Discovery of elevated cholestatic liver biochemistries (alkaline phosphatase and gamma GT) in an asymptomatic IBD patient should always prompt consideration of the diagnosis of PSC. Blood tests typically fluctuate over time and at times may even return completely to normal. Autoantibody tests are of little diagnostic significance. IgM concentrations are increased in about 50 % of patients with advanced PSC.

Serum IgG4 levels should be measured in all patients with suspected PSC. They are elevated in about 9 % of patients and are associated with a worse outcome [33] (see Section Histology).

The first imaging modality recommended in the workup for a patient with cholestatic LFTS is a transabdominal ultrasound but in the vast majority of PSC patients, this will be nondiagnostic (Fig. 52.2). Usually, a diagnosis of PSC is made when cholangiography (MRCP or endoscopic retrograde pancreatography (ERCP)) demonstrates characteristic bile duct changes of multifocal stricturing and segmental dilatations, causing a “beaded appearance,” in the absence of a secondary cause (Fig. 52.3).

Patients with small duct PSC have normal cholangiographic findings—this subgroup, who share similar biochemical and clinical features to large duct PSC, are instead diagnosed when histological changes of PSC are demonstrated on liver biopsy [35] (see Section Laboratory investigations).

Magnetic resonance cholangiopancreatography (MRCP) is now the investigation of choice over endoscopic retrograde cholangiopancreatography (ERCP) . MRCP is noninvasive, does not involve radiation, avoids ERCP complications such as pancreatitis and is comparable to ERCP for diagnosis of PSC with good interobserver agreement [36, 37]. ERCP may still have a place in patients where the diagnosis remains uncertain after MRCP and is most useful for imaging subtle abnormalities in the intrahepatic biliary tree . As yet, there is no data on the utility of CT cholangiography for PSC diagnosis.

When radiological findings support the diagnosis of PSC, histological examination of the liver is not required to confirm this and only exposes patients to unnecessary morbidity.

Histology is diagnostic in only one third of PSC patients, although in another third there may be findings suggestive of biliary disease. The characteristic early biopsy findings of PSC are inflammation with periductal “onion-skin” fibrosis, portal edema, and bile ductular proliferation resulting in expansion of portal tracts (Fig. 52.4 ). With disease progression, bridging fibrosis eventually leads to cirrhosis. The focal nature of both early and late changes in PSC can make “staging” liver biopsies unreliable [38].

When a diagnosis of PSC is suggested by imaging and histology, there are a number of causes of secondary sclerosing cholangitis (SSC) that must be considered (Table 52.2). At times, it can be very difficult to distinguish these from PSC, particularly in PSC patients who have coexisting pathology such as choledocholithiasis. In these patients, the clinical history, presence of IBD, and distribution of cholangiographic abnormalities are most helpful in identifying the predominant disease process [39]. It is particularly important to exclude IgG4-related sclerosing cholangitis.

Small duct PSC is normally diagnosed in the patients with IBD who have cholestatic serum biochemistry with a normal cholangiogram, but it may occur in patients without IBD. It is characterized by histological changes on liver biopsy characteristic of PSC. It occurs in approximately 10 % of the PSC population [40–42]. A recent study from the Calgary health region in Canada has shown an incidence of small duct PSC as 0.15/100,000. In children the incidence rate was 0.23/100,000 compared with 1.11/100,000 in adults [7].

Small duct PSC typically runs a milder course than large duct disease with a reduced likelihood of progression to cirrhosis and with a significantly improved survival compared with large duct disease. To date there have been no reports of cholangiocarcinoma in the small duct PSC patient population [43, 44]. Approximately one quarter of small duct PSC patients will subsequently develop large duct disease over a period of 10 years [43–47].

Various studies suggest that between 1.4 % and 8 % of PSC patients have coexisting autoimmune liver disease (AIH)—recently defined as PSC-AIH syndrome [48–51]. PSC-AIH is more commonly found in children and young adults and characterized by clinical, biochemical, and histological features of AIH in the presence of cholangiographic findings identical to PSC [44, 46, 52–55]. PSC-AIH should be considered if the aminotransferase level is elevated more than twice the upper limit of normal and the serum IgG. Rarely, AIH features can develop in patients with established PSC. A liver biopsy should always be performed in these patients to confirm the diagnosis before treating with immunosuppressants. Immunosuppression is only helpful in improving disease progression in this selected group [56–59].

Elevated IgG4 serum levels were first reported more than 10 years ago in patients with autoimmune pancreatitis (AIP) [60]. These patients commonly have associated intra- and extrahepatic biliary stricturing that may mimic PSC. In 2004, a Japanese case series found a subset of PSC patients had a significant infiltrate of IgG4-positive plasma cells isolated to the biliary tree (in the absence of any pancreatic abnormality) [61]. This group is now recognized to represent a distinct clinical entity that, like autoimmune pancreatitis may be responsive to immunosuppression. The term IgG4-related sclerosing cholangitis is used to encompass both this patient group and those patients who have biliary pathology in association with autoimmune pancreatitis [34].

Recent retrospective studies have found elevated serum IgG4 levels in 9–12 % of PSC patients with levels up to twice the upper limit of the normal range [33, 62]. IgG4 positive PSC patients have a reduced incidence of IBD and a more severe disease course when compared with patients who are negative [33].

A recent reevaluation of 98 consecutive liver transplants for patients originally diagnosed with PSC, found that 23 (23 %) of explanted livers stained positive for IgG4. Serum IgG4 levels were elevated in 18 of these patients. Tissue IgG4 positivity was associated with a more aggressive clinical course manifested by shorter time to transplant and a higher likelihood of disease recurrence after liver transplantation [63].

The important clinical implications and potential benefit of immunosuppression in this patient group make it imperative that IgG4 cholangitis is recognized and serum IgG4 levels should be checked in all PSC patients. It is now established that IgG4 disease is a distinct disease entity from PSC patients with and without elevated IgG4 levels [64].

There is a strong association between IBD and PSC. In patient series from Northern Europe and North America, the prevalence of IBD in people with PSC ranges between 60 and 80 % [65–67]. There are, however, significant geographical variations in the reported prevalence of IBD in PSC with a much weaker association found in certain countries (Table 52.3). This discrepancy may partly be explained because studies reporting a weaker association often only used sigmoidoscopy (and/or no biopsies) subsequently missing a significant proportion of colitis, which may be right sided and microscopic in PSC patients.

The predominant form of IBD is ulcerative colitis (UC) with approximately 85 % of IBD/PSC patients affected [67, 74]. IBD can be diagnosed at any time in patients with PSC but most commonly the diagnosis predates PSC by several years [66, 74, 75]. Interestingly, even patients who have received liver transplantation for PSC continue to be at increased risk of developing IBD [76].

All patients with a new diagnosis of PSC should have a full colonoscopy with biopsies, regardless of the presence of gastrointestinal symptoms to evaluate for IBD [34]. Because asymptomatic colitis and rectosigmoid sparing are common features in IBD/PSC, a flexible sigmoidoscopy is insufficient for screening. It is unclear whether interval endoscopies should be performed if a patient remains symptom free following a colonoscopy with normal colonic histology. Some clinicians advocate repeat endoscopy every 5 years.

It is likely that PSC/UC represents a distinct UC genotype and phenotype [74]. The natural history of UC in patients with PSC has a more benign course than in those patients with UC alone despite the fact that the disease usually involves the whole colon [77]. Common features in the PSC/IBD group include: rectal sparing (52 % versus 6 %), backwash ileitis (51 % versus 7 %), and an increased prevalence of pouchitis (following colectomy and ileo-anal pouch formation). The reason for increased rates of pouchitis is unknown (Table 52.4).

Early small, uncontrolled series suggested that PSC/Crohn’s typically manifests as extensive colitis and also that isolated small bowel Crohn’s is not associated with PSC [67, 82]. A recent case-controlled study specifically examined the course of Crohn’s disease in 39 patients with PSC/Crohn’s compared with Crohn’s patients without PSC. The study confirmed that isolated ileal disease is rare (6 % vs. 31 %). Interestingly, unlike PSC/UC , the two groups followed similar disease courses, as judged by the need for surgical intervention or significant medical therapy (defined as requiring biological therapy with TNFα antagonists or greater than 5 courses of corticosteroids). The PSC/Crohn’s patients were more likely to be female than PSC/UC patients (50 % vs. 28 %) and were more likely to have small duct PSC. Interestingly, they were less likely to progress to cancer, liver transplantation and death [17].

The true prevalence of PSC within the IBD patient population is unknown because until recently, accurate data have required invasive cholangiography to be carried out on unselected patient groups and PSC patients may have normal liver biochemistry. Work underway using noninvasive MRCP will hopefully provide this data in the near future. Available evidence suggests that approximately 5 % of UC patients have coexisting PSC although this is likely to be an underestimate [83]. This data comes from three major studies (Table 52.5). The largest of these, found that 5 % of a cohort of 1500 UC patients had elevated alkaline phosphatase levels, and in those who subsequently underwent ERCP, 85 % had evidence of PSC [4].

The prevalence of PSC in Crohn’s disease is significantly less than that of UC, with an estimated percentage of 3–4 % of patients affected [82]. Typically these patients have either ileo-colonic or extensive colonic Crohn’s disease.

Both the development of PSC and its outcome are independent of the activity of colitis. It may even occur after proctocolectomy. Interestingly, however, colectomy prior to transplantation for PSC is protective against the development of PSC in the transplanted liver in male patients [86].

The increased risk of colorectal cancer (CRC) in ulcerative colitis compared with the general population is well established [79, 88–91]. Based on a meta-analysis of 11 studies, in UC patients with coexisting PSC this risk is elevated five times higher again (OR of 4.79 (95 % CI 3.58–6.41)). This risk increases with time and continues even after liver transplantation [80, 92] (Fig. 52.5).

Although CRC can present at any time in the disease course, the median time from diagnosis of colitis to development of colorectal cancer is 17 years [94]. Interestingly, the majority of these cancers (76 %) are right sided [79]. It has been proposed that this right-sided predominance may result from a carcinogenic effect of increased cecal concentrations of secondary bile acids such as lithocholic acid [88].

Considering the absolute risk for colonic dysplasia or cancer in PSC/UC approaches 31 % after 20 years of colitis [95] it is understandable that guidelines recommend 1–2 year interval colonoscopies with biopsies from the time of diagnosis of PSC/IBD. Patients with Crohn’s/PSC are included in this recommendation, although a recent study has shown a low prevalence of dysplasia and cancer in the PSC patients with associated Crohn’s colitis compared to PSC/UC [96]. This was not confirmed in a Swedish cohort [97].

Ursodeoxycholic acid (UDCA) treatment in patients with PSC and UC may decrease the risk of colorectal dysplasia and colorectal cancer [98, 99] (see Section Cholangitis).

In patients who develop PSC-associated colorectal malignancy , proctocolectomy with ileo-anal pouch formation is the preferred surgical management as it avoids ) the complication of peristomal varices in patients with an ileal stoma.

Cholangiocarcinoma complicates the clinical course of PSC in 10–20 % of patients, with an annual incidence (starting 1 year after diagnosis of PSC) of 0.5–1.5 % [65, 100–103]. Male gender, smoking and a long history of IBD were identified as risk factors for CCA in a case-controlled review of 39 PSC patients presenting with CCA [104].

One third of patients who develop CCA are diagnosed within 1 year of the diagnosis of their PSC. The likely explanation for this is that the development of symptomatic cholangiocarcinoma brings a number of patients with previously unrecognized PSC to medical attention.

The diagnosis of CCA can be challenging in PSC patients and early detection is difficult. Often patients with CCA are asymptomatic and when symptoms develop they are nonspecific, typically indicate metastatic disease and mimic PSC disease progression [105]. Worsening jaundice/bilirubin levels, pruritus, weight loss, and abdominal pain in any PSC patient should always prompt evaluation for CCA. Unfortunately, computed tomography (CT), ultrasonography (US), and MRCP have poor sensitivity for early detection of CCA [106]. Annual MRCP has been advocated as a surveillance method for CCA in patients with large duct PSC, but there is no evidence to support this approach.

Tumor markers play a limited role in the early detection of CCA [34, 103, 107–110]. Using a cut-off level for Carbohydrate Antigen 19-9 (CA19-9) of 130U/ml (normal <55), the sensitivity and specificity is 79 % and 98 % respectively [108]. CA19-9 can be elevated in benign biliary disease as well as other malignancies including pancreas, colon, stomach, and gynecological cancers. It is important to note that 7 % of the general population (those negative for Lewis antigen) will not produce CA 19-9 regardless of their CCA status. Generally, CA 19-9 has only been useful to identify patients with advanced, unresectable CCA and has no role in cancer surveillance in PSC. Elevated levels of serum trypsinogen-2 were reported to represent an early marker of CCA, but this needs confirmation in a larger study [111].

CCA can develop in any part of the biliary tree. Perihilar CCAs (the so-called Klatskin tumors) are the most common tumor site. These are typically infiltrating, desmoplastic tumors. Mass lesions in the liver make up about 20 % of cancers and distal CBD tumors are uncommon (<10 %) [112].

The discovery of a dominant peripheral stricture on imaging presents a significant diagnostic challenge (Fig. 52.6). The majority of these strictures will be benign. It is uncommon for a coexisting mass to be identified, but if found, this has a virtually 100 % sensitivity and specificity for CCA [103]. In a recent retrospective study of 89 patients with dominant strictures, MRCP was found to be superior to ERCP for the diagnosis of CCA with a sensitivity, specificity and accuracy of 96 %, 85 %, 81 % compared with 80 %, 75 %, and 78 % respectively [113]. One advantage of ERCP, however, is that it does allow acquisition of cytological specimens using brushings. It is also important to exclude IgG4-related disease in this patient group, as it may mimic PSC and CCA on cholangiography.

Conventional brush cytology specimens obtained through ERCP have poor sensitivity, 18–40 % [103, 110, 114–116]. However, when positive, the specificity approaches 100 %. New diagnostic modalities such as direct image analysis (DIA) and fluorescent in situ hybridization (FISH) have been employed in attempts to increase diagnostic yield on cytological specimens. Early studies indicated a significant improvement in sensitivity [115], but more recently this has been brought into question with the most recent study using FISH polysomy demonstrating a sensitivity of 46 %, only slightly higher than cytology alone [117].

Distinguishing PSC and CCA by direct visualization of the biliary tree with techniques such as endoscopic intraductal US and cholangioscopy is promising. A recent study reported a sensitivity of 92 % and specificity of 93 % for diagnosing malignant strictures compared to 66 % and 51 % respectively using ERCP alone [118]. However, these techniques have not yet been tested in large patient populations. PET is not helpful to distinguish PSC from CCA [119].

Unfortunately, the prognosis for CCA is poor. The median survival is 5–11 months [101, 120, 121]. Most patients are treated palliatively. Even with surgical resection, the 3-year survival is <30 % [122, 123]. The Mayo Clinic in the USA has pioneered the use neoadjuvant chemoradiotherapy, followed by liver transplantation in highly selected patients with nonmetastatic cancers of less than 3 cm in diameter. This approach in this highly selected group, has achieved remarkable 5-year survival rates of up to 82 % [124, 125].

PSC is recognized as a major risk factor for gallbladder carcinoma . In a study examining 286 patients with PSC, 18 (6 %) of patients were found to have gallbladder mass lesions of which 10 were gallbladder cancer [126]. Guidelines recommend annual US surveillance, and that cholecystectomy should be performed for any mass lesion (such as polyps) identified regardless of size.

One study has suggested that the risk of pancreatic carcinomas may also be increased in PSC. This remains to be confirmed [87].

Due to an absence of population-based studies, the exact prevalence of dominant strictures is unknown but estimated to occur in approximately 10–58 % of PSC patients during follow-up [71, 127, 128]. The majority of these strictures are benign, but their discovery should always prompt consideration of a CCA (see Section IBD in PSC patients). Dominant strictures are associated with bacterial cholangitis as well as biliary stones.

Patients with symptoms from dominant strictures such as cholangitis, jaundice, pruritus, right upper quadrant pain or worsening biochemical indices are appropriate candidates for therapy. ERCP with sphincterotomy and balloon dilatation (with prophylactic pre-procedural antibiotics) has the best evidence of benefit [129–134]. Stenting should be reserved for strictures that are refractory to repeat dilatation. Large retrospective studies suggest that endoscopic therapy results in clinical improvement and prolonged survival.

It is very unusual for PSC patients to develop cholangitis in the absence of a previous history of instrumentation of the biliary tree or a dominant stricture [135, 136]. The presence of a dominant stricture is thought to predispose to cholangitis through bacterial colonization of stagnant of bile acids.

A recent study demonstrated that 40.5 % of patients with a dominant stricture had bacterial colonization of their biliary tree (In the absence of a stricture, patient bile samples were all sterile) [137]. Most infections are caused by aerobic, enteric organisms such as E. coli, Klebsiella, and E. faecalis [138]. Unfortunately, short-course antibiotics are poorly effective in eradicating bacteria. However, biliary drainage procedures combined with antibiotics are usually effective. Patients with recurrent bacterial cholangitis may benefit from cyclical prophylactic long-term antibiotics usually quinolones. In patients who fail antibiotic therapy, recurrent cholangitis may be the primary indication for OLTx.

Fatigue is a common problem and has a major impact on the quality of life in PSC patients [138]. It occurs independently of the severity of liver disease. No medical therapy is of proven benefit although the antinarcolepsy agent modenafil has been tried off-label with varying degrees of success.

Pruritus is also common in patients with PSC. 80–90 % of patients will respond to first line therapy with cholestyramine [139].

The antibiotic rifampicin is used as a second line agent for patients who have not responded to cholestyramine. It has shown to be effective treatment for pruritus in cholestatic liver disease [140–142]. The mechanism of action is unknown but it probably induces the hepatic microsomal drug-metabolizing system causing metabolism of endogenous pruritogenic compounds. It may alternatively have a direct antimicrobial effect in the intestinal lumen, causing a change in synthesis of secondary bile acids. It is also a nuclear receptor agonist of PXR.

There is an upregulation of opioid receptors in patients with chronic cholestasis as well as an increased central opioidergic tone and opioid antagonists (nalmefene, naloxone, and naltrexone) have been shown to be effective in pruritus. These drugs can cause a withdrawal-like reaction by preventing binding of endogenous opioids but this complication is transient and settles within the first couple of days [143–145].