|Year : 2020 | Volume
| Issue : 7 | Page : 1634-1640
Hilar cholangiocarcinoma: Value of high-resolution enhanced magnetic resonance imaging for preoperative evaluation
Yinghui Xin1, Qingwei Liu1, Jie Zhang1, Jun Lu2, Xie Song2, Haiyu Zhan3, Xinyi Chen4, Zhixin Cao5, Yan Li6, Zhaoqin Huang1
1 Department of Radiology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University; Department of Radiology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, China
2 Department of Radiology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, China
3 Department of Radiology, The First Affiliated Hospital of Shandong First Medical University, Jinan, Shandong, China
4 Department of Nuclear Medicine, Jinling Hospital, School of Medicine, Nanjing University, Nanjing, China
5 Department of Pathology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, China
6 Department of Ultrasound, The Second Hospital of Shandong University, Jinan, Shandong, China
|Date of Submission||05-Feb-2020|
|Date of Decision||18-Aug-2020|
|Date of Acceptance||27-Oct-2020|
|Date of Web Publication||9-Feb-2021|
Department of Ultrasound, The Second Hospital of Shandong University, 247# Beiyuan Road, Jinan, Shandong Province
Department of Radiology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, 9677# Jingshi Road; Department of Radiology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, 250021
Source of Support: None, Conflict of Interest: None
Objectives: The objective is to assess the accuracy of high-resolution (HR) enhanced magnetic resonance imaging (MRI) images in the preoperative evaluation of biliary and vascular invasion in hilar cholangiocarcinomas.
Methods: This retrospective study included 36 patients with hilar cholangiocarcinoma who underwent enhanced HR-MRI with an effective section thickness of 1.2 mm at 3.0 T before surgery. Combined HR-MRI and magnetic resonance cholangiopancreatography (MRCP) images were compared with MRCP in evaluating the extent of biliary infiltration according to the Bismuth–Corlette classification. To determine the suitable criterion for HR-MRI in predicting vessel invasion, Labeling 180 and 90 of circumferential contact of the tumor with the vessel were used to predict the invasion. The correlation between imaging findings and surgical and histopathological records was statistically analyzed.
Results: The accuracy in detecting biliary neoplastic invasion was higher for combined HR-MRI images (97.2%) than MRCP images (86.1%). HR-MRI images increased the accuracy in delineation of the tumor biliary extent (P < 0.05). The accuracy of Labeling 90 (98.6% in portal venous system and 98.0% in hepatic arterial system) was higher than that of Labeling 180 (96.5% in portal venous system and 94.6% in hepatic arterial system). However, there was no significant statistic difference between them (P > 0.05). Interobserver agreement was high with respect to biliary tract, portal venous, and hepatic arterial system involvement.
Conclusions: Enhanced HR-MRI images showed excellent capability for assessing tumor extent and vascular invasion in hilar cholangiocarcinomas. More than 90° of circumferential contact of the tumor with the vessel on HR-MRI may be an appropriate criterion for predicting invasion.
Keywords: Hepatic artery, hepatic duct, hilar cholangiocarcinoma, magnetic resonance imaging, portal vein
|How to cite this article:|
Xin Y, Liu Q, Zhang J, Lu J, Song X, Zhan H, Chen X, Cao Z, Li Y, Huang Z. Hilar cholangiocarcinoma: Value of high-resolution enhanced magnetic resonance imaging for preoperative evaluation. J Can Res Ther 2020;16:1634-40
|How to cite this URL:|
Xin Y, Liu Q, Zhang J, Lu J, Song X, Zhan H, Chen X, Cao Z, Li Y, Huang Z. Hilar cholangiocarcinoma: Value of high-resolution enhanced magnetic resonance imaging for preoperative evaluation. J Can Res Ther [serial online] 2020 [cited 2021 Mar 8];16:1634-40. Available from: https://www.cancerjournal.net/text.asp?2020/16/7/1634/308760
| > Introduction|| |
Hilar cholangiocarcinoma is the most common type of cholangiocarcinoma, accounting for approximately 60% of all cases. Hilar cholangiocarcinoma is typically a small, slow-growing, locally invasive tumor with a very poor prognosis., Surgery is the main curative therapy. The median survival of patients who undergo complete resection is 30–46 months, and the 5-year survival rate ranges from 25% to 40%.,, However, about 30%–60% patients who undergo a laparotomy were found to have unresectable disease., A detailed preoperative delineation of the tumor extent and an assessment of vascular involvement are crucial for identifying patients who should undergo resection.
With advances in magnetic resonance imaging (MRI) technology, the combination of conventional contrast material-enhanced three-dimensional (3D) magnetic resonance (MR) and MR cholangiopancreatography (MRCP) has shown improved diagnostic accuracy for hilar cholangiocarcinoma.,,, However, conventional MR images cannot be reconstructed or provide detailed information about smaller bile ducts and blood vessels. Up to now, there is few reports assessing the value of high-resolution (HR) enhanced MR images for assessing the tumor extent and vascular involvement of hilar cholangiocarcinomas.
The purpose of this retrospective study was to assess the accuracy of preoperative HR enhanced MR in predicting biliary and vascular involvement in hilar cholangiocarcinomas.
| > Methods|| |
Seventy-one patients pathologically diagnosed as hilar cholangiocarcinoma who underwent MR in our department between December 2016 and December 2018 were eligible for inclusion. The inclusion criteria were as follows: (a) complete MR examinations with good imaging quality available for analysis; (b) surgery performed in our institution within 2 weeks after MR examination; (c) no biliary tract surgery or any anticancer treatments performed before MR examinations; and (d) no history of clonorchis sinensis infestation, intrahepatic stone disease, or primary sclerosing cholangitis. After selection, 36 patients (29 males and seven females; age range, 35–74 years; mean age, 61 years) were included in the present study.
Magnetic resonance imaging
All MRI scans were performed on a 3.0-T MR imaging unit (Ingenia; Philips). Two types of liver-specific MR contrast medium were used in this study: Gadobenate dimeglumine (Gd-BOPTA, MultiHance; Bracco Imaging) and gadoxetate disodium (Gd-EOB-DTPA, Primovist, Bayer Schering Pharma). Prior to contrast medium administration, all patients were imaged with unenhanced sequences [Table 1]. Dynamically enhanced thin-slice images were obtained during the hepatic arterial phase (15–25 s), portal venous phase (60–70 s), delayed phase (2 and 10 min), and hepatobiliary phase (HBP) (20 min for Gd-EOB-DTPA and 90 min for Gd-BOPTA) after the bolus injection of gadolinium using a fat-suppressed, T1-weighted, 3D gradient-echo pulse sequence (mDixon) with an effective section thickness of 1.2 mm. Coronal oblique, thin-slice images were acquired along the plane of the common hepatic duct and primary confluence during the delayed phase (3-4 min) and HBP enhancement. Gd-BOPTA was infused at a standard dose of 0.1 mmol/kg of body weight using an automated injection at a rate of 2 ml/s followed by a 20-ml saline flush. Gd-EOB-DTPA was infused at a standard dose of 0.025 mmol/kg of body weight using an automated injection at a rate of 2 ml/s followed by a 20-ml saline flush.
Image analysis was first performed by two radiologists (reader 1 and reader 2, with 11 and 20 years' experience in abdominal imaging diagnosis, respectively) on a picture archiving and communication system (PACS) workstation (Centricity PACS RA1000; General Electric). The enhanced thin-slice images were subjected to interactive real-time multiplanar reconstruction (MPR), maximum intensity projection (MIP), and minimum intensity projection (MinIP) to evaluate the involvement of the vascular and biliary system. Differences between the two observers were resolved by consensus.
Evaluation of the biliary system
We compared the MRCP with combined HR MR and MRCP images in evaluating the extent of biliary infiltration according to the Bismuth–Corlette classification [Figure 1]. The assessment of the tumor extent and involvement of the biliary system were evaluated: Common hepatic duct; primary biliary confluence; left hepatic duct; right hepatic duct; and left and right secondary confluence. On MR images, complete ductal occlusion, abrupt and/or irregular ductal narrowing, loss of continuity of the bile duct, irregularly shaped intraluminal filling defects, and segmental enhancing wall thickening of the bile duct were considered to indicate tumor invasion.,, The bile duct was marked as Grade 0 for no involvement and Grade 1 for involvement.
Evaluation of vascular involvement
We evaluated the extent of vascular infiltration according to the new staging system for perihilar cholangiocarcinoma. The portal vein system includes main portal vein, portal vein bifurcation, left portal vein, and right portal vein. The involvement of different parts of the hepatic artery system was also evaluated: Proper hepatic artery; hepatic artery bifurcation; left hepatic artery; right hepatic artery. Any anatomic variations regarding these vessels were recorded. In cases of variant vascular anatomy, the aberrant vessels were evaluated according to the same grading scale. The degree of involvement of the vessels was evaluated using a four-point grading scale as follows: Grade 0, no contact; Grade 1, tumor contact <90°; Grade 2, tumor contact 90°–180° of the circumference of the vessel, but with no narrowing of the lumen; and grade 3, >180° contact, luminal narrowing (eccentric or concentric) or vessel occlusion.
The new staging system labels arterial and venous involvement when the tumor encompasses >180° of the circumference of the vessel. Thus, Grades 0, 1, and 2 represent no vascular involvement, whereas Grade 3 represents vascular invasion. This criterion is named Labeling 180. To determine whether this criterion is suitable for HR MRI, another labeling was also used. More than 90° of circumferential contact of the tumor with the vessel predicts invasion (Labeling 90). Then, Grades 0 and 1 represent no vascular involvement, whereas Grades 2 and 3 represent vascular invasion.
All patients underwent surgery with a median interval of 5 days (range, 2–8 days) between MR imaging and the surgical procedure. Two board-certified surgeons, who have mainly served as hepatobiliary specialists for 13 and 19 years, respectively, performed the procedures. Exposure of the biliary confluence and assessment of the vascular involvement were achieved, and histologic assessment of the resection margin was performed. The final diagnosis of all tumors was based on histopathologic examination of the surgical specimen and resection margin assessment. The surgical and histopathological records represented the standard for reference.
All raw data were analyzed using a statistical software package (SPSS for Windows,version 20.0;IBM Corp.,Armonk,NY,USA). The performance of MRCP and combined HR MR and MRCP images in terms of the Bismuth classification were calculated. The performances were compared using Fisher's exact test. The sensitivity, specificity, and accuracy were calculated. For vascular invasion, a comparison was made between Labeling 180 and Labeling 90 using the McNemar test. Null hypotheses of no difference were rejected if P < 0.05. Cohen's k-statistic was calculated to quantify the interobserver agreement. A k >0.75 indicated excellent agreement, and 0.40≤ k ≤0.75 denoted moderate agreement.
| > Results|| |
Of the 36 patients, eleven (30.6%) underwent left hemi-hepatectomies, seven (19.4%) underwent extended left hemi-hepatectomies, four (11.1%) underwent right hemi-hepatectomies, nine (25%) underwent extended right hemi-hepatectomies, and the remaining five patients (13.9%) underwent bile duct resections and hepaticojejunostomies.
Evaluation of the biliary system
The assessment of the biliary system using MRCP was consistent with surgical/pathologic findings, in 31/36 patients (86.1%). Comparing the combination of HR MR and MRCP images with surgical/pathologic findings, the assessment of the biliary system was correctly evaluated by MRCP in 35/36 patients (97.2%) [Table 2]. In comparison with MRCP images, combined HR MR and MRCP images correctly changed the classification of tumor status according to the Bismuth classification system in 4/36 cases (11.1%), two cases from II to IIIA and IIIB, one case from IIIB to IV and one case from IV to IIIB [Table 2]. HR MR images increased the MR accuracy in delineation of the tumor biliary extent (P<0.05) [Figure 2].
|Figure 2: (a) High-resolution Gd-BOPTA enhanced portal venous phase image in 35-year-old man with Bismuth-Corlette Type IIIb hilar cholangiocarcinoma. The image shows clear thickening of the primary biliary confluence, left hepatic duct and left secondary confluence. (b) Minimum intensity projection reconstruction image of Gd-EOB-DTPA enhanced high-resolution delayed phase in 58-year-old man with Bismuth-Corlette type II hilar cholangiocarcinoma. Primary biliary confluence, left hepatic duct and right hepatic duct are involved|
Click here to view
|Table 2: The correlation between MRCP and the combination of HR-MR and MRCP with surgical/pathologic findings according to bismuth classification|
Click here to view
The overall accuracy for the determination of tumor extent along the biliary system was 97.7% (211 of 216 bile duct segments). High accuracy and sensitivity (100%) were shown in the assessments of the common hepatic duct, primary biliary confluence, and left hepatic duct [Table 3]. Regarding the assessment of the right hepatic ducts, the accuracy was 97.2% (35 of 36 cases), and one case was overestimated (specificity, 95.2%). The accuracy of the assessment of the left secondary confluence was 94.4% (34 of 36 cases), and two cases were underestimated (sensitivity, 84.6%). The accuracy of the assessment of the right secondary confluence was 94.4% (34 of 36 cases), with one case underestimated (sensitivity, 92.9%) and one case overestimated (specificity, 95.5%). The observers were in near-complete agreement regarding biliary system involvement [Table 3]. Interobserver agreement was statistically significant (All P < 0.05).
|Table 3: Magnetic resonance imaging prediction of biliary system involvement: Comparison with surgical and histopathological findings|
Click here to view
Evaluation of vascular involvement
Portal vein system involvement
High accuracy, sensitivity, and specificity (100%) were shown in the assessments of the main portal veins and portal vein bifurcations [Table 4]. Among the 144 portal venous segments, two segments (one left portal vein and one right portal vein) that were scored as having >180° of contact between the vessel and tumor on MRI were confirmed as non being involved. These two segments were overestimated due to the presence of an extensive peritumoral desmoplastic reaction within the connective tissue of the porta hepatis. Three segments (two left portal veins and one right portal vein) that were scored as having >90° and <180° of contact between the vessel and tumor were shown to be involved on surgical and histopathologic findings [Figure 3].
|Figure 3: Portal vein system involvement in different patients. (a) Enhanced high-resolution image shows the tumor (long arrow) encompasses 65.7° of the circumference of the left portal vein (short arrow). The left portal vein was not involved pathologically. (b) Another tumor (long arrow) encompasses 160.8° of the right portal vein (short arrow), with no narrowing of the lumen. The right portal vein proven to be involved pathologically (c and d) images in 59-year-old man. (c) The tumor (long arrow) is close to the left portal vein (short arrow). (d) Reconstruction image along the axial plane of the left portal vein shows no invasion of the vessel|
Click here to view
|Table 4: Magnetic resonance imaging prediction of portal vein system involvement: Comparison with surgical and histopathological findings|
Click here to view
The accuracy of Labeling 90 (98.6%, 142 of 144 portal venous segments) is higher than that of Labeling 180 (96.5%, 139 of 144 portal venous segments). However, there is no significant statistic difference between them (P = 0.25).
Hepatic artery system involvement
Arterial anatomy was conventional in 30 patients. Two right hepatic arteries (RHAs) arose directly from the superior mesenteric artery, and one left hepatic artery (LHA) arose directly from the left gastric artery. In three other patients, there were accessory RHAs arising separately from the left gastric and superior mesenteric arteries.
Three segments (one PHA, one LHA and one RHA) that were scored as having >180° of contact between the vessel and tumor on MRI were determined to not be involved based on surgical and histopathologic findings [Table 5]. They were overestimated as the tumor extended into the perivascular connective tissue. Five segments (one hepatic artery bifurcation, two LHAs, and two RHAs) were scored as having >90° and <180° of contact between the vessel and tumor on MRI and were confirmed as being involved [Figure 4]. The accuracy of Labeling 90 (98.0%, 144 of 147 hepatic arterial segments) is higher than that of Labeling 180 (94.6%, 139 of 147 hepatic arterial segments). There is no significant statistic difference between them (P = 0.062).
|Figure 4: Hepatic artery system involvement in two patients with hilar cholangiocarcinoma. (a) Abrupt cut-off of the hepatic artery (short arrow) was seen on the high-resolution hepatic artery phase image. The hepatic artery was confirmed surgically and histologically to be involved. (b) Enhanced high-resolution hepatic artery phase image shows the hepatic artery (short arrow) passing round the tumor (long arrow) naturally. The hepatic artery was not involved|
Click here to view
|Table 5: Magnetic resonance imaging prediction of hepatic artery system involvement: Comparison with surgical and histopathological findings|
Click here to view
The observers were in good agreement regarding vascular system involvement [Table 4] and [Table 5]. Interobserver agreement was statistically significant (All P < 0.05).
| > Discussion|| |
Hilar cholangiocarcinoma is the most common cholangiocarcinoma but it is challengeable to identify patients who are suitable for surgery. In determining unresectability, biliary, and vascular involvements are considered the most important factors. Thus, a detailed preoperative delineation of the tumor extent and assessment of the vascular involvement are essential. The small size and infiltrative nature of hilar cholangiocarcinomas make detection and determination of proximal extent difficult. Our study results demonstrated that HR MR images increased the MR accuracy in delineation of the tumor biliary extent (P < 0.05). The addition of HR MR to MRCP imaging improve the accuracy of the MR in assessing the classification of tumor status according to the Bismuth classification system.
In the current study, the overall accuracy for the determination of tumor extent was 97.7%, which is superior to the accuracy reported in previous studies.,, The superior performance could be attributed to several factors. First, HR images using 3D liver acquisition with volume acceleration sequences can be reconstructed along any plane, such as the biliary bifurcation plane. Second, the slice thickness of dynamic MRI in our study was 1.2 mm after zip 2 reconstruction (interpolation). Considering that many bile duct tumors are small, this extremely thin slice thickness may have contributed to the better results achieved in our study. During the pretest phase, we tried three different slice thicknesses (1.0 mm, 1.2 mm, and 1.5 mm). The signal-to-noise ratio of the 1.0-mm images was low. The display of smaller bile ducts and blood vessels was suboptimal on the 1.5-mm images. We chose 1.2 mm as the HR MRI thickness. During image analysis, we used MPR and MinIP to evaluate the tumor extent along the biliary tree.
The other crucial aspect in the staging of hilar cholangiocarcinomas is the assessment of vascular invasion. The new staging system labels arterial and venous involvement when there is evidence that the tumor encompasses >180° of the circumference of the vessel. The criteria was mostly based on data in a series of patients with pancreatic cancer and a small series of hilar cholangiocarcinoma with computed tomography (CT). In the current study, three cases of the portal venous system and five cases of the hepatic arterial system which were 90°–180° contact between the vessel and tumor were proved as involvement by surgical and histopathologic findings. The overall accuracies of assessing the involvement of the vascular system were 96.5% (139 of 144 portal venous segments) and 94.6% (139 of 147 hepatic arterial segments) for the portal venous and hepatic arterial systems, respectively in our study. Chryssou et al. reported that tumors with >90° of circumferential contact with the portal vein may be a useful criterion for predicting invasion of the portal vein or its segmental branches in nonstented patients with 1.5T MR. Zhou et al. reported that the modified Lu's standard (>1-quarter) performed best in the CT evaluation of portal venous invasion by hilar cholangiocarcinomas. For our study data, if we defined tumors in contact with >90° of the circumference of vessels as the criterion for predicting invasion, the overall accuracies would be 98.6% (142 of 144 portal venous segments) and 98.0% (144 of 147 hepatic arterial segments), although the difference is not statistically significant. For HR MRI of hilar cholangiocarcinoma, we propose labeling arterial and venous involvement when there is evidence that the tumor encompasses >90° of the circumference of the vessel.
Observers were in excellent agreement regarding vascular system involvement. In the current study, the sensitivity, specificity, and accuracy for distinguishing vascular invasion were superior to those in previous studies.,, Using MR angiography, the vessel lumen is visualized clearly. Consequently, MR angiography provides little information about primary tumors surrounding the vessel. The MPR and MIP images not only show vessel wall involvement but also show the vessel lumen with high reproducibility. A previous study conducted by Ni et al. demonstrated that postprocessing techniques such as MinIP and MPR can comprehensively reveal the relationship between tumors and blood vessels. Our study had two limitations. First, our study was limited by the small number of surgical patients. Our study population only included patients with confirmed hilar cholangiocarcinomas who had undergone surgical resection. Patients with advanced stage hilar cholangiocarcinomas who could not undergo surgery were not included. This exclusion criterion could have resulted in selection bias. Second, 14/71 (19.7%) initially selected patients were excluded from this study due to insufficient MR image quality due to motion artifacts, which implied that the diagnostic accuracy was only valid in cases with proper image quality.
| > Conclusions|| |
HR, enhanced images with 3D reconstruction show excellent capability in the preoperative assessment of the tumor extent and vascular involvement in patients with hilar cholangiocarcinomas and may contribute to identify real patients suitable for surgery; using tumor contact >90° of the vessel on preoperative HR-enhanced MR demonstrates high accuracy in evaluation of vascular involvement in hilar cholangiocarcinomas.
The correlation between MRCP and the combination of HR MR and MRCP with surgical/pathologic findings as the standard of reference for determining the biliary infiltration according to Bismuth classification.
Financial support and sponsorship
This study has received funding by Primary Research and Development Plan of Shandong Province (No. 2016GSF201095 and No.ZR2017PH 001).
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Khan SA, Thomas HC, Davidson BR, Taylor-Robinson SD. Cholangiocarcinoma. Lancet 2005;366:1303-14.
Hoffmann K, Luible S, Goeppert B, Weiss KH, Hinz U, Büchler MW, et al
. Impact of portal vein resection on oncologic long-term outcome in patients with hilar cholangiocarcinoma. Surgery 2015;158:252-1260.
Groot Koerkamp B, Jarnagin WR. Surgery for perihilar cholangiocarcinoma. Br J Surg 2018;105:771-2.
Sun Z, Feng Y, Liu H, Shen F, Xiao J, Kang X, et al
. Analysis of treatment methods and prognostic factors in 354 cases of hilar cholangiocarcinoma: A cohort study. J Cancer Res Ther 2020;16:230-7.
Ito F, Cho CS, Rikkers LF, Weber SM. Hilar cholangiocarcinoma: Current management. Ann Surg 2009;250:210-8.
Groot Koerkamp B, Wiggers JK, Gonen M, Doussot A, Allen PJ, Besselink MG, et al
. Survival after resection of perihilar cholangiocarcinoma-development and external validation of a prognostic nomogram. Ann Oncol 2016;27:753.
Coelen RJ, Ruys AT, Besselink MG, Busch OR, van Gulik TM. Diagnostic accuracy of staging laparoscopy for detecting metastasized or locally advanced perihilar cholangiocarcinoma: A systematic review and meta-analysis. Surg Endosc 2016;30:4163-73.
Coelen RJ, Ruys AT, Wiggers JK, Nio CY, Verheij J, Gouma DJ, et al
. Development of a risk score to predict detection of metastasized or locally advanced perihilar cholangiocarcinoma at staging laparoscopy. Ann Surg Oncol 2016;23:904-10.
Sun HY, Lee JM, Park HS, Yoon JH, Baek JH, Han JK, et al
. Gadoxetic acid-enhanced MRI with MR cholangiography for the preoperative evaluation of bile duct cancer. J Magn Reson Imaging 2013;38:138-47.
Arrivé L, Hodoul M, Arbache A, Slavikova-Boucher L, Menu Y, El Mouhadi S. Magnetic resonance cholangiography: Current and future perspectives. Clin Res Hepatol Gastroenterol 2015;39:659-64.
Wu XP, Ni JM, Zhang ZY, Lu FQ, Li B, Jin HH, et al
. Preoperative evaluation of malignant perihilar biliary obstruction: Negative-Contrast CT Cholangiopancreatography and CT angiography versus MRCP and MR Angiography. AJR Am J Roentgenol 2015;205:780-8.
Park HS, Lee JM, Choi JY, Lee MW, Kim HJ, Han JK, et al
. Preoperative evaluation of bile duct cancer: MRI combined with MR cholangiopancreatography versus MDCT with direct cholangiography. AJR Am J Roentgenol 2008;190:396-405.
Chryssou E, Guthrie JA, Ward J, Robinson PJ. Hilar cholangiocarcinoma: MR correlation with surgical and histological findings. Clin Radiol 2010;65:781-8.
Kim MJ, Mitchell DG, Ito K, Outwater EK. Biliary dilatation: Differentiation of benign from malignant causes-value of adding conventional MR imaging to MR cholangiopancreatography. Radiology 2000;214:173-81.
Kim HJ, Lee JM, Kim SH, Han JK, Lee JY, Choi JY, et al
. Evaluation of the longitudinal tumor extent of bile duct cancer: Value of adding gadolinium-enhanced dynamic imaging to unenhanced images and magnetic resonance cholangiography. J Comput Assist Tomogr 2007;31:469-74.
Deoliveira ML, Schulick RD, Nimura Y, Rosen C, Gores G, Neuhaus P, et al
. New staging system and a registry for perihilar cholangiocarcinoma. Hepatology 2011;53:1363-71.
Ryoo I, Lee JM, Chung YE, Park HS, Kim SH, Han JK, et al
. Gadobutrol-enhanced, three-dimensional, dynamic MR imaging with MR cholangiography for the preoperative evaluation of bile duct cancer. Invest Radiol 2010;45:217-24.
Zhang H, Zhu J, Ke F, Weng M, Wu X, Li M, et al
. Radiological imaging for assessing the respectability of hilar cholangiocarcinoma: A systematic review and meta-analysis. Biomed Res Int 2015;2015:497942.doi:10.1155/2015/497942.
Yin LL, Song B, Xu J, Li YC. Hilar cholangiocarcinoma: Preoperative evaluation with a three dimensional volumetric interpolated breath-hold examination magnetic resonance imaging sequence. Chin Med J (Engl) 2007;120:636-42.
Nakanishi Y, Tsuchikawa T, Okamura K, Nakamura T, Tamoto E, Murakami S, et al
. Prognostic impact of the site of portal vein invasion in patients with surgically resected perihilar cholangiocarcinoma. Surgery 2016;159:1511-9.
Zhou Q, Guan Y, Mao L, Zhu Y, Chen J, Shi J, et al
. Modification and establishment of CT criteria in preoperative assessment of portal venous invasion by hilar cholangiocarcinoma. HPB (Oxford) 2018;20:1163-71.
Lee MG, Park KB, Shin YM, Yoon HK, Sung KB, Kim MH, et al
. Preoperative evaluation of hilar cholangiocarcinoma with contrast-enhanced three-dimensional fast imaging with steady-state precession magnetic resonance angiography: Comparison with intraarterial digital subtraction angiography. World J Surg 2003;27:278-83.
Ni Q, Wang H, Zhang Y, Qian L, Chi J, Liang X, et al
. MDCT assessment of resectability in hilar cholangiocarcinoma. Abdom Radiol (NY) 2017;42:851-60.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]