Primary Sclerosing Cholangitis – Approach to Diagnosis

Introduction

Primary sclerosing cholangitis (PSC) is a progressive disease of the biliary tract characterized by diffuse inflammation and fibrosis of both intra- and extrahepatic bile ducts. According to population-based studies, its prevalence is estimated at 1-10 per 100,000 persons in the United States.[1,2] Disease progression is slow and highly variable, but eventually culminates in cirrhosis and complications related to portal hypertension, with a 10-year survival of approximately 65%.[1,3,4] Problems more specific to PSC can also develop, such as recurrent ascending cholangitis and cholangiocarcinoma.

Although neither the etiology nor the pathogenesis of PSC is well understood, PSC is thought to be an immune-mediated process and is indeed associated with other autoimmune diseases.[5] The vast majority of PSC cases occur in the setting of inflammatory bowel disease (IBD) and its association with IBD is well recognized, especially ulcerative colitis. Although PSC occurs in approximately 5% of all patients with ulcerative colitis, up to 75% of those with PSC will have associated ulcerative colitis.[1,6–8]

It can be extremely difficult to diagnose PSC with confidence, particularly outside of tertiary care centers. This article uses case scenarios to illustrate the diagnostic process in PSC, and follows with a review of the current understanding of how clinical presentation, laboratory tests, endoscopy, radiology, and histology facilitate the diagnosis of PSC.

Case Descriptions

Case 1

A 19-year-old white man was found to have abnormal serum liver biochemistry test results 3 years ago when he was first diagnosed as having ulcerative colitis. He denied any previous transfusions, gallstone disease, intravenous drug use, tattoos, or significant alcohol consumption. The patient had no family history of liver disease and, with the exception of his ulcerative colitis, he reported that he was healthy. Even after stopping all medications, his liver function test results remained abnormal. His laboratory studies were significant for increased serum alkaline phosphatase (ALP) and aminotransferase levels elevated to 3 times the upper limit of normal. His bilirubin level was normal (< 1 mg/dL). Endoscopic retrograde cholangiopancreatography (ERCP) was performed and showed multiple stenoses and dilatations throughout the intrahepatic biliary system, with a pruning type pattern. The common bile duct and extrahepatic portion of the common hepatic duct appeared normal, without strictures. These biliary findings coupled with the presence of inflammatory bowel disease (IBD) and the absence of any alternative etiology led to a confident diagnosis of PSC.

Case 2

A 47-year-old white woman was found to have abnormal liver function test results during a routine health maintenance examination. She had always been asymptomatic and denied any social risk factors for acquired liver disease. Laboratory studies showed the following: serum aspartate aminotransferase (AST), 68 U/mL (normal, 10-39 U/mL); serum alanine aminotransferase (ALT), 50 U/mL (normal, 10-39 U/mL); serum total bilirubin, 0.8 mg/dL (normal, 0.1-1.0 mg/dL); and ALP, 289 U/L (normal, 45-129 U/L). Serologic markers, including antimitochondrial antibody and antinuclear antibody, were normal, as were immunoglobulins, including IgG4 and carbohydrate antigen 19-9 (CA19-9). Magnetic resonance cholangiopancreatography (MRCP) was performed and showed hepatic ductal dilatation, with no evidence of mass lesions. The patient subsequently underwent ERCP at an outside facility, which showed right hepatic ductal dilatation without visualization of the left hepatic ducts. This finding was indicative for obstruction, and malignancy could not be ruled out. Therefore, the patient was referred to our institution for tumor resection. A repeat MRCP was performed and showed intrahepatic ductal dilatation with tortuous and diffuse dilatation most consistent with PSC. Focal obstruction at the confluence of right and left hepatic ducts was still indicative for a Klatskin tumor. However, repeat ERCP showed tight strictures of the left hepatic duct and the first division of the right hepatic duct. Several smaller strictures were noted in the smaller intrahepatic ducts. Biopsies and cytology specimens obtained during this ERCP showed no evidence of malignancy. This patient has been observed for 1 year and her benign course corroborated a diagnosis of PSC.

Clinical Presentation

Case 1 illustrates the classic patient with PSC: a young male with history of ulcerative colitis who is found to have a cholestatic pattern on serum liver biochemistry measurement. Indeed, up to 70% of patients with PSC are male, with the mean age of 40 years at time of diagnosis.[2,9,10]

Case 2 demonstrates the diagnostic challenge of PSC because of its varying clinical presentation. Approximately 15% to 40% of patients with PSC are asymptomatic.[6-8,11] Despite the lack of symptoms, these patients have a decreased life expectancy compared with the general population.[12,13] This may be due to the fact that up to 17% of asymptomatic patients are found to have cirrhosis at diagnosis.[8]

When present, symptoms of PSC are vague and nonspecific, and usually are of little benefit in making the diagnosis. Typically, patients will go through successive periods of exacerbation and remission, with the most common complaints consisting of fatigue, right upper quadrant abdominal pain, and pruritus.[14,15] Also, particular attention should be paid to clinical signs of infection, because these patients are at increased risk for the development of bacterial cholangitis secondary to biliary obstruction. Jaundice may appear during these episodes of acute cholangitis, but unrelieved jaundice usually means progression of the underlying disease. In general, symptomatic patients are more likely to have advanced liver disease as well as a decreased survival free of liver transplantation. Finally, although unusual, PSC can present as acute liver failure. This diagnosis should especially be considered in individuals with IBD who present with acute liver failure of indeterminate etiology.[16]

Laboratory Studies

Elevated serum ALP is the most common laboratory abnormality seen in PSC. ALP levels are often found to be 3 to 10 times the upper limit of normal, whereas serum AST and ALT levels are typically 2-3 times the upper limit of normal.[9,10] The majority of patients with PSC have a normal serum total bilirubin level. An increase in bilirubin may result from stricturing due to advanced disease, acute cholangitis, choledocholithiasis, or development of malignancy. As with most liver diseases, altered prothrombin time and serum albumin levels reflect progression of disease.[3,6,11,17]

Changes in immunologic markers are not unusual. Serum immunoglobulin levels are frequently elevated, particularly IgG and IgM. Although serum IgG4 levels are usually within normal limits, approximately 9% of patients can present with elevated IgG4 (> 140 mg/dL).[18] These patients differ from those with normal serum IgG4 levels in that they have a lower frequency of associated IBD, higher PSC Mayo risk scores, and shorter time to liver transplantation, possibly representing a more severe course of disease. The pancreatogram may be abnormal in these patients, causing confusion with another entity known as autoimmune sclerosing pancreatitis.[19–21] Additional studies are needed to better define diagnostic criteria and to determine whether this subgroup of patients with PSC and increased IgG4 levels will respond similarly to steroid therapy as patients with autoimmune sclerosing pancreatitis.

As many as 97% of patients with PSC have at least one detectable autoantibody;[22] however, the presence of multiple antibodies does not correlate with disease activity.[23] Anti-smooth muscle antibodies and antinuclear antibodies are found in up to 75% of patients with PSC.[24] But when antimitochondrial antibodies are detected, consideration should be given to primary biliary cirrhosis as a more likely diagnosis. Cases of PSC-primary biliary cirrhosis overlap have been described, although exceedingly rare.[25–27] Perinuclear antineutrophil cytoplasmic antibody (p-ANCA) has been described in up to 80% of patients with PSC, as well as in 30% of their unaffected family members.[28] P-ANCA in PSC appears to be unrelated to ulcerative colitis, as it may be seen in patients with PSC without ulcerative colitis. Recent investigation has shown that antibodies to the baker's yeast Saccharomyces cerevisiae (ASCA) are found in up to 44% of patients with PSC irrespective of the presence of IBD,[23] but its significance is yet to be determined.

Gastroenterologists frequently use serum CA 19-9 as a screening test for cholangiocarcinoma in patients with PSC. However, this tumor marker is not specific for cholangiocarcinoma and levels can be elevated in many circumstances, including both malignant and benign conditions. Thus, an abnormal serum CA 19-9 level can be found not only in patients with cholangiocarcinoma, pancreatic cancer, and hepatocellular carcinoma, but also in those with PSC without cancer, alcoholic liver disease, cholangitis, autoimmune hepatitis, chronic viral hepatitis, and pancreatitis. We have previously evaluated the utility of CA 19-9 as a screening tool in patients with PSC and found that a serum value > 129 U/mL could adequately differentiate between benign and malignant strictures with a sensitivity of 78.6% and specificity of 98.5%. These numbers were in agreement with data reported by other investigators. However, the positive predictive value was only 56.6%, and almost all cases of cholangiocarcinoma that had an elevated CA 19-9 level were too advanced to qualify for any curative treatment. These findings suggest that serum CA 19-9 does not perform well as a screening test. Other diagnostic modalities are needed to identify patients who could benefit from early intervention.[29]

Radiologic Studies

ERCP traditionally had been the gold standard for the diagnosis of PSC. With its significant safety advantages and advances in quality, magnetic resonance cholangiopancreatography (MRCP) has challenged this concept. Early cholangiographic changes can include fine or deep ulcerations of the common bile duct.[2] In a small subgroup of patients, these changes can affect the cystic duct or gallbladder.[30] As PSC progresses, segmental fibrosis develops within the bile ducts, with saccular dilatation of the normal areas between them, leading to the typical “beads-on-a-string” appearance seen on cholangiography (Figure 1).[31,32] Although these strictures can be found anywhere on the biliary tree, the intrahepatic and extrahepatic bile ducts are simultaneously involved in the vast majority of cases.[2]

Figure 1.

ERCP from a patient with elevated serum ALP and a history of ulcerative colitis. The right and left intrahepatic branches show multiple diffuse areas of attenuation, stenosis, and dilatation consistent with PSC.

A subgroup representing 5% to 10% of all PSC patients will have “small-duct PSC,” with histologic features and cholestatic liver test findings typical of PSC, yet no cholangiographic changes.[24,28] Small-duct PSC may progress into large-duct disease, although the actual proportion of patients who experience progression is unknown.

The emergence of MRCP is a noninvasive method of diagnosing PSC. The typical finding on MRCP is high T2 signal intensity in wedge-shaped areas with bile duct dilatation (Figure 2 and Video). Multiple studies have compared the diagnostic accuracy of MRCP to that of ERCP (Table 1). Most of these studies have shown that the effectiveness of MRCP, read by experienced radiologists, may approach that of invasive cholangiography. Although MRCP will never replace ERCP completely, it may eventually become a better diagnostic option given its obvious safety advantages. Recent studies have also shown that MRCP may be more cost effective in certain clinical situations.[33]

Figure 2.

MRCP from the same patient as in Figure 1, showing areas of narrowing within the intrahepatic branches.

Table 1.

Efficacy of Magnetic Resonance Cholangiopancreatography

Study	Sensitivity	Specificity
Angulo et al.[39]	83%	98%
Ernst et al.[40]	100%	100%
Fulcher et al.[41]	85%-88%	92%-97%
Textor et al.[42]	88%	99%

Download video file ^{(870.8KB, mov)}

Video. MRCP “movie” showing multiple areas of stenosis and dilatations within the intrahepatic ducts and mild narrowing of the common bile duct.

Occasionally, ultrasound changes can be seen in PSC. Radiologists experienced in biliary tract imaging may pick up subtle findings such as gallbladder wall thickening, increased gallbladder volume, thickening of the bile duct, or biliary tract dilation.[34] While these changes may not be specific, they are suggestive of PSC and often lead to its eventual diagnosis.

Cholangiography performed by interventional radiologists is useful in cases where ERCP or MRCP is not available or not able to provide adequate imaging. Cholangiography is also helpful in the biopsy, drainage, and stenting of intrahepatic and extrahepatic strictures that are not reachable with ERCP.

Histopathology

Because histologic findings in PSC tend to be nonspecific, liver biopsy is used primarily for staging the disease.[35,36] The staging is based on the degree of inflammation, ductal proliferation, and ductopenia. In general, early PSC is marked by inflammation and later, as the disease progresses, fibrosis predominates. Eventually the bile duct can be replaced by a solid fibrous cord.[2,36,37] However, it is sometimes difficult to stage PSC histologically because various manifestations of the disease can occur within the same liver. Further complicating this issue, the classic description of periductal “onion skin” fibrosis (Figure 3) is present in less than 40% of biopsy specimens and is not pathognomonic for PSC.[38]

Figure 3.

Liver biopsy from a patient with stage II PSC showing concentric periductal fibrosis (“onion skin”) involving interlobular bile duct; hematoxylin-eosin, X40.

The PSC staging system is similar to that used for primary biliary cirrhosis[38]:

• Stage I (portal stage) – focal inflammation limited to portal triad, with or without bile duct abnormalities, without fibrosis;

• Stage II (periportal stage) – enlargement of portal tracts, periportal fibrosis with or without periportal inflammation;

• Stage III (septal stage) – extension of septal or bridging fibrosis with bridging necrosis; and

• Stage IV – biliary cirrhosis.

Differential Diagnosis

As previously highlighted by Case 2, one of the diagnostic dilemmas with PSC is that it can often be challenging to distinguish from secondary causes of sclerosing cholangitis and malignancy. Causes of secondary sclerosing cholangitis must be excluded before a diagnosis of PSC can be established. The most commonly described secondary causes include stone disease, infection, pancreatitis, and surgical/procedural trauma. A more comprehensive list of differential diagnoses of PSC is provided in Table 2 .

Table 2.

Differential Diagnosis of Primary Sclerosing Cholangitis

Autoimmune disorders

Immune deficiency (acquired and congenital)

Previous biliary surgery

Ischemic bile duct damage

Pancreatic disease

Infection

Infiltrative processes

Biliary neoplasm

Drug injury

Idiopathic fibrosis

Congenital abnormalities of the biliary tract

Recurrent stone disease

Another caveat is the fact that the presence of cirrhosis may interfere with interpretation of ERCP/MRCP findings. In particular, subtle intrahepatic ductal changes in PSC can be difficult to distinguish from biliary tract changes seen with cirrhosis, and although a liver biopsy can be helpful, history alone may be the only suggestion of PSC. A high index of suspicion will be necessary to make an assertive diagnosis in this setting, keeping in mind that most cases of PSC occur in patients with inflammatory bowel disease.

Conclusion

The best approach to diagnosing PSC involves a systematic investigation including historical examination, laboratory studies, and cholangiographic evaluation. Particular attention should be paid to a history of IBD, autoimmune disease, or active cholangitis.

Laboratory testing should be directed at ruling out secondary causes of sclerosing cholangitis as well as looking for autoimmune markers, which can be more indicative of PSC. Thus, once initial evaluation with a hepatic panel has suggested cholestasis, further testing with serum autoantibodies and immunoglobulins is recommended.

As shown in Table 1, the sensitivity and specificity of MRCP in facilities with radiologists experienced in biliary tract diseases approaches that of ERCP. Given the improved safety profile of MRCP compared with ERCP, we anticipate that MRCP will soon be considered as a first-line imaging investigation in centers with skilled radiologists. Once signs of obstruction, dominant strictures, or inconclusive findings are described, the clinician may proceed with an ERCP for diagnostic and therapeutic maneuvers. If the patient's presentation is suggestive of cholangitis, an ERCP should be the procedure of choice.

Finally, liver biopsy should be reserved for cases in which other studies are inconclusive or when a diagnosis of small-duct PSC or an overlap disorder is being considered.