Indian Journal of Radiology Indian Journal of Radiology  

   Login   | Users online: 1265

Home Bookmark this page Print this page Email this page Small font sizeDefault font size Increase font size     


MAJOR PAPERS Table of Contents   
Year : 2002  |  Volume : 12  |  Issue : 1  |  Page : 37-42
Utility of MR cholangiography in planning transhepatic biliary interventions in malignant hilar obstructions

Department of Radiology, Amrita Institute of Medical Sciences, Amrita lane, Elamakkara, Cochin-682026, Kerala, India

Click here for correspondence address and email


Aim: To evaluate the accuracy of Magnetic resonance cholangiography (MRC) in defining the extent of biliary ductal involvement in malignant hilar obstruction and its usefulness in planning percutaneous palliative drainage. Materials and Methods: Fifteen patients with malignant hilar obstructions were examined with MRC. All patients underwent Percutaneous Transhepatic Cholangiography (PTC) and biliary drainage. The site of obstruction and a hypothetical drainage plan based on the MRC findings were compared to the PTC and the actual drainage performed. Results: MRC accurately predicted the site of obstruction in 11 out of 14 patients (78 percent). One patient had an inadequate MRC. The hypothetical drainage plan based on MRC matched the actual drainage in 12 out of 14 cases (85 percent). Conclusion: Preprocedural MRC is a valuable tool for defining the site of obstruction. It provides a 'road map' for performing percutaneous biliary drainage.

Keywords: Hilar malignancy, MR Cholangiography, Percutaneous Transhepatic Drainage

How to cite this article:
Pillai A K, Sreekumar K P, Prabhu N K, Moorthy S. Utility of MR cholangiography in planning transhepatic biliary interventions in malignant hilar obstructions. Indian J Radiol Imaging 2002;12:37-42

How to cite this URL:
Pillai A K, Sreekumar K P, Prabhu N K, Moorthy S. Utility of MR cholangiography in planning transhepatic biliary interventions in malignant hilar obstructions. Indian J Radiol Imaging [serial online] 2002 [cited 2020 Jul 11];12:37-42. Available from:

   Introduction Top

Malignant hilar biliary obstructions are caused by tumors involving the major confluence of the hepatic ducts and upper third of the common bile duct. A spectrum of disease entities, namely, hilar cholangiocarcinoma, gallbladder carcinoma, metastasis, icteric hepatocellular carcinoma and lymphoma can produce hilar obstruction.

Majority of these tumors are inoperable at the time of diagnosis [1]. Palliation of jaundice relieves the patient of itching and improves the quality of life. Surgical, endoscopic and percutaneous palliative measures are available. Surgical palliation carries a significant mortality. Endoscopic palliation is useful in lower CBD obstruction. Percutaneous palliation is the treatment of choice in hilar biliary obstruction.

Magnetic resonance cholangiography (MRC) is an evolving non-invasive technique for biliary tract imaging. Ductal dilatation, strictures, stones and anatomical variation are well depicted by this imaging technique [2]. The ability to display the biliary tree by combining projectional and cross sectional views, makes this modality well suited for planning the optimal therapeutic approach for patients with biliary obstruction.

   Materials and Methods Top


Fifteen consecutive patients, 8 females and 7 males, who underwent biliary intervention between August 2000 and December 2001 were enrolled in our study. The oldest patient was 73 years and the youngest 45 years (mean 59 years) and MRC was performed in all. These patients had evidence of high grade biliary obstruction as depicted by IHBR dilatation on ultrasound/CT. No patient had previous history of biliary surgery. Specific tumor types were hilar cholangiocarcinoma (n=10), invasive gallbladder carcinoma (n=3) and periportal metastasis (n=2). Among the latter group, one patient had histologically proved breast carcinoma and the other had carcinoma of the stomach.

MR imaging

All patients fasted for 4 hours before the MRC in order to reduce artifacts from bowel peristalsis. The MR Cholangiograms were acquired using commercially available software in a clinical MR scanner (Magnetom Symphony 1.5T scanner, Siemens Medical Systems, Erlangen, Germany) with a body coil. All sequences were performed during breath holding. After localizer, heavily T2 weighted 3D thick slab single shot fast spin echo (SSFSE) and coronal, axial and oblique coronal thin multislice half Fourier turbo spin echo sequence (HASTE) were obtained.

Imaging parameters used for SSFSE sequence were as follows: TR 2800ms, TE 1000ms, slab thickness 80mm, FOV 300cms, Flip angle 150, matrix 240*265, acquisition time-4 seconds

Imaging parameters for multislice HASTE were as follows: TR 1300msec, TE 90ms, Slice thickness 4mm, FOV 300cms, flip angle 150, matrix 240*256, acquisition time 24 seconds.

Image interpretation

MR images were interpreted at an independent workstation (Sienet; Siemens Medical Systems; Erlangen). Complete sets of source images and projection rendering were evaluated individually or in a cine mode.

Biliary strictures were typed as per Bismuth Corlette classification. According to this classification system, type 1 obstruction occurs distal to the confluence of the right and left hepatic ducts (primary confluence), type 2 involves the primary confluence but not the secondary confluence [Figure - 2], type 3 involves the primary confluence and either the right (3a) [Figure - 3] or left (3b) secondary confluence, and type 4 involves the secondary confluence of both the right and left hepatic ducts [3]. In addition, the presence of lobar atrophy and anatomical variations were also evaluated.

Biliary drainage

The following drainage protocol is followed in our institution. Generally, For type 1 and type 2 lesions with no lobar atrophy, a single drain through the right anterior system is considered appropriate. Type 3 and 4 lesions require two drains i.e. right anterior and right posterior or right anterior and left systems. Lobes/segments which are atrophic are excluded from the proposed drainage plan. Drainage procedure however, is tailored according to the patient condition and disease process. For example, in cases of invasive GB malignancy a more conservative drainage approach is followed due to the short life expectancy. Based on the MRC findings, a hypothetical drainage plan was proposed. This was compared to the Percutaneous Transhepatic Cholangiography (PTC) and the actual drainage performed.

   Results Top

MRC was unsuccessful in one patient who could not cooperate. She was dyspnoeic consequent to septicemia induced by cholangitis. PTC was used to classify the stricture in all patients. Two patients had type 1 (13%), 4 patients had type 2 (26%), 8 patients had type 3a (53%) and 1 patient had type 4 (7%) lesion. None of the patients had a type 3b stricture. Lobar atrophy [Figure - 2] was noted in 5 patients (33%). In one patient (7%) there was an anatomic variation whereby the right posterior sectoral duct drained into the left duct [Figure - 4].

Patients with type 1 lesion underwent unilateral biliary drainage through the right anterior sectoral system. MRC typed the lesions accurately and the proposed drainage plan matched with that of PTC in these patients.

Of the four patients who had a type 2 stricture, MRC over-graded the tumor as type 3a in 2 cases [Table 2]. All four patients underwent a unilateral drainage through the right anterior sectoral duct. In these two patients the MRC based drainage plan did not match with the final drainage performed.

MRC correctly staged the disease in all 7 patients with type 3a stricture [Figure - 2]. The anticipated drainage plan matched with that of PTC in all the cases.

One patient had a type 4 lesion that was under-graded by MR. However the drainage plan did not change in this case as our protocol for type 3a and type 4 lesions, entails bisectoral drainage. Overall, MR accurately typed the site of obstruction in 11/14 cases (78%). The hypothesized drainage plan on MRI correlated with the actual drainage done in 12/14 patients (85%).

   Discussion Top

Only 20-30% malignant hilar tumors are resectable at the time of diagnosis. Although resection offers the only chance of cure, the five-year survival rates after potential curative resection for hilar cholangio carcinoma has been reported to be from 0-22% [4]. Most patients with metastic disease affecting the porta hepatis and invasive gallbladder carcinomas causing hilar biliary obstruction are not candidates for curative resection.

Management options include some form of biliary decompression and/or supportive care. Indications for biliary decompression in inoperable patients are intractable pruritus, cholangitis, and the need for access for intraluminal radiotherapy and to allow recovery of hepatic parenchymal function in patients receiving chemotherapeutic agents.

Biliary decompression can be obtained either by percutaneous transhepatic puncture or by endoscopic stent placement. Hilar tumors are more difficult to traverse with the endoscopic technique [1],[5],[6]. Moreover, the failure rates and incidence of subsequent cholangitis are high. Thus, endoscopic biliary drainage is not preferred in patients with irresectable hilar tumors.

The goal of imaging is to predict the site of obstruction and to look for lobar atrophy and intrahepatic metastasis. Frequently, hilar tumors isolate all 3 major hilar ducts (left hepatic, right anterior sectoral hepatic and right posterior sectoral hepatic), and 2 or more stents must be placed for adequate drainage [Figure - 2]. Though the necessity for drainage of all obstructed sectors has been emphasized by some authors [6], current opinion seems to favour a conservative approach. It is well known that drainage of 30% of the hepatic parenchyma will palliate the clinical and metabolic sequel of biliary obstruction [7]. Percutaneous drainage through an atrophic lobe usually does not relieve jaundice and should be avoided. The presence of multiple intrahepatic metastases or ascites may also add to the technical difficulty of the procedure.

Knowledge of the site of obstruction with respect to the primary and secondary biliary confluence is, thus vital for planning percutaneous access to the biliary tree. Ultrasound and helical computed tomography (CT) have a limited role in providing this information [8,9]. PTC and ERCP are considered the gold standards for assessing biliary anatomy. PTC is generally considered superior to ERCP for evaluation of biliary ducts in hilar obstruction [4]. However, incomplete opacification of the intrahepatic ducts is not an uncommon problem. Multiple projections with tilting and rolling of the patient are often necessary to improve the diagnostic accuracy of PTC. In patients with non-communicating intrahepatic segments of the biliary tree, multiple punctures of the individual ducts maybe necessary. Complications of PTC include pain, intra-abdominal hemorrhage, hemobilia and pneumothorax. Cholangitis is a dreaded complication occurring in 3.4-4.8% of patients resulting from failure of drainage of contrast medium injected into ducts with high-grade stenosis [9].

MRC is a powerful non-invasive tool for evaluation of malignant biliary strictures at the hepatic hilum. Unlike PTC, it does not require contrast administration and

incomplete opacification is not a problem. All the dilated bile ducts are depicted clearly whether isolated or not [Figure - 3]. Details regarding the site of obstruction, length of the stricture, anatomical variations of drainage and lobar atrophy are accurately brought out by this imaging modality. Lopera et al reported 96% accuracy in grading the stricture by MRC. PTC was used as the yardstick in this study [12]. Zidi et al compared MRC to ERC. MRC correctly graded the stricture in 14 out of 17 cases (78%) [11]. In our series MRC predicted the Bismuth grade accurately in 11 of the 14 cases (78%).

In two cases of invasive GB malignancies, MRC overestimated the grade of obstruction. The degree of ductal narrowing tends to be overestimated at MRC compared with direct cholangiography. This results from the lower spatial resolution of MRC. Moreover, MRC depicts the biliary tree in a physiologic state, whereas direct cholangiography depicts the biliary tree under pressure from contrast material injection [13].

In a case of cholangiocarcinoma the site of obstruction was typed 3a but on PTC the lesion was type 4. A short stricture of segment 3 adjacent to the left confluence was not picked up on the MR. This stricture was overlooked on the MRC possibly due to partial-volume averaging resulting from the use of a 4-mm slice thickness with the multislice half-Fourier method. Underestimation of the lesion by MRC has been reported by other authors, Zidi et al reported underestimation of the type of stricture in 22% of cases in his study [11]. Ta-Sen Yeh et al compared MRC with histopathologic correlation for the site of obstruction. In his series underestimation of stricture grade was noted in 20% of cases [14].

The hypothesized drainage plan matched with the actual drainage in 12 of the 14 patients (85%) in our study. Other investigators have also reported similar result [12]. Having prior knowledge of anatomical variants has important implications in percutaneous access and drainage planning. In one of our patients, the right posterior duct drained directly into the left hepatic duct [Figure - 4]. The left ductal approach alone resulted in drainage of large volumes of liver.

MR technology is improving at a dramatic rate. Compared with the body coil (used in our study), phased array coils provide a two to four fold improvement in the signal-to-noise ratio [15]. New sequences like simultaneous acquisition of spatial harmonics (SMASH) and sensitivity encoding (SENSE) which uses the intrinsic sensitivity variation of the phased-array coils to speed the process of phase encoding reduces imaging time by two fold. Recently, the first hepatobiliary agent was introduced for clinical MR imaging of the liver. Mangafodipir trisodium contains paramagnetic Manganese, which shortens T1[16]. The primary route of excretion is through the biliary system resulting in intense enhancement of the biliary tree on T1 weighted sequences. The combination of phased array coils, faster magnetic field gradients, contrast agents and newer sequences may eventually provide spatial resolution that will improve the performance of this modality. The results of this study indicate that MRC can correctly type the bile duct strictures in most patients with hilar malignancies. Based on the MRC, the appropriate ductal access and the sectors to be drained can be determined. Puncture into isolated segmental ducts can be avoided thereby reducing the risk of cholangitis. Though not evaluated in our study, we feel that the prior MRC 'road map' reduces the procedure time and thus, the radiation exposure to the patient and radiologist.

In conclusion, our results indicate that MRC is a valuable tool in evaluation of malignant hilar obstructions and helps in planning the optimal therapeutic approach for percutaneous drainage.

   Acknowledgement Top

We would like to acknowledge the support of our colleagues in the department of medical and surgical Gastroenterology and Mr. George KV, senior MRI technician.

   References Top

1.Blumgart LH. Cholangiocarcinoma. In: Blumgart LH, ed. Surgery of the liver and biliary tract. Edinburgh: Churchill Livingstone, 1988; 1: 721-753.  Back to cited text no. 1    
2.Adamek HE, Albert J, Weitz M, Breer H, Schilling D, Riemann JF, A Prospective evaluation of Magnetic Resonance Cholangiopancreatography in patients with suspected bile duct obstructions. Gut 1998; 43: 680-683.  Back to cited text no. 2    
3.Bismuth H, Corlette MB. Intrahepatic cholangioenteric anastomosis in carcinoma of the hilus of the liver. Surg GYnecol obstet 1975; 140: 170-178.  Back to cited text no. 3  [PUBMED]  
4.Garg PK, Tandon RK. Preoperative assessment of cholangiocarcinoma. Meeting the challenge. J gastroenterol hepatol 1999; 14: 615-617.  Back to cited text no. 4    
5.England RE, Martin DF, Endoscopic and Percutaneous Intervention in Malignant Obstructive Jaundice. Cardiovasc Intervent Radiol 1996; 19: 381-387.  Back to cited text no. 5    
6.Coons HG, Self-expanding stainless steel biliary stents. Radiology 1989; 170: 979-983.  Back to cited text no. 6    
7.Polydorou AA, Chisholm EM, Romanos AA. A Comparison of right veses left hepatic duct endoprosthesis insertion in malignant hilar obstruction. Endoscopy 1989; 21: 266-271.  Back to cited text no. 7    
8.Han JK, Choi BI, Kim TK, Kim SW, Han MC, Hilar Cholangiocarcinoma: Thin section spiral CT findings with cholangiographic correlation. Radiographics 1997; 17: 1475-1485.  Back to cited text no. 8    
9.Bloom C, Langer B, Wilson SR, Role of US in the detection, characterization and staging of cholangiocarcinoma. Radiographics 1999; 19: 1199-1218.  Back to cited text no. 9    
10.Ishizaki Y, Wakayama T, Okada Y. Magnetic resonance Cholangiography for evaluation of obstructive Jaundice. Am J gastroenterol 1993; 88: 2072-2076.  Back to cited text no. 10    
11.Zidi SH, Prat F, O Le Guen, Y Rondeau, G Pelletier, Performance Characteristics of Magnetic Resonance Cholangiography in the staging of Malignant Hilar Strictures. Gut 2000; 46: 103-106.  Back to cited text no. 11    
12.Lopera JE, Soto JA, Munera F, Malignant hilar and perihilar biliary obstruction: Use of MR Cholangiography to define the extend of biliary ductal involvement and plan percutaneous interventions. Radiology 2001; 220: 90-96.  Back to cited text no. 12    
13.Watanabe Y, Dohke M, Ishimori T, Diagnostic Pitfalls of MR Cholangiopancreatography in the Evaluation of the Biliary Tract and gallbladder Radiographics 1999; 19: 415-429.  Back to cited text no. 13    
14.Ta-Sen, Jan Y, Tseng J, malignant perihilar biliary obstruction: Magnetic Resonance Cholangiopancreatography. Am J gastroenterol 2000; 95: 432-440.  Back to cited text no. 14    
15.Keogan MT, Edelman RR, Technological advances in abdominal MR imaging. Radiology 2001: 220: 310-320.  Back to cited text no. 15    
16.Marchal G, Ni Y, Zhang X, Yu J, Lodemann KP, Baert AL. Mn-DPDP enhanced MRI in experimental bile duct obstruction. J Comput Assist Tomogr 1993; 17: 290-296.  Back to cited text no. 16  [PUBMED]  

Correspondence Address:
A K Pillai
Department of Radiology, Amrita Institute of Medical Sciences, Amrita lane, Elamakkara, Cochin-682026, Kerala
Login to access the Email id

Source of Support: None, Conflict of Interest: None

Rights and PermissionsRights and Permissions


[Figure - 1], [Figure - 2], [Figure - 3], [Figure - 4]


    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Email Alert *
    Add to My List *
* Registration required (free)  

    Materials and Me...
    Article Figures

 Article Access Statistics
    PDF Downloaded0    
    Comments [Add]    

Recommend this journal