strictures in cases of subacute decreased arterial blood flow. These subacute cases are not necessarily related to arterial thrombosis and may be due to arterial stenosis, postreperfusion injuries, ABO incompatibility, or long ischemic time [17, 18]. The routine use of MRCP in these patients allows us to perform retransplantation without previous biliary drainage or drainage of an infected biloma.
In cases with hilar nonanastomotic stenosis, biliary cast is usually an associated finding. Previous studies have suggested an association between ischemia and biliary debris formation [19]. In our series, 71.4% (5/7) of patients with hilar stenosis had associated biliary sludge.
Although biliary complications after OLT have been extensively reported using direct cholangiography, few series have reported imaging findings with MRCP [8, 11]. The results of our series are similar to the results of these previously reported studies. In the series of Laghi et al. [11], MRCP features of biliary complications in a group of 11 patients were reported. However, the diagnostic criteria of anastomotic and nonanastomotic strictures were not clearly defined. In our study, we classified a stricture as anastomotic if it was focal at the site of biliary anastomosis (which in our institution is usually an end-to-end hepaticohepaticostomy). A stricture was considered as nonanastomotic (ischemic cholangitis) if there was a long hilar stricture (not focal at the site of anastomosis) or if there were signs of intrahepatic biliary necrosis or biloma.
We also introduced a new concept, the donor-to-recipient common bile duct disproportion, which to the best of our knowledge has not been previously reported. This imaging feature involves a recipient's common bile duct diameter that is substantially different from the diameter of the donor's common bile duct in the absence of intrahepatic biliary dilatation. Usually in cases of common bile duct disproportion, the donor's common bile duct is thin and smooth, and the recipient's common bile duct is slightly enlarged. Occasionally the donor's common bile duct may be abnormally enlarged, whereas the receptor duct is thin and smooth. These cases may be misdiagnosed as an anastomotic stricture. In our series, one case of donor-to-recipient common bile duct disproportion was misdiagnosed as an anastomotic stricture on MRCP. In general, the absence of intrahepatic dilatation and the fact that the bile lumen is decreased smoothly without complete focal obliteration may suggest the diagnosis. However, in day-to-day practice, comparison with intraoperative cholangiographic findings is mandatory because it allows assessing stability of the imaging findings and therefore suggests the correct diagnosis.
Biliary stone formation or intraductal debris accumulation is another complication after liver transplantation and can be explained by alterations in bile composition. On the basis of findings on cholangiography and sonography, Barton et al. [20, 21] reported a prevalence of 13% of sludge in transplant recipients. The diagnostic accuracy of MRCP for the detection of stones is very high, with reported sensitivities ranging between 90% and 95% [22, 23]. In our series, the prevalence of isolated biliary stones or sludge was 14.2%; however, there were 25 patients (39%) with another biliary complication that also presented associated biliary sludge or stones. The accuracy of MRCP for diagnosis of stones in our series (88.9%) is not substantially different from the accuracy reported in previously published series [22, 23].
There are some limitations to this study. First, not all transplant recipients underwent MRCP, so the true prevalence of biliary complications was probably underestimated. However, these patients were closely monitored clinically and analytically, and therefore, it is unlikely that a patient with biliary complications was overlooked. In addition, the number of biliary complications in our series is relatively small. In conclusion, MRCP can be safely used as the only imaging method in the diagnosis of late biliary complications after liver transplantation. In terms of diagnosis, MRCP had a sensitivity of 95.3% (41/43), a positive predictive value of 97.6% (41/42), and a diagnostic accuracy of 95.2% (60/63). We found that 96.8% (61/63) patients were adequately imaged using only MRCP, and 3.2% (2/63) required additional ERCP or PTHC. Due to its wide availability and lack of side-effects, MRCP emerges as the only imaging method required to assess the biliary tree in transplant recipients. In our experience, direct cholangiography is not needed for diagnosis and should be reserved for therapeutic procedures.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Novellas, T. Caramella, M. Fournol, J. Gugenheim, and P. Chevallier MR Cholangiopancreatography Features of the Biliary Tree After Liver Transplantation Am. J. Roentgenol., July 1, 2008; 191(1): 221 - 227. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Hoeffel, L. Azizi, M. Lewin, V. Laurent, C. Aube, L. Arrive, and J.-M. Tubiana Normal and Pathologic Features of the Postoperative Biliary Tract at 3D MR Cholangiopancreatography and MR Imaging RadioGraphics, November 1, 2006; 26(6): 1603 - 1620. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
