University of Pisa
Department of Translational Research and New Technologies
in Medicine and Surgery
Residency Program in Diagnostic Radiology
(2012-2017)
Chairman: Prof. Davide CaramellaBiliary complications after liver transplantation: evaluation with MR
cholangiography and MR imaging at 3T device
Supervisor
Candidate
Prof. Davide Caramella, MD Dr. Federica Pacciardi, MD Academic Year 2015-2016ABSTRACT
Biliary complications after liver transplantation: evaluation with MR cholangiography and MR imaging at 3T device
Purpose: To assess the diagnostic value of MR cholangiography (MRC) and MR imaging at 3T device
when evaluating biliary adverse events after liver transplantation.
Material and Methods: A series of 374 MRI examinations in 232 liver transplant subjects with
suspected biliary complications (impaired liver function tests and/or biliary abnormalities on ultrasound) were performed at 3T device (GE-DISCOVERY MR750; GE Healthcare). After the acquisition of axial 3D dual-echo T1-weighted images and T2-weighted sequences (propeller and SS-FSE), MRC was performed through coronal thin-slab 3D-FRFSE and coronal oblique thick-slab SSFSE T2w sequences. DW-MRI of the liver was also performed using an axial spin-echo echo-planar sequence with multiple b values (150, 500, 1000, 1500 sec/mm2) in all diffusion directions. Contrast
enhanced-MRC was performed in 25/232 patients. All MR images were blindly evaluated by two experienced abdominal radiologists in conference to determine the presence of biliary complications, whose final diagnosis was based on direct cholangiography, surgery and integrating clinical follow-up with ultrasound and/or MR findings.
Results: In 113 patients no biliary abnormality was observed. The remaining 119 subjects were
affected by one or more of the following complications: non anastomotic strictures including typical ischemic-type biliary lesions (n=67), anastomotic strictures (n=34), ampullary dysfunction (n=4), anastomotic leakage (n=4), stones, sludge and casts (n=65), vanishing bile duct (n=1). The sensitivity, specificity, PPV, NPV and diagnostic accuracy of the reviewers for the detection of all types of biliary complications were 99%, 96%, 95%, 99% and 97%, respectively.
Conclusion: MR cholangiography and MR imaging at 3T device are reliable for detecting biliary
complications after liver transplantation.
Introduction
Despite refinements of surgical techniques and improvements in organ preservation, biliary complications (BCs) after liver transplantation (LT) are still common (ranging between 5.8% and 30%) and represent important causes of morbidity and graft dysfunction [1, 2] [3]. State-of-the-art MRI using 1.5 or 3T magnets is the preferred non-invasive modality to investigate BCs [4]. In a recent meta-analysis published by Xu et al. [5] these authors concluded that MR cholangiography (MRC) is a highly accurate diagnostic technique for diagnosis of BCs in liver transplant recipients, with a pooled sensitivity and specificity of 95% and 92%, respectively. Although large multicentric trials on proper target populations are lacking, a review of the literature suggests the high negative predictive value of MRC in excluding BCs in patients with low-to-moderate risk. On the other hand, positive MRC provides a detailed road map for planning interventional procedures or a surgical approach, thus further contributing to reduce morbidity[6, 7]. The recent innovation of 3T device, including technical advances such as ultra-fast imaging sequences and parallel imaging techniques, provides improved image quality over that at 1.5T [8]. Several studies have shown benefits for 3T MRC compared with 1.5T, in particular with higher SNR, CNR, and improved intrahepatic duct visualization (even in a non-dilated biliary system) [9, 10]. To the best of our knowledge, no study has specifically investigated 3T MRI in the evaluation of post-transplant BCs and so our study was aimed to assess the diagnostic value of MRC and MR imaging at 3T device when evaluating BCs in liver transplant recipients.
Material and methods Patients Between June 2011 and December 2016 we retrospectively selected from the radiological database of our Institution 396 exams in 244 orthotopic transplant patients (175 males and 69 females; age range: 21-77 years; mean age: 55 years) who were studied with MR imaging at 3T device. LTs were all full grafts from deceased donors excepted for a case of split cadaveric liver transplant. The indications for LT were represented by hepatocellular carcinoma (n=79), HCV-related hepatic cirrhosis (n= 32), HBV-related hepatic cirrhosis (n= 12), mixed HBV/HDV- related hepatic cirrhosis (n = 15), alcoholic cirrhosis (n = 44), fulminant hepatic failure (n = 9 ), cryptogenic hepatic cirrhosis (n = 7), Budd- Chiari syndrome (n=2), primary biliary cirrhosis (n=16), primary sclerosing cholangitis (n=17), Caroli’s disease (n = 1), cholangiocarcinoma (n= 2), polycystic hepatic disease (n=3), hemochromatosis (n =1), autoimmune hepatitis (n= 4).
The MR examinations were executed for a clinical suspicion of a biliary complication that was mainly based on altered liver function tests (elevated ALT, AST, gamma-glutamyl-transpeptidase, alkaline phosphatase, bilirubin) associated or not to abnormalities at abdominal ultrasound (biliary dilation; biliary wall thickening; lithiasis; liver parenchymal lesions).
In our series, biliary reconstruction was performed with choledocho-choledochostomy (CC) in 198 patients (185 with and 13 without a T-tube stent) and with bilio-enteric anastomosis (BEA) in 46 subjects. When a T-tube was placed through a choledochotomy in the recipient common bile duct, it was left in place for about three months. In our liver transplant center T-tube cholangiography is performed if necessary during this period and in all cases, before the removal of T-tube.
Twelve patients were excluded from our statistical analysis since a gold standard technique and/or clinical-radiological follow-up were not available; as a consequence our statistical evaluation was limited to 232 patients who performed 374 MR examinations.
Direct cholangiographic techniques, represented by endoscopic retrograde cholangiography (ERC) or percutaneous trans-hepatic cholangiography (PTC), were utilized in 87 patients, and were considered the “gold standard” methods for biliary tree imaging. In 12 subjects requiring surgery (8 BEA reconstructions and 4 re-transplantations) surgical findings were used to validate MRI results. In 20 patients with slight anastomotic and non-anastomotic strictures in which an invasive treatment was considered inadequate by our liver transplant center, the final diagnosis was assumed by integrating clinical/ laboratoristic findings with various imaging modalities including MRI and by subsequent follow-up (ultrasound every 2-3 months and MR exam every 6 months). In the remaining 113 recipients in whom no biliary abnormality was identified at MR examination, clinical follow-up was integrated with ultrasound and/or MR imaging for a period of at least twelve months. Institutional review board approval was obtained for this study and all the enrolled patients signed a written informed consent after the nature of the procedure had been fully explained.
MR imaging MR examinations were performed with a 3T scanner (GE-DISCOVERY MR750; GE Healthcare) utilizing an eight-channel phased-array body coil (maximum gradient strength over 50 mT/m, maximum slew rate 200 mT/m/s).
Immediately before starting MR imaging, scopolamine methyl-bromide (Buscopan® 20mg/ml, Boehringer Ingelheim) was intramuscularly administered in order to avoid peristaltic artefacts.
After the acquisition of T1- and T2w sequences with and without fat-suppression in the axial and coronal plane (FRFSE-Propeller T2w images, fast spin-echo (FSE) T2w sequence, spoiled gradient-echo (SPGR) T1w images) conventional MR cholangiography was performed through a respiratory-triggered, thin-slab, three-dimensional, heavily T2w fast spin-echo sequence and through breath-hold, thick-slab, single-shot T2w sequences.
Diffusion-weighted MR imaging of the liver was performed using an axial respiratory-triggered spin-echo echo-planar sequence with multiple b gradient factor values in all diffusion directions.
In 25 patients a three-dimensional fat-suppressed LAVA (Liver Acquisition with Volumetric Acceleration) T1w sequence was performed in the axial and coronal plane before and 20-120 minutes after intravenous administration of Gd-EOB-DTPA (Primovist®, Bayer Schering Pharma) with a power
injector (Empower, MR ACIST medical system, Bracco) via the antecubital vein at a dosage of 0,1 ml/Kg of body weight and a flow rate of 2 ml/sec, followed by injection of 20 ml of isotonic saline solution. MRI protocol is reported in Table 1.
Table1. MR imaging protocol
Sequence TR (ms) TE(ms) FA(D*) NEX Thickness/Spacing BH/RTr**
Pre-contrast imaging 2D Axial SPGR Dual Echo T1w 150 TE1=1,3;TE2=2,5 50 0.71 3.20/1.60 BH 2D Axial-Coronal SSFSE T2w FS Propeller T2w 1600 3500-6000 94 65-90 90 110 0.6 0.2 5/0 5/0 BH RTr MRCP 3D 2727 857 90 0.5 2/1 RTr MRCP thick slab 6000 602 90 0.8 10-50/0 BH 2D-Axial SE-EPI Diffusion T2w Multiple b-values (0.150,500,1000,1500 mm2/sec) 4500-7000 Minimum 90 1-6 5/0 RTr Post-contrast imaging 3D Axial-Coronal-LAVA Flex T1w 5.2 Minimum full 25-30 0.72 5/2.5 BH Statistical analysis
The distribution of the qualitative variables was expressed as the relative frequency of the various modalities under observation. The distribution of the quantitative variables was expressed as the mean, standard deviation, minimum, maximum, and number of observations.
MRI findings were correlated with direct cholangiography, surgery and/or clinical-radiological follow-up and defined as true positive when they correctly identified biliary complications confirmed by the final diagnosis reference standards; false positive when they were not confirmed by direct cholangiography, surgery and/or clinical-radiological follow-up; false negative when complications detected by direct cholangiography, surgery and/or clinical-radiological follow-up were not observed by MR examination; true negative when the absence of complications was confirmed by direct cholangiography, surgery and/or clinical-radiological follow-up.
To evaluate the diagnostic yield of MR imaging and MR cholangiography, we determined the sensitivity, specificity, diagnostic accuracy, positive predictive value (PPV), and negative predictive value (NPV) of the reviewers for the detection of all BCs. TR, repetion time; TE, echo time; FA flip angle; NEX number of experiments; BH breath holding; RTr Respiratory trigger;
Results
A regular or normal biliary tree anatomy was observed in 113 subjects, whereas in the remaining 119 patients (51%) BCs were identified; in this second group of patients one or more of the following complications were detected: anastomotic strictures (ASTs) (n = 34), non-anastomotic strictures (NASs) including typical “ischemic-type biliary lesions” (ITBL) (n = 67), sphincter of Oddi dysfunction (SOD) (n = 4), biliary stones, sludge, and casts (n = 64), biliary leakage (n = 4) and “vanishing bile duct syndrome” (n=1) (see Table 2).
Table 2. Biliary complications in our series of 232 liver transplanted patients. Non-anastomotic stricture including typical ITBL 67 Anastomotic stricture 34 Sphincter of Oddi dysfunction (SOD) 4 Biliary stones, sludge and casts 64 Biliary leakage 4 Vanishing bile duct syndrome 1 Patients with regular or normal biliary tree anatomy 113
In our series NASs including typical ITBL were the most frequent adverse events; they prevalently involved the hepatic confluence and extra hepatic biliary system of the graft and were confirmed by our gold standard methods in all 67 cases.
On the other hand, in 33 out of 34 cases of ASTs, reviewers correctly assessed the site of biliary obstruction and the dilated biliary system above the stenosis was completely visualized on MR images; there was a 100% correlation between MRC and standard of reference. In 13 of these patients no significant excretion of Gd-EOB-DTPA was observed in the physiological times after contrast agent administration. Direct cholangiography did not confirm ASTs diagnosed on MRC images in 2 patients (2 false positive cases); in these subjects the discrepancy in the calibre of the donor and recipient common bile ducts (ratio 2:1) was interpreted as a stricture at the anastomotic site, but it was due to the anatomical conformation of pre- and post-anastomotic biliary tree. Another BEA stricture graded as mild on PTC was misdiagnosed as an artefact at MRC (1 false negative case). Biliary stones, sludge, and casts were seen alone or associated with both ASTs and NASs and were correctly identified in the intra- and/or extra-hepatic biliary tract on MR images in 64 patients. Stones were typically recognized as low-signal-intensity areas surrounded by high-signal-intensity bile in the posterior portion of the ducts on axial images whereas sludge and cast showed high hyperintensity on T1 images. All four anastomotic leaks were diagnosed due to the fact that Gd-EOB-DTPA-enhanced MRC was able to directly visualize contrast material extravasation into the fluid collection, allowing to identify the site of bile leakage. At last, in the 4 patients with sphincter of Oddi dysfunction we observed on MR images a significant dilatation of both recipient and donor common bile duct in the presence of narrowing of the terminal
portion of recipient’s common bile duct. However, in two patients an anatomical discrepancy between donor’s and recipient’s common bile ducts was misdiagnosed as SOD (2 false positives cases); in these subjects contrast-enhanced MRC was not performed.
The only case of vanishing bile duct was correctly interpreted by MRI and confirmed by histophatology after liver re-transplantation.
Overall, on the basis of reviewers’ readings, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy of the MRC and MR imaging at 3T device in the detection of biliary complications after LT were 99%, 96%, 95%, 99% and 97%, respectively (see Table 3).
Table 3. Diagnostic value of MRC and MR imaging at 3T device in the detection of BC after LT Test parameters Reviewers True positive 118 True negative 109 False positive 4 False negative 1 Statistical measures Sensitivity (%) 99 Specificity (%) 96 Accuracy (%) 97 PPV (%) 95 NPV (%) 99 *The reference standards used for the final diagnosis are given in the text. Discussion
Nowadays, biliary reconstruction during LT is mainly performed with an end- to-end anastomosis between donor and recipient choledochal duct (CC) or with an end-to-side anastomosis between the donor hepatic duct and a recipient jejunal loop (BEA). The first one is technically simpler and preserves the sphincter of Oddi as a barrier against bacterial colonization of the biliary tract, while the BEA is usually reserved for cases of primary sclerosing cholangitis (PSC), surgical salvage after BCs or re-transplantation[11].
Despite a gradual decrease in incidence, BCs are still frequent, mainly represented by bile leakage, ASTs and NASs, and bile ducts obstructions. According to the literature, the main risk factors are technical complications, ischemia/reperfusion injury, ABO mismatch, hepatic artery complications (thrombosis and stenosis), donor age and cytomegalovirus infection[12].
Bile leaks involved 5% of liver transplant recipient and more than 70% occur within the 1st postoperative month. Leaks occur most often at the T-tube site and rarely at the site of anastomosis. Bile may leaks freely into the peritoneal cavity or may form a perihepatic collection. Treatment includes stent placement and drainage of collections [2].
usually related to tension at the surgical anastomosis site, ischemia, or size mismatch between the donor and recipient ducts. NASs manifest early after LT if associated to hepatic artery thrombosis (properly said NASs), or late if presenting with patent hepatic artery (ITBL)[13-15].
Sludge, casts and stone usually are associated to NASs or ASTs as the consequence of the necrosis of biliary epithelium or bile stasis [16].
Vanishing bile duct syndrome is characterized by progressive loss of small intrahepatic ducts and chronic rejection and ischemia are the most common causes [17].
Finally SOD is reported to be up to 7% in liver transplant recipients and it is attributed to the denervation of the sphincter during LT. After the removal of a T-tube (if used), the only means of directly visualizing the biliary system in the past was to use invasive procedures, such as ERC or PTC, which have been considered for a long time the standard of reference in imaging patients with CC and BEA, respectively. The advent of MRC has changed this situation. Nowadays ERC and PTC are usually reserved when interventional procedures are necessary while MRC is use for routine screening of BCs and/or for planning surgical or endoscopic procedures[4, 7, 11, 18-21]. MRC is a useful test to identify patients who needed interventional or surgical treatment of BCs after LT and several authors have reported very encouraging results as concerns as the MRC evaluation of BCs after LT [2, 4-6, 8, 10, 11, 17, 18, 22-28].
The main advantage of conventional T2w MRC is that the bile ducts are non-invasively depicted in their normal state, both below and above obstruction sites. 3D-imaging techniques provide better image quality compared to 2D-sequences, even though the combination of different MRCP sequences has proven to be valuable in the assessment of bile duct anatomy and pathology[6, 29]. However depiction of anatomy and lesion detection can be inadequate in a non-dilated biliary system; besides, free fluid and leak in the vicinity obscures the biliary anatomy due to overlapping. T1w contrast-enhanced MRC with intravenous administration of Gd-EOB-DTPA may provide both anatomical and functional information on the biliary tract demonstrating bile leakage or evaluating BEA and bile cast syndrome[6, 28]. Kinner et al.[26] concluded that Gd-EOB-DTPA T1w-MRC is a valuable technique for the assessment of biliary complications in patients after liver transplantation and the use of high-resolution navigator-based methods and higher flip angles, which optimize the SNR and CNR of biliary imaging using T1w-MRC, may improve our detection[6, 24, 30]. Only a few studies have compared 1.5 and 3T MRC in the evaluation of the biliary tract [29, 31] [32] and, to the best of our knowledge, no study has specifically investigated 3T MRI in the evaluation of BCs in liver transplant patients. The new generation of clinical MR systems operating at very high field strengths has, in general, the potential to significantly improve spatial resolution due to greater signal to noise ratio (SNR) and is faster. Studies of MRCP at 3T compared with 1.5T showed a trend toward increased CNR and therefore depiction of anatomy and lesion detection in a non-dilated biliary system is significantly better at 3T, improving the evaluation of intrahepatic ductal disease [10, 31].
strengths, magnetic field inhomogeneity are increased, susceptibility effects are greater and the specific absorption rate of radio- frequency power deposition may become a limiting factor [31] [10, 33].
In our study group about half of our patients do not present BCs since ultrasound is not always reliable to exclude BCs and so a second level imaging is necessary. Nevertheless, complications after liver transplant often present subtle and/or nonspecific signs and symptoms showing a large spectrum of overlapping clinical and laboratory manifestations. Once imaging excludes causes of BCs requiring surgical and/or interventional approach, further steps include the exclusion of primary graft dysfunction, rejection or drug toxicity. Not surprisingly, liver biopsy is mandatory for diagnosing and monitoring primary parenchymal complications[25]. On the other hand, in our series biliary complications were identified in 51% of liver recipients. Non-anastomotic strictures including typical “ischemic-type biliary lesions” (ITBL) were the most frequent complications and were identified in 100% of cases. We confirmed also at 3T device that MRC is the non-invasive technique of choice to evaluate the number, site and extent of strictures, which appears as tight, smoothly delineated focal zones of decreased signal intensity along bile ducts, with a typical, extensive involvement of the hepatic confluence and the donor tract of the common bile duct [25](Fig.1-2).
Besides, five out of 17 liver transplant patients for PSC showed disease relapse developing multiple biliary strictures alternating with dilation of bile ducts (Fig.3). 3T MRC accurately exhibited a typical “pruned tree” appearance of the biliary system with multiple stenoses alternating with normal or slightly dilated bile ducts. All these biliary findings and in particular the intrahepatic bile ducts variations could be more easily demonstrated at 3T device[8]. DWI was useful to detect inflammation of bile ducts in particular in patients with ITBL and with PSC relapse, showing a significant restriction of diffusivity when using high b values (1000-1500)[13]. In the experience of our transplant centre MRC is a very useful tool for the management of these patients since it often plans interventional procedures or surgical approach in multidisciplinary team meetings.
Fig.1 Non-anastomotic biliary stricture: Ischemic type biliary lesion. MIP reconstruction of 3D thin slab magnetic resonance cholangiogram (a) demonstrates a stenosis of the pre-anastomotic biliary tract with an irregular dilation of the intrahepatic biliary system; Axial T2- (b) and axial T1w (c) images well exhibit the presence of circumferential wall thickening (red arrows) at the level of common bile duct and hepatic bifurcation and hyperintense endoluminal casts (white arrow), respectively; on diffusion-weighted imaging the liver parenchyma appears inhomogeneous with multiples areas of persistent high signal intensity in highest b-value acquisitions (b=1000) (d).
a
b
c
d
Fig.2 Non-anastomotic biliary stricture in patient with CC anastomosis.
Single-shot thick slab MRC (a) and coronal SSFSE T2w images (b) show a dilated intra- and extrahepatic biliary system with a non-anastomotic stricture at the left hepatic duct. T1w image (c) well demonstrates an endoluminal cast above the stricture. Gd-EOB-DTPA enhanced 3D T1w MRC at 20 minutes (d) shows a regular bile excretion associated to a filling defect at the level of the stenosis.
Fig.3 Recurrence of primary sclerosing cholangitis in a transplant patient with hepatico-jejunostomy.
Single-shot thick slab MRC (a) and axial Propeller T2w (b) images show multifocal stenosis with intervening dilation affecting the intrahepatic biliary system; coronal SSFSE T2w image shows also an irregular appearance of the extra hepatic bile duct (c). On diffusion-weighted MR imaging with high b-values (b=1500) liver parenchyma appears markedly inhomogeneous with areas of persistent high signal intensity in all b-value acquisitions (d). PTC (e) confirmed the irregularities of the intrahepatic biliary tract and a biliary drainage is placed.
On MRCP the AST appears as a short, focal calibre reduction and/or biliary signal loss at the common bile duct, lying between the donor and recipient cystic duct stumps, with biliary dilation upstream. In our study there is an overestimation of the ASTs in two patients with CC (2 false positive cases) due to the anatomical conformation of pre- and post-anastomotic biliary tree. Since only significant stenoses are treated, it is important to know post transplant biliary anatomy (using intraoperative cholangiography and/or trans-Kher cholangiography when it is used) in order to avoid an overdiagnosis (Fig.4).
d
c
b
a
a
b
e
c
d
b
Fig.4 Anastomotic biliary stricture in a patient with a CC anastomosis. Coronal T2 image (a) shows dilation of pre- anastomotic biliary tract with a cast above the anastomotic site’s stricture. MIP reconstruction of 3D thin-slab fast spin-echo T2-weighted image (b) better demonstrates a circumscribed narrowing at the level of the surgical anastomosis associated with dilation of the pre-anastomotic biliary tract. Axial T1-weighted images confirms the presence of calculi in the biliary tract (c). PTC images show anastomotic stricture and stones which were removed by a bilioplastic treatment (d-f).
The degree and extent of BEA strictures are much more difficult to be assessed, since biliary dilatation is often present also when there is not a significant stenosis at the anastomotic site. The identification of pneumobilia is indicative of anastomosis patency; besides, contrast-enhanced MRC with Gd-EOB-DTPA can be used to determine whether there is impended contrast flow into the duodenum [24, 25]. In our series a BEA stricture graded as mild on PTC was misdiagnosed as an artefact at MR cholangiography (1 false negative case).
In 64 patients of our study group 3T MR imaging and MRC confirmed a high diagnostic accuracy in the detection of biliary stones, sludge, and casts in the intra- and/or extra-hepatic biliary tract. They are usually associated to NASs or ASTs as the consequence of the necrosis of biliary epithelium or bile stasis. On MRCP, sludge and stones appear as an intense filling defects surrounded by a thin rim of hyperintense bile, whereas biliary cast syndrome is hiperintense on T1w images [25]. Superior depiction of hepatic ducts at 3T MRI may improve the detection of small intraductal calculi and casts [34](Fig.5).
a
b
c
d
e
f
Fig.5 Anastomotic biliary stricture with lithiasis in a patient with hepatico-jejunostomy. Coronal/oblique MIP reconstructions of 3D thin-slab MRC image (a) and SSFSE image (b) show dilation of the biliary system above the hepatico-jejunal anastomosis with a concomitant coarse stone into the common hepatic duct at the level of the anastomotic site. Patient is treated by multiple bilioplastics through PTC (c-e). All 4 biliary leaks were correctly identified on MRI after the administration of Gd-EOB-DTPA contrast agent (in one case waiting until 120 minutes). On the basis of our experience contrast-enhanced MRC is able to demonstrate active biliary leakage by visualizing contrast media extravasation into the fluid collection. In this setting, we suggest to use contrast material when there is a strong suspicion of biliary leakage (fluid collection in the perianastomotic region) and to wait more than 20 minutes in order to reliably exclude the bile leak [10] (Fig.6).
a
b
e
d
c
Fig.6 Anastomotic bile leakage
Axial SSFSE fs T2w image (a) and conventional coronal thick-slab single-shot MRC (b) show a sub-hepatic fluid collection (red arrow) partially obscuring the biliary tract. Coronal (c) and MIP reconstructions (d) obtained from Gd-EOB-DTPA enhanced 3D T1w MRC at 20 and 60 minutes respectively demonstrate the presence of a leakage at the anastomotic site with contrast-enhanced bile filling of the sub-hepatic collection. ERC reveals a contained leakage in the perianastomotic region (e,f); a biliary endoprosthesis is therefore applied (g). We correctly evaluated all 4 cases of SOD that showed a significant dilatation of both recipient and donor bile duct on MRC associated to a protrusion of the enlarged ampullary region into the duodenal lumen. In these cases, Gd-EOB-DTPA-enhanced MRC was added to T2w MR cholangiography in order to obtain functional information on the degree of biliary obstruction and increase the diagnostic accuracy of MR imaging. Two patients with a discrepancy between donor’ and recipient’s common bile ducts were misdiagnosed as SOD but in these cases contrast material was not administered (2 false positive cases). In the only patient with a vanishing bile duct syndrome, MRC was able to accurately show paucity of small intrahepatic bile ducts suggesting this disease entity[17] (Fig.7).
a
d
c
b
e
f
g
Fig.7 (a-d). Vanishing bile duct syndrome. Axial Propeller T2w image (a) and coronal SSFSE T2w image (b) show the absence of intrahepatic biliary system; these findings are better appreciable on single shot thick slab MRC image (c) where only right-left hepatic ducts and extrahepatic pre- and post-anastomotic biliary tract are visible while the entire intrahepatic biliary system is missed; axial Gd-EOB-DTPA LAVA T1w image (d) show the absence of contrast excretion into the biliary system at 20 minutes.
On the basis of the reviewers’ readings, sensitivity and specificity of 3T MRC and MR imaging in detecting BC after LT were higher then that reported in the most recent metanalysis of the literature when using 1.5T scanner[5, 7]; besides, this diagnostic modality can predict BC in at least 95%, and excludes them in 99% of cases. Our data may have important implications in the management of liver transplant recipients since MRI allows guide for further treatment. Moreover, this study has several limitations: first, in 113 patients with a regular or normal biliary tract anatomy and in 20 patients with ASTs and/or NASs the final diagnosis was obtained not by ERC/PTC/surgery but by integrating clinical – laboratory and radiological findings; however MR imaging has been validated as a reliable tool to detect biliary complications and all these patients underwent serial clinical and imaging follow-up with ultrasound and MRC.
Two reviewers in conference performed imaging analysis, so the inter-observer agreement was not evaluated in order to establish the reproducibly of our results.
At least this is a retrospective study and so prospective studies are advocated to evaluate the diagnostic accuracy of 3T MR in this setting; however, 3T MR devices are not easily available, entail
b
d
c
high costs and require highly qualified medical personnel.
Conclusions
In conclusion, in liver transplant recipients MR cholangiography and MR imaging at 3T device are extremely reliable for detecting biliary complications and should be recommended before planning any therapeutic interventions.
References
1. Welling, T.H., et al., Biliary complications following liver transplantation in the model for end-stage liver disease era: effect of donor, recipient, and technical factors. Liver Transpl, 2008. 14(1): p. 73-80. 2. Caiado, A.H., et al., Complications of liver transplantation: multimodality imaging approach. Radiographics, 2007. 27(5): p. 1401-17. 3. Camacho, J.C., et al., Nonvascular post-liver transplantation complications: from US screening to cross-sectional and interventional imaging. Radiographics, 2015. 35(1): p. 87-104. 4. Girometti, R., et al., Post-operative imaging in liver transplantation: state-of-the-art and future perspectives. World J Gastroenterol, 2014. 20(20): p. 6180-200. 5. Xu, Y.B., et al., Diagnostic value of magnetic resonance cholangiopancreatography for biliary complications in orthotopic liver transplantation: a meta-analysis. Transplant Proc, 2013. 45(6): p. 2341-6. 6. Girometti, R., et al., Magnetic resonance cholangiography in the assessment and management of biliary complications after OLT. World J Radiol, 2014. 6(7): p. 424-36. 7. Jorgensen, J.E., et al., Is MRCP equivalent to ERCP for diagnosing biliary obstruction in orthotopic liver transplant recipients? A meta-analysis. Gastrointest Endosc, 2011. 73(5): p. 955-62. 8. Boraschi, P., et al., Biliary complications following orthotopic liver transplantation: May contrast-enhanced MR Cholangiography provide additional information? Eur J Radiol Open, 2016. 3: p. 108-16. 9. Bhatti, L., et al., Advanced magnetic resonance techniques: 3 T. Radiol Clin North Am, 2015. 53(3): p. 441-55. 10. Boraschi, P., M.C. Della Pina, and F. Donati, Graft complications following orthotopic liver transplantation: Role of non-invasive cross-sectional imaging techniques. Eur J Radiol, 2016. 85(7): p. 1271-83. 11. Boraschi, P., et al., Detection of biliary complications after orthotopic liver transplantation with MR cholangiography. Magn Reson Imaging, 2001. 19(8): p. 1097-105. 12. Memeo, R., et al., Management of biliary complications after liver transplantation. World J Hepatol, 2015. 7(29): p. 2890-5. 13. Wang, J., et al., Could diffusion-weighted imaging detect injured bile ducts of ischemic-type biliary lesions after orthotopic liver transplantation? AJR Am J Roentgenol, 2012. 199(4): p. 901-6. 14. Collettini, F., et al., Ischemic-type biliary lesions after ortothopic liver transplantation: diagnosis with magnetic resonance cholangiography. Transplant Proc, 2011. 43(7): p. 2660-3. 15. den Dulk, A.C., et al., Value of Magnetic Resonance Cholangiopancreatography in Assessment of Nonanastomotic Biliary Strictures After Liver Transplantation. Transplant Direct, 2015. 1(10): p. e42.16. Katabathina, V.S., et al., Adult bile duct strictures: role of MR imaging and MR cholangiopancreatography in characterization. Radiographics, 2014. 34(3): p. 565-86. 17. Boraschi, P. and F. Donati, Postoperative biliary adverse events following orthotopic liver transplantation: assessment with magnetic resonance cholangiography. World J Gastroenterol, 2014. 20(32): p. 11080-94. 18. Linhares, M.M., et al., Magnetic resonance cholangiography in the diagnosis of biliary complications after orthotopic liver transplantation. Transplant Proc, 2004. 36(4): p. 947-8. 19. Stadnik, A., et al., Biliary complications after liver transplantation: The role of MR imaging using different hydrographic sequences in patients with biliary-enteric and duct-to-duct biliary anastomosis. Ann Transplant, 2013. 18: p. 460-70. 20. Fulcher, A.S. and M.A. Turner, Orthotopic liver transplantation: evaluation with MR cholangiography. Radiology, 1999. 211(3): p. 715-22. 21. Faleschini, G., et al., Predictors of endoscopic treatment outcome in the management of biliary complications after orthotopic liver transplantation. Eur J Gastroenterol Hepatol, 2015. 27(2): p. 150-4. 22. Boraschi, P., et al., Complications after liver transplantation: evaluation with magnetic resonance imaging, magnetic resonance cholangiography, and 3-dimensional contrast-enhanced magnetic resonance angiography in a single session. Can Assoc Radiol J, 2008. 59(5): p. 259-63. 23. Boraschi, P., et al., MR cholangiography in orthotopic liver transplantation: sensitivity and specificity in detecting biliary complications. Clin Transplant, 2010. 24(4): p. E82-7. 24. Boraschi, P. and F. Donati, Biliary-enteric anastomoses: spectrum of findings on Gd-EOB-DTPA-enhanced MR cholangiography. Abdom Imaging, 2013. 38(6): p. 1351-9. 25. Girometti, R., et al., Imaging of liver transplantation. Eur J Radiol, 2017. 26. Kinner, S., et al., Added value of gadoxetic acid-enhanced T1-weighted magnetic resonance cholangiography for the diagnosis of post-transplant biliary complications. Eur Radiol, 2017. 27. Long, B. and A. Koyfman, The emergency medicine approach to transplant complications. Am J Emerg Med, 2016. 34(11): p. 2200-2208. 28. Kinner, S., et al., Biliary complications after liver transplantation: addition of T1-weighted images to MR cholangiopancreatography facilitates detection of cast in biliary cast syndrome. Radiology, 2012. 263(2): p. 429-36. 29. O'Regan, D.P., et al., A comparison of MR cholangiopancreatography at 1.5 and 3.0 Tesla. Br J Radiol, 2005. 78(934): p. 894-8. 30. Ogul, H., et al., The efficiency of Gd-EOB-DTPA-enhanced magnetic resonance cholangiography in living donor liver transplantation: a preliminary study. Clin Transplant, 2014. 28(3): p. 354-60. 31. Schindera, S.T. and E.M. Merkle, MR cholangiopancreatography: 1.5T versus 3T. Magn Reson Imaging Clin N Am, 2007. 15(3): p. 355-64, vi-vii. 32. Girometti, R., 3.0 Tesla magnetic resonance imaging: A new standard in liver imaging? World J Hepatol, 2015. 7(15): p. 1894-8. 33. Merkle, E.M., B.M. Dale, and E.K. Paulson, Abdominal MR imaging at 3T. Magn Reson Imaging Clin N Am, 2006. 14(1): p. 17-26. 34. Patel, H.T., et al., MR cholangiopancreatography at 3.0 T. Radiographics, 2009. 29(6): p. 1689-706.
