Original Research
Gastrointestinal Imaging
February 20, 2014

Comparative MRI Analysis of Morphologic Patterns of Bile Duct Disease in IgG4-Related Systemic Disease Versus Primary Sclerosing Cholangitis

Abstract

OBJECTIVE. The purpose of this study was to perform a retrospective MRI-based comparative analysis of the morphologic patterns of bile duct disease in IgG4-related systemic disease (ISD, also called autoimmune pancreatitis) compared with primary sclerosing cholangitis (PSC) and the autoimmune liver diseases autoimmune hepatitis and primary biliary cirrhosis.
MATERIALS AND METHODS. This study included 162 consecutively registered patients (47 with ISD, 73 with PSC, and 42 with autoimmune liver diseases). Two abdominal radiologists retrospectively reviewed MR images in consensus. Imaging findings on the bile ducts, liver, pancreas, and other organs were analyzed to establish disease patterns.
RESULTS. ISD was associated with contiguous thickening of intrahepatic and extrahepatic bile ducts (p < 0.001), pancreatic parenchymal abnormalities (p < 0.001), renal abnormalities (p < 0.001), and gallbladder wall thickening (p < 0.03). The severity of common bile duct wall thickness was significantly different in ISD (p < 0.001). The mean single wall thickness in the ISD group was 3.00 (SD, 1.47) mm, in the PSC group was 1.89 (SD, 0.73) mm, and in the autoimmune liver disease group was 1.80 (SD, 0.67) mm. PSC was associated with liver parenchymal abnormalities (p < 0.001). We did not find statistical significance between the three groups in location (p = 0.220) or length (p = 0.703) of extrahepatic bile duct strictures, enhancement of bile duct stricture (p = 0.033), upper abdominal lymphadenopathy, or retroperitoneal fibrosis. Although presence of intrahepatic bile duct stricture was statistically significant when all three groups were compared, it was not useful for differentiating ISD from PSC.
CONCLUSION. The presence of continuous as opposed to skip disease in the bile ducts, gallbladder involvement, and single-wall common bile duct thickness greater than 2.5 mm supports a diagnosis of ISD over PSC. ISD and PSC could not be differentiated on the basis of location and length of common bile duct stricture.
IgG4-associated systemic disease (ISD) is an uncommon but clinically important multisystem disorder characterized by an elevated serum level of IgG4. ISD affects multiple organ systems and has been given several names according to the site of involvement: autoimmune pancreatitis [1], biliary tree IgG4-associated cholangitis [27], and kidney [8], retroperitoneal, and salivary gland disease [9]. Conclusive diagnosis can be difficult and is usually arrived at after multiple investigations, delaying treatment. ISD often presents as a focal pancreatic mass, mimicking pancreatic cancer. A misdiagnosis can lead to unnecessary cancer surgery, because ISD is treated with steroids [1012]. Approximately 2–3% of patients undergoing pancreatic resection for suspected cancer have ISD [13, 14].
Biliary tree abnormalities found in other pancreaticobiliary diseases can create diagnostic challenges to confirming ISD, most commonly primary sclerosing cholangitis (PSC). Bile duct involvement may also be observed in other autoimmune liver diseases; even though it is not the predominant feature in these diseases. Several authors have tried to assess the pattern of biliary disease in ISD, but the studies have usually been ERCP based [2, 15] or an evaluation of a mixture of modalities for imaging the pancreas [3]. MRCP and pancreaticobiliary MRI are well established and preferred for noninvasive imaging evaluation of the biliary tree [16, 17]. In this study we used MRI to systematically analyze the pattern of involvement of the biliary tree in ISD and describe any features that may be specific to ISD. We compared the findings with those in PSC and autoimmune liver diseases.

Materials and Methods

Patients

Institutional ethics board approval was obtained, and the requirement for informed consent was waived. From a clinical database, we retrospectively identified the cases of consecutively registered patients with the diagnosis of ISD (or autoimmune pancreatitis) (Figs. 1A1D and 2) who had undergone MRI evaluation of the abdomen. To attain statistical significance, a random consecutive selection of patients from our liver clinic database who had undergone pancreaticobiliary MRI with a clinical diagnosis of PSC, small-duct PSC, primary biliary cirrhosis, or autoimmune hepatitis were included. Patients with probable overlap syndromes were excluded. The American Association for the Study of Liver Diseases diagnostic criteria [18] were used for PSC and the Mayo Clinic HISORt (histology, imaging findings, serology, other organ involvement, and response to steroid treatment) criteria were used for diagnosis of ISD [19]. Thus, three cohorts comprising patients with ISD, PSC, and other autoimmune liver diseases (small-duct PSC, primary biliary cirrhosis, autoimmune hepatitis) were assessed. The third group was used to improve the statistical power and decrease bias rather than to evaluate the pattern of bile duct disease in these disorders.
Fig. 1A —55-year-old woman with elevated serum IgG4 level.
A, Axial T2-weighted (A), arterial phase (B), and delayed phase (C) contrast-enhanced MR images show enlarged pancreas (asterisk) with delayed contrast enhancement. C shows marked thickening and enhancement of common bile duct (arrow).
Fig. 1B —55-year-old woman with elevated serum IgG4 level.
B, Axial T2-weighted (A), arterial phase (B), and delayed phase (C) contrast-enhanced MR images show enlarged pancreas (asterisk) with delayed contrast enhancement. C shows marked thickening and enhancement of common bile duct (arrow).
Fig. 1C —55-year-old woman with elevated serum IgG4 level.
C, Axial T2-weighted (A), arterial phase (B), and delayed phase (C) contrast-enhanced MR images show enlarged pancreas (asterisk) with delayed contrast enhancement. C shows marked thickening and enhancement of common bile duct (arrow).
Fig. 1D —55-year-old woman with elevated serum IgG4 level.
D, Coronal delayed phase contrast-enhanced MR image shows involvement of entire common bile duct (straight arrows) in continuous manner with no intervening normal areas. Curved arrow indicates thickened gallbladder.
Fig. 1E —55-year-old woman with elevated serum IgG4 level.
E, MR image shows measurement of CBD wall thickness. Larger measurement relates to outer wall diameter and lesser measurement to inner diameter.
Fig. 2A —65-year-old man with obstructive jaundice and elevated serum IgG4 level.
A, Coronal T2-weighted slab MR image shows intrahepatic and extrahepatic bile duct dilatation with stricture in hilum and proximal common bile duct (CBD).
Fig. 2B —65-year-old man with obstructive jaundice and elevated serum IgG4 level.
B, Arterial phase contrast-enhanced MR image shows enlarged poorly enhancing pancreas (asterisk) with thickening and enhancement of proximal CBD (arrow).
Fig. 2C —65-year-old man with obstructive jaundice and elevated serum IgG4 level.
C, Delayed phase contrast-enhanced MR image shows thickening and enhancement of distal CBD (arrow).
Fig. 2D —65-year-old man with obstructive jaundice and elevated serum IgG4 level.
D, Coronal delayed phase contrast-enhanced MR image shows thickening and enhancement of entire CBD (arrows) in keeping with continuous disease.

Imaging

The MRI evaluation was identical in all groups. All but two patients (who underwent unenhanced MRCP) had undergone standard pancreaticobiliary MRI that included MRCP and gadolinium-enhanced imaging. Pancreaticobiliary MRI at our institution was performed with a mix of 1.5-T and 3-T systems. Torso phased-array coils were used in all cases. Gadobutrol (Gadovist, Bayer Healthcare) and gadopentetate dimeglumine (Magnevist, Bayer Healthcare) were the contrast agents. At our institution pancreaticobiliary MRI is performed with a 1.5-T system with an eight-channel body coil. The sequences are as follows: coronal single-shot fast spin-echo (SSFSE) (TE, 70), axial SSFSE (TE, 180), axial dual fast spoiled gradient-recalled echo (GRE) (FSPGR), axial 3D dynamic contrast-enhanced liver acquisition with volume acceleration, radial slabs, coronal 3D fast recovery fast spin-echo with respiratory triggering, and coronal 2D FSPGR delayed contrast-enhanced.
Acquisition performed with 1.5- and 3-T systems (Magneto Avanto Syngo, Siemens Healthcare) with eight-channel body coils consisted of the following sequences: coronal HASTE (TE, 180), axial in- and out-of-phase T1-weighted GRE, axial T2-weighted fat-suppressed motion correction with radial blades (BLADE) (TE, 90), coronal T2-weighted turbo spin-echo fat-suppressed 3D volume with respiratory triggering, T2-weighted HASTE radial slabs, axial unenhanced and dynamic contrast-enhanced volumetric interpolated breath-hold examination (VIBE), and coronal delayed VIBE.

Imaging Review

All studies were randomized. Two abdominal radiologists (10 and 1.5 years of experience) blinded to diagnoses performed consensus readings. Demographic and clinical chart data were analyzed separately. Most patients underwent several MRI examinations during their courses of treatment. For the purposes of this study, the initial MRI examination at the first presentation was evaluated for all patients. The earliest MRI examination included in the study was from February 2001, and the last dated MRI was performed in August 2009. There was a substantial time gap between the primary MRI reading and the consensus reading for this study, eliminating any recall bias. The parameters evaluated for the study are detailed in Table 1. The readers’ provisional diagnosis and degree of confidence (high, medium, low) were documented.
TABLE 1: Summary of Study Parameters Recorded by the Readers
Bile DuctsPancreasLiverOther
Presence of CBD stentBase T1 signal intensityLobar redistributionRenal lesions
CBD stricturesEnlargement [43], focal massCirrhosisPeripheral cortical nodules
 Site (proximal, distal, hilar)AtrophyPortal vein thrombusWedge-shaped cortical lesions
 NumberNormal contrast enhancementSplenomegalyDiffuse patchy involvement
 Length [42] < 2 mm (short), 3–10 mm (long), > 10 mm (segmental)Delayed contrast enhancement Retroperitoneal fibrosis
 Location of dominant stricturePresence of peripancreatic rind Upper abdominal lymphadenopathy
 Enhancement of dominant stricture   
Intrahepatic bile duct involvement   
 Order of ducts involved (1–4)   
 Beading   
 Pruning   
Continuous or skip involvement   
 Presence of contiguous wall thickening instead of discrete strictures or skip lesions   
Wall thickening at wall stricture   
 Diameter of lumen   
 Diameter of external walls   
Gallbladder   
 Previous cholecystectomy   
 Gallstones   
 Wall thickening   
 Abnormal wall enhancement   

Note—CBD = common bile duct.

Statistical Analysis

Before the initiation of the study a sample size calculation was performed. The association between the bile duct features and the three clinical groups was examined with the Kruskal-Wallis (common bile duct [CBD] wall thickness) and chi-square (other features) tests. A paired Student t test was performed on the ISD and PSC groups, excluding the third group. A value of p < 0.05 was considered to indicate a significant association.

Results

A total of 162 patients were included in the study. The patients were divided into three clinical groups: 47 patients had ISD (37 men, 10 women; average age, 55.9 years), 73 patients had PSC (34 men, 39 women; average age, 42.8 years), and 42 patients had autoimmune hepatitis, primary biliary cirrhosis, or small-duct PSC (11 men, 31 women; average age, 45.4 years). The results are summarized in Table 2.
TABLE 2: Summary of Findings
FindingIgG4-Related Systemic Disease (Autoimmune Pancreatitis)Primary Sclerosing CholangitisOtherp
Presence of CBD stricture28/47 (59.5)42/73 (57.5)6/42 (14.2) 
Location of CBD stricture   0.220
 Distal1060 
 Proximal231 
 Hilum10224 
 Diffuse6111 
Length of bile duct stricture   0.703
 Band (1–2 mm)110 
 Segmental (> 3 mm)152 
 Long (> 10 mm)26364 
Intrahepatic bile duct strictures30/47 (63.8)68/73 (93.2)19/42 (45.2)< 0.001
Enhancement of bile duct stricture20/28 (71.4)25/42 (59.5)2/6 (33.3)0.033
CBD wall thickness (mm)    
 Mean3.001.891.80< 0.001
 SD1.470.730.67 
 Median2.752.01.5 
Gallbladder thickening19/37 (51.3)17/61 (27.8)0/27< 0.030
Continuous disease in bile ducts19/28 (67.8)4/42 (9.5)1/6 (16.7)< 0.001
Hepatic abnormalities9/47 (19.1)42/73 (57.5)28/42 (66.7)< 0.001
Pancreatic abnormalities41/47 (87.2)2/73 (2.7)3/42 (7.1)< 0.001
Renal abnormalities14/47 (29.7)0/730/42< 0.001
Retroperitoneal fibrosis1/47 (2.1)0/731/42 (2.4)0.300
Upper abdominal lymphadenopathy7/47 (14.9)16/73 (21.9)8/42 (19.0)0.358

Note—Except for CBD wall thickness, values are number of patients with percentage in parentheses. CBD = common bile duct.

Dominant Bile Duct Stricture

A dominant bile duct stricture is defined as a CBD lumen narrower than 1.5 mm and a common hepatic duct lumen narrower than 1 mm [20]. Patients with visible dominant strictures at MRI were included in this analysis. Twenty-eight of 47 (59.5%) ISD patients had a CBD stricture; 42 of 73 (57.5%) patients with PSC had extrahepatic bile duct strictures; and 6 of 42 (14.3%) patients in the group with other autoimmune hepatic diseases had extrahepatic biliary disease. The distribution is shown in Table 2. The difference between the three groups with respect to location of CBD strictures was not statistically significant (p = 0.220).

Intrahepatic Bile Duct Disease

Patients with intrahepatic bile duct disease were included in this part of the analysis. Thirty of 47 (63.8%) patients in the ISD group had intrahepatic bile duct disease, as did 68 of 73 (93.2%) patients in the PSC group and 19 of 42 (45.2%) patients in the other autoimmune disease group. A comparison of all three groups showed an association between intrahepatic bile duct disease and ISD and PSC with a high rate of occurrence in both the ISD and the PSC groups (p < 0.001). The findings did not allow statistically significant differentiation between ISD and PSC.

Extent and Enhancement of Biliary Strictures

Patients with a CBD stricture were included in this analysis. Length of stricture was classified as in ERCP: band (1–2 mm), segmental (> 3 mm), and long (> 10 mm). All three groups had predominance of long strictures, but the difference was not statistically significant (p = 0.703). Among the patients in the ISD group with CBD stricture, 20 of 28 (71.4%) had stricture enhancement. In the PSC group 25 of 42 (59.5%) had enhancement of CBD strictures. In the third group two of six patients had enhancing CBD strictures. Overall there was poor correlation between the presence of enhancement in a CBD stricture in the three groups (p = 0.033).

Common Bile Duct Wall Thickness

The 3-minute delayed contrast-enhanced coronal or axial images were used for measurement of CBD wall thickness. This included only patients with CBD strictures. The inner diameter was subtracted from the outer diameter, and the resulting value was divided by 2 (Fig. 1E). The mean CBD wall thickness in patients with ISD was 3.00 (SD, 1.47) mm, in PSC patients was 1.89 (SD, 0.73) mm, and in the third group was 1.80 (SD, 0.67) mm. There was a significant difference in wall diameters between the ISD group and the other two groups (p < 0.001). The box plot of CBD wall thickness in the three groups is shown in Figure 3.
Fig. 3 —Box plot shows common bile duct wall thickness in three study groups. Symbols indicate outliers; horizontal lines in boxes, median wall thickness; whiskers, 1.5 x interquartile range. AIH = autoimmune hepatitis, ISD = IgG4-related systemic disease, PSC = primary sclerosing cholangitis.

Contiguous Biliary Disease

The analysis of contiguous biliary disease was based on the presence of contiguous bile duct disease in the intrahepatic and extrahepatic biliary tree in patients with multiple strictures. In the ISD group 19 of 32 (59.4%) patients had continuous disease, whereas in the PSC group, 4 of 65 (6.2%) patients had continuous disease, and 61 of 64 (95.3%) patients had skip disease. In the other group 1 of 17 (5.9%) patients had continuous disease. The presence of continuous disease in the biliary tree is a strong predictor of ISD (p < 0.001).

Gallbladder Involvement

Patients with cholelithiasis and those who had undergone surgical removal of the gallbladder were excluded from this analysis. Gallbladder wall thickening and enhancement were taken as abnormal features. Nineteen of 37 (51.4%) patients in the ISD had an abnormal gallbladder, as did 17 of 61 (27.9%) patients with PSC. No patient in the third group had an abnormal gallbladder. The presence of an abnormal gallbladder favored a diagnosis of ISD in our group of patients (p < 0.03).

Liver

Liver parenchymal abnormalities such as lobar redistribution (combination changes of atrophy and hypertrophy that occur with chronic liver disease), fibrosis, and cirrhosis were seen in 9 of 47 (19.1%) patients in the ISD group, but in the PSC group 42 of 73 (57.5%) patients had liver parenchymal abnormalities. In the group with other hepatic autoimmune disease 28 of 42 (66.7%) patients had liver parenchymal abnormalities. Thus the presence of liver parenchymal abnormalities such as cirrhosis and fibrosis in the presence of biliary disease supports a diagnosis of PSC over ISD (p < 0.001).

Pancreatic Abnormalities

Focal or diffuse pancreatic T1 signal-intensity abnormalities, atrophy, or an abnormal enhancement pattern were seen in 41 of 47 (87.2%) patients in the ISD group. In the PSC group only 2 of 73 (2.7%) had an abnormal pancreas. In the third group 7% of patients had an abnormal pancreas. The presence of abnormal pancreas correlates with a diagnosis of ISD (p < 0.001).

Renal and Retroperitoneal Abnormalities

Peripheral cortical abnormalities such as cortical nodules and wedge-shaped lesions along with diffuse patchy renal parenchymal involvement are well described in ISD. In our study, 14 of 47 (29.8%) patients in the ISD group had focal peripheral renal cortex abnormalities. No similar abnormalities were identified in the other two groups. The presence of renal abnormalities supported the diagnosis of ISD (p < 0.001). Only one patient with ISD had retroperitoneal abnormality; hence statistical association could not be evaluated.

Abdominal Lymphadenopathy

We did not find statistical significance to the presence of upper abdominal lymph nodes in differentiating the three groups of patients (p = 0.358). Although it was not our intent to assess the diagnostic accuracy of imaging for ISD, PSC, and other autoimmune diseases, the overall accuracy of the readers in giving the correct diagnosis was 85% for ISD, 88% for PSC, and 60% for other diseases (p = 0.001).

Discussion

ISD, also called autoimmune pancreatitis, most commonly affects the pancreas (Fig. 1). In a Japanese survey the prevalence of ISD was found to be 0.71 per 100,000 population [21]. IgG4-associated cholangitis (Fig. 2) has been well described and is used as one of the criteria for diagnosis of ISD. It is also an important imaging finding for differentiating pancreatic cancer from ISD causing pancreatitis (autoimmune pancreatitis). Involvement of the biliary tree in ISD without a pancreatic lesion [2224] sometimes can mimic hilar cholangiocarcinoma [25], thus diagnosing ISD in these patients can be challenging. It is also well recognized that some patients with ISD do not have an elevated IgG4 level and that an elevation in IgG4 level is not specific for ISD [26] and can be seen in other conditions, such as atopic dermatitis, pemphigus vulgaris, pancreatic cancer, and cholangiocarcinoma without PSC [27].
Patients with PSC have elevated serum IgG4 levels in 6.3–9% of cases [28, 29]. Zhang et al. [29] found that the liver explants of patients with PSC have periductal IgG4-positive plasma cell infiltrates in 23% of cases. Twenty-two percent of the study population had an increased IgG4 level in stored serum obtained before transplant. Because the clinical features of these patients were different from those of patients with ISD, the authors concluded that IgG4-positive PSC is a subtype of PSC. Nasr et al. [30] found elevated serum IgG4 levels in 28% of patients who underwent liver transplants for sclerosing cholangitis. Alswat et al. [31] found that pancreatic abnormalities are seen in PSC patients with and without elevated serum IgG4 levels. Thus an isolated finding of elevated serum IgG4 level or pancreatic abnormality in a patient with biliary disease may not be helpful for differentiating PSC from ISD.
Several diagnostic criteria have been proposed for aiding the diagnosis of ISD. These include the HISORt [19], Japanese, Korean [32], and Asian consensus [33] criteria. In a study of extrapancreatic lesions associated with ISD, Fujinaga et al. [9] found bile duct involvement in 77.8% of their study subjects. Diagnostic pathways have been devised to help differentiate ISD from pancreatic cancer. Involvement of the proximal CBD (because distal CBD involvement could be due to a pressure effect from pancreatic swelling) [34] or intrahepatic bile ducts [35, 36] is one of the imaging findings used in these diagnostic criteria.
Histologic descriptions of bile duct involvement in IgG4-associated cholangitis show that it is a separate entity from PSC. These studies also show that the CBD wall thickening in ISD was caused by fibrosis and lymphoplasmacytic infiltration with an intact bile epithelium and is different from PSC [5, 7]. Studies of surgical resection specimens and intraductal ultrasound studies have shown wall thickening even in parts of the CBD that do not exhibit strictures at ERCP [37]. This finding was also confirmed at ultrasound evaluation in a study by Nishino et al. [4] and is different from the bile duct thickening in PSC, which is confined to the site of stricture. However, histologic examination of the bile ducts to differentiate PSC from ISD is challenging and not practical as a routine discriminatory test.
Involvement of the gallbladder in ISD has been described. Abraham et al. [38] found in a comparison of surgical resection specimens from patients with PSC and ISD that there was no statistical significance in the number of patients with abnormal gallbladders in each group but that the histologic findings in the gallbladders of ISD and PSC patients differed. The ISD group had an increased incidence of transmural inflammation. Wang et al. [39] found that IgG4 immunohistochemical analysis of the gallbladder can be used to diagnose ISD. Kamisawa et al. [34] found statistically significant gallbladder wall thickening in their ISD study group. At histologic examination they found transmural thickening of the involved gallbladder and CBD.
The bile duct findings in the PSC group in our study show presence of intrahepatic bile duct disease with skip lesions in almost all patients; 57.5% of the PSC group patients had CBD strictures. Beading and a pruned tree appearance of the biliary tree were also noted. The findings in the PSC group are in keeping with those of previous MRCP-based studies [16, 40] (Fig. 4).
Fig. 4A —43-year-old woman with primary sclerosing cholangitis.
A, Thick-slab (A) and coronal T2-weighted fat-saturated (B) MR images show multiple intrahepatic bile duct strictures (thin arrows) and extrahepatic bile duct strictures (thick arrows, B).
Fig. 4B —43-year-old woman with primary sclerosing cholangitis.
B, Thick-slab (A) and coronal T2-weighted fat-saturated (B) MR images show multiple intrahepatic bile duct strictures (thin arrows) and extrahepatic bile duct strictures (thick arrows, B).
Fig. 4C —43-year-old woman with primary sclerosing cholangitis.
C, Axial contrast-enhanced images show common bile duct (thin arrow) and gallbladder (thick arrow, D) thickening. Absence of continuous common bile duct wall thickening is evident and is finding of IgG4-related systemic disease.
Fig. 4D —43-year-old woman with primary sclerosing cholangitis.
D, Axial contrast-enhanced images show common bile duct (thin arrow) and gallbladder (thick arrow, D) thickening. Absence of continuous common bile duct wall thickening is evident and is finding of IgG4-related systemic disease.
Our findings show that in the ISD group, the presence of CBD wall thickening, continuous involvement of the CBD, and an abnormal gallbladder have a statistically significant association with ISD and can be used to differentiate it from PSC. The presence of intrahepatic duct disease and stricture enhancement had some association, but it was not statistically significant.
The average CBD single-wall thickness in the ISD group was 3.00 (SD, 1.47; median, 2.75) mm, whereas in the PSC group it was 1.89 (SD, 0.73; median, 2.00) mm, so the presence of single-wall CBD thickness greater than 2.5 mm is suggestive of ISD. The percentage of ISD patients from our data with wall thickness greater than 2.5 mm was 50%, whereas in the PSC group it was 9%. Wall thickening and enhancement were best observed on 3-minute-delayed contrast-enhanced images.
When there was more than one site of abnormality in the biliary tree, the presence of abnormal wall thickening in the intervening part was interpreted as continuous disease. If the intervening part was completely normal, it was called skip disease. The presence of continuous disease is a good predictor of ISD.
The presence of thickened and enhancing gallbladder in the absence of stone disease also shows good association with ISD. The presence of liver parenchymal abnormalities in patients with biliary disease has a statistically significant negative association with ISD. Thus presence of liver parenchymal abnormalities should steer the diagnosis away from ISD toward PSC or another autoimmune liver disease, such as primary biliary cirrhosis or autoimmune hepatitis. It is possible that there may be liver parenchymal changes in ISD in the chronic phase, but we evaluated the first available image and may have underestimated liver changes.
The presence of renal and retroperitoneal abnormalities in ISD has been well described in the literature. Our study also shows the association between renal abnormalities and ISD, but the number of patients with retroperitoneal abnormalities was too small for useful conclusions.
Although ISD was associated with intrahepatic bile duct disease when the three study groups were compared, the association with PSC was also significant, and differentiation between ISD and PSC would be difficult on the basis of involvement of the intrahepatic bile ducts alone.
We found no statistical association between the location of the dominant strictures in the three groups. Nakazawa et al. [15] described the ERCP findings in ISD and PSC in a series of 55 patients. They found distal CBD stricture more frequently in patients with ISD. They also found an association of segmental and long strictures with ISD and band strictures with PSC. Several previous studies predominantly based on ERCP had similar conclusions. The biliary findings in those studies are summarized in Table 3.
TABLE 3: Collated Data on Bile Duct and Gallbladder Involvement
StudyYearConditionnDistal CBDProximal CBDDiffuse CBDIntrahepatic BDIntrahepatic and ExtrahepaticGallbladder Abnormality
Nakazawa et al. [15]2004AIP2623 (88)8 (30) 9 (35)  
  PSC2912 (41)5 (17)    
Nakazawa et al. [37]2005AIP2018 (90)11 (55) 7 (35)10 (50) 
  PSC27   2 (7)25 (93) 
Nishino et al. [2]2005AIP1613 (81)5 (31) 2 (13) 10 (63)
Nishino et al. [4]2007AIP2423 (96)1 (4)   15 (63)
  PSC243 (13)4 (17)   4 (17)
Ghazale et al. [3]2008AIP5337 (70)18 (34) 19 (36)  
Kamisawa et al. [34]2008AIP1715 (88)  2 (12)15 (88) 
  Pancreatic cancer7035 (50)  035 (50) 
Chari et al. [35]2009AIP45    31 
  Pancreatic cancer100      
Naitoh et al. [44]2011AIP1916 (84)10 (53) 7 (37)  
  PSC22   3 (14)19 (86) 
Tokala et al.Current studyISD4710 (21)12 (26)6 (13)30 (64) 19/37 (51)
  PSC736 (8)25 (34)11 (15)68 (93) 17/61 (28)

Note—Values are number of cases with percentages in parentheses. Missing values indicate inability to deduce a value from the data in an article. CBD = common bile duct, AIP = autoimmune pancreatitis (same condition as IgG4-related systemic disease), PSC = primary sclerosing cholangitis, ISD = IgG4-related systemic disease.

In clinical practice, MRI is the preferred noninvasive modality for imaging of the biliary tree after sonography. The Asian consensus criteria [33] for diagnosis of ISD are ERCP for the initial diagnosis and MRI for follow-up. Our MRI-based findings do not show preponderance to the involvement of the distal CBD. This may be because our patient group had a mixture of acute and chronic presentations of ISD. The case reports that describe involvement of the biliary tree in ISD without a pancreatic mass [2224] do not show the current accepted ERCP pattern involvement of CBD. It is possible that a distal CBD stricture is seen in acute ISD with pancreatic mass due to inflammatory change and mass effect from the pancreas.
Khalili et al. [41] described renal cortical lesions and other extrapancreatic features of ISD and their usefulness for differentiating ISD from pancreatic cancer. Kamisawa et al. [34] and Chari and Sugumar and their colleagues [35, 36] described strategies for differentiating ISD from pancreatic cancer. They included the presence of intrahepatic or proximal extrahepatic bile duct strictures as criteria for differentiating ISD presenting as a pancreatic mass from pancreatic cancer. There is a high incidence of biliary involvement in ISD, as shown by our data. Our study showed no association between the location or the length of CBD strictures and the presence of intrahepatic bile duct disease in ISD compared with PSC. In patients with a pancreatic mass, however, our data support the criteria of biliary strictures used in both these strategies for differentiating ISD from pancreatic cancer.
We acknowledge limitations to this study. The group of patients with other autoimmune diseases was mainly used to increase statistical power and decrease bias. The study was retrospective in design. However, to achieve larger patient numbers and be inclusive of multiple study cohorts, a retrospective study was considered most practical for preventing an unusually long prospective study. Consensus reader analysis was another limitation. However, because the primary purpose of the study was observational (to report on morphologic patterns of biliary involvement to seek any distinguishable pattern for differentiating ISD from PSC and other autoimmune diseases) rather than a test of diagnostic imaging accuracy for these diseases, we consider consensus reading acceptable if not optimal. During our study we found one case that had been reclassified from PSC to ISD. The patient underwent a liver transplant because of liver failure. The preoperative diagnosis was PSC, but immune staining of the liver explant showed ISD. Although the diagnosis of PSC and other autoimmune liver disease was based on the American Association for the Study of Liver Diseases diagnostic criteria, it is still possible that there was overlap in diagnosis in the context of current understanding of these diseases. Imaging was performed with a combination of 1.5-T and 3-T MRI systems. This is not unusual because the study included patient cohorts over a long period, and evolving technology is bound to create heterogeneity in the MRI platforms. Furthermore, we continue to use a mix of MRI systems in clinical practice, and the study reflects this practice. Finally, MRI to evaluate temporal changes in the bile ducts after steroid treatment and in patients with relapses would help confirm the distribution of biliary disease in ISD, but the objective of our observational study was limited to evaluation of morphologic patterns at clinical presentation.

Conclusion

We propose that increased single-layer CBD wall thickness greater than 2.5 mm, continuous involvement of the biliary tree, gallbladder involvement, and absence of liver parenchymal abnormalities are accurate predictors of ISD and can be useful for differentiating it from PSC and other autoimmune diseases, such as primary biliary cirrhosis and autoimmune hepatitis, particularly when the pancreatic abnormality is not contributory to a diagnosis of ISD.

References

1.
Yoshida K, Toki F, Takeuchi T, et al. Chronic pancreatitis caused by an autoimmune abnormality: proposal of the concept of autoimmune pancreatitis. Dig Dis Sci 1995; 40:1561–1568
2.
Nishino T, Toki F, Oyama H, et al. Biliary tract involvement in autoimmune pancreatitis. Pancreas 2005; 30:76–82
3.
Ghazale A, Chari ST, Zhang L, et al. Immunoglobulin G4–associated cholangitis: clinical profile and response to therapy. Gastroenterology 2008; 134:706–715
4.
Nishino T, Oyama H, Hashimoto E, et al. Clinicopathological differentiation between sclerosing cholangitis with autoimmune pancreatitis and primary sclerosing cholangitis. J Gastroenterol 2007; 42:550–559
5.
Uehara T, Hamano H, Kawa S, Sano K, Honda T, Ota H. Distinct clinicopathological entity ‘auto-immune pancreatitis associated sclerosing cholangitis'. Pathol Int 2005; 55:405–411
6.
Hirano K, Shiratori Y, Komatsu Y, et al. Involvement of the biliary system in autoimmune pancreatitis: a follow-up study. Clin Gastroenterol Hepatol 2003; 1:453–464
7.
Notohara K, Burgart LJ, Yadav D, Chari ST, Smyrk TC. Idiopathic chronic pancreatitis with periductal lymphoplasmacytic infiltration: clinicopathologic features of 35 cases. Am J Surg Pathol 2003; 27:1119–1127
8.
Bodily KD, Takahashi N, Fletcher JG, et al. Auto-immune pancreatitis: pancreatic and extrapancreatic imaging findings. AJR 2009; 192:431–437
9.
Fujinaga Y, Kadoya M, Kawa S, et al. Characteristic findings in images of extrapancreatic lesions associated with autoimmune pancreatitis. Eur J Radiol 2010; 76:228–238
10.
Kojima E, Kimura K, Noda Y, Kobayashi G, Itoh K, Fujita N. Autoimmune pancreatitis and multiple bile duct strictures treated effectively with steroid. J Gastroenterol 2003; 38:603–607
11.
Park DH, Kim MH, Chari ST. Recent advances in autoimmune pancreatitis. Gut 2009; 58:1680–1689
12.
Sahani DV, Sainani NI, Deshpande V, Shaikh MS, Frinkelberg DL, Fernandez–del Castillo C. Autoimmune pancreatitis: disease evolution, staging, response assessment, and CT features that predict response to corticosteroid therapy. Radiology 2009; 250:118–129
13.
Wolfson D, Barkin JS, Chari ST, et al. Management of pancreatic masses. Pancreas 2005; 31:203–217
14.
Hardacre JM, Iacobuzio-Donahue CA, Sohn TA, et al. Results of pancreaticoduodenectomy for lymphoplasmacytic sclerosing pancreatitis. Ann Surg 2003; 237:853–859
15.
Nakazawa T, Ohara H, Sano H, et al. Cholangiography can discriminate sclerosing cholangitis with autoimmune pancreatitis from primary sclerosing cholangitis. Gastrointest Endosc 2004; 60:937–944
16.
Fulcher AS, Turner MA, Franklin KJ, et al. Primary sclerosing cholangitis: evaluation with MR cholangiography—a case-control study. Radiology 2000; 215:71–80
17.
Dave M, Elmunzer BJ, Dwamena BA, Higgins PD. Primary sclerosing cholangitis: meta-analysis of diagnostic performance of MR cholangiopancreatography. Radiology 2010; 256:387–396
18.
Chapman R, Fevery J, Kalloo A, et al. Diagnosis and management of primary sclerosing cholangitis. Hepatology 2010; 51:660–678
19.
Chari ST, Smyrk TC, Levy MJ, et al. Diagnosis of autoimmune pancreatitis: the Mayo Clinic experience. Clin Gastroenterol Hepatol 2006; 4:1010–1016
20.
Aljiffry M, Renfrew PD, Walsh MJ, Laryea M, Molinari M. Analytical review of diagnosis and treatment strategies for dominant bile duct strictures in patients with primary sclerosing cholangitis. HPB (Oxford) 2011; 13:79–90
21.
Nishimori I, Onishi S, Otsuki M. Nationwide survey for autoimmune pancreatitis in Japan. Pancreas 2008; 37:117–118
22.
Ryuzaki H, Yamamoto T, Masuoka S, et al. An unusual case of IgG4-related sclerosing cholangitis of the peripheral bile duct with no pancreatic lesion. Hepatol Int 2009; 3:1936
23.
Haraoka K, Morishita Y, Urata A, Miyase S, Takaaki IT, Fujiyama S. A case of immunoglobulin G4-related sclerosing cholangitis without pancreatic lesions. Gastroenterol Endosc 2008; 50:3054–3061
24.
Nijs J, Macken E, Struyf N, et al. Autoimmune pancreatitis with evolution to cholangitis: a case report. Acta Gastroenterologica Belg 2004; 346–350
25.
Hamano H, Kawa S, Uehara T, et al. Immunoglobulin G4-related lymphoplasmacytic sclerosing cholangitis that mimics infiltrating hilar cholangiocarcinoma: part of a spectrum of auto-immune pancreatitis? Gastrointest Endosc 2005; 62:152–157
26.
Ghazale A, Chari ST, Smyrk TC, et al. Value of serum IgG4 in the diagnosis of autoimmune pancreatitis and in distinguishing it from pancreatic cancer. Am J Gastroenterol 2007; 102:1646–1653
27.
Oseini AM, Chaiteerakij R, Shire AM, et al. The utility of serum immunoglobulin G4 in distinguishing immunoglobulin G4-associated cholangitis from cholangiocarcinoma. Hepatology 2011; 54:940–948
28.
De Valle MB, Bjornsson E, Lindkvist B. Prevalence of elevated serum immunoglobulin G4 in a population based cohort of primary sclerosing cholangitis. Gastroenterology 2010; 138 (suppl 1) S9–S10
29.
Zhang L, Lewis JT, Abraham SC, et al. IgG4+ plasma cell infiltrates in liver explants with primary sclerosing cholangitis. Am J Surg Pathol 2010; 34:88–94
30.
Nasr JY, Nalesnik M, Demetris AJ, et al. A comparative study of orthotopic liver transplant (OLTX) for sclerosing cholangitis in patients with and without elevated serum IgG4 levels. Gastroenterology 2011; 140 (suppl 1):S907
31.
Alswat K, Al-Harthy N, Mazrani W, et al. The spectrum of sclerosing cholangitis and the relevance of IgG4 elevations in routine practice. Am J Gastroenterol 2012; 107:56–63
32.
Kwon S, Kim MH, Choi EK. The diagnostic criteria for autoimmune chronic pancreatitis: it is time to make a consensus. Pancreas 2007; 34:279–286
33.
Otsuki M, Chung JB, Okazaki K, et al. Asian diagnostic criteria for autoimmune pancreatitis: consensus of the Japan-Korea Symposium on Autoimmune Pancreatitis. J Gastroenterol 2008; 43:403–408
34.
Kamisawa T, Imai M, Chen PY, et al. Strategy for differentiating autoimmune pancreatitis from pancreatic cancer. Pancreas 2008; 37:e62–e67
35.
Chari ST, Takahashi N, Levy MJ, et al. A diagnostic strategy to distinguish autoimmune pancreatitis from pancreatic cancer. Clin Gastroenterol Hepatol 2009; 7:1097–1103
36.
Sugumar A, Takahashi N, Chari ST. Distinguishing pancreatic cancer from autoimmune pancreatitis. Curr Gastroenterol Rep 2010; 12:91–97
37.
Nakazawa T, Ohara H, Sano H, et al. Clinical differences between primary sclerosing cholangitis and sclerosing cholangitis with autoimmune pancreatitis. Pancreas 2005; 30:20–25
38.
Abraham SC, Cruz-Correa M, Argani P, et al. Lymphoplasmacytic chronic cholecystitis and biliary tract disease in patients with lymphoplasmacytic sclerosing pancreatitis. Am J Surg Pathol 2003; 27:441–451
39.
Wang WL, Farris AB, Lauwers GY, et al. Autoimmune pancreatitis–related cholecystitis: a morphologically and immunologically distinctive form of lymphoplasmacytic sclerosing cholecystitis. Histopathology 2009; 54:829–836
40.
Vitellas KM, Keogan MT, Freed KS, et al. Radiologic manifestations of sclerosing cholangitis with emphasis on MR cholangiopancreatography. RadioGraphics 2000; 20:959–975
41.
Khalili K, Doyle DJ. Chawla TP, et al. Renal cortical lesions in patients with autoimmune pancreatitis: a clue to differentiation from pancreatic malignancy. Eur J Radiol 2008; 67:329–335
42.
Craig DA, MacCarty RL, Wiesner RH, Grambsch PM, LaRusso NF. Primary sclerosing cholangitis: value of cholangiography in determining the prognosis. AJR 1991; 157:959–964
43.
Heuck A, Maubach P, Reiser M, et al. Age-related morphology of the normal pancreas on computed tomography. Gastrointest Radiol 1987; 12:18–22
44.
Naitoh I, Zen Y, Nakazawa T, et al. Small bile duct involvement in IgG4-related sclerosing cholangitis: liver biopsy and cholangiography correlation. J Gastroenterol 2011; 46:269–276

Information & Authors

Information

Published In

American Journal of Roentgenology
Pages: 536 - 543
PubMed: 24555589

History

Submitted: November 23, 2012
Accepted: June 3, 2013

Keywords

  1. autoimmune pancreatitis
  2. IgG4
  3. primary sclerosing cholangitis
  4. systemic disease

Authors

Affiliations

Ajay Tokala
Department of Medical Imaging, Abdominal Imaging, University Health Network, Mount Sinai Hospital and Women's College Hospital, University of Toronto, 610 University Ave, Toronto, ON M5G 2M9, Canada.
Present address: Department of Radiology, Royal Preston Hospital, Fulwood, Preston, UK.
Korosh Khalili
Department of Medical Imaging, Abdominal Imaging, University Health Network, Mount Sinai Hospital and Women's College Hospital, University of Toronto, 610 University Ave, Toronto, ON M5G 2M9, Canada.
Ravi Menezes
Department of Medical Imaging, Abdominal Imaging, University Health Network, Mount Sinai Hospital and Women's College Hospital, University of Toronto, 610 University Ave, Toronto, ON M5G 2M9, Canada.
Gideon Hirschfield
Liver Centre, University Health Network, Mt. Sinai Hospital, University of Toronto, Toronto, ON, Canada.
Present address: Centre for Liver Research, University of Birmingham, Birmingham, UK.
Kartik S. Jhaveri
Department of Medical Imaging, Abdominal Imaging, University Health Network, Mount Sinai Hospital and Women's College Hospital, University of Toronto, 610 University Ave, Toronto, ON M5G 2M9, Canada.

Notes

Address correspondence to K. S. Jhaveri ([email protected]).

Metrics & Citations

Metrics

Citations

Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

Media

Figures

Other

Tables

Share

Share

Copy the content Link

Share on social media