EASL Clinical Practice Guidelines

Primary sclerosing cholangitis

Primary sclerosing cholangitis (PSC) is a chronic, cholestatic liver disease that is characterized by an inflammatory and fibrotic process affecting both intra- and extrahepatic bile ducts [74]. The disease leads to irregular bile duct obliteration, including formation of multifocal bile duct strictures. PSC is a progressive disorder that eventually develops into liver cirrhosis and liver failure. The etiology of PSC is unknown, but there is evidence that genetic susceptibility factors are involved [75]. The male to female ratio is approximately 2:1. PSC can be diagnosed in children as well as in the elderly, but mean age at diagnosis is around 40 years. Up to 80% of PSC patients have concomitant inflammatory bowel disease (IBD) that in the majority of cases is diagnosed as ulcerative colitis (UC). Thus, the “typical” PSC patient is a young to middle-aged man with IBD who presents with biochemical and/or clinical signs of a cholestatic liver disease.

Diagnosis of PSC

A diagnosis of PSC is made in patients with elevated serum markers of cholestasis (AP, γGT) not otherwise explained, when magnetic resonance cholangiopancreatography (MRCP) or endoscopic cholangiopancreatography (ERCP) show characteristic bile duct changes with multifocal strictures and segmental dilatations, and causes of secondary sclerosing cholangitis [76] and other cholestatic disorders are excluded. Patients who present with clinical, biochemical and histological features compatible with PSC, but have a normal cholangiogram, are classified as small duct PSC.

Signs and symptoms

About 50% of PSC patients are symptomatic at first presentation. Typical symptoms include pruritus, pain in the right upper abdominal quadrant, fatigue, weight loss, and episodes of fever and chills, which are reported in a variable number of patients [77]. Symptoms of liver cirrhosis and portal hypertension with ascites and variceal hemorrhage are more rarely reported at diagnosis in PSC. Hepatomegaly and splenomegaly are the most frequent findings at clinical examination at the time of diagnosis in PSC. Osteopenic bone disease is a complication of advanced PSC, although less frequent than that reported in PBC. Fat malabsorption with steatorrhea and malabsorption of fat-soluble vitamins occur only with prolonged cholestasis.

Biochemical tests

Elevation of serum AP is the most common biochemical abnormality in PSC [[77], [78], [79]]. However, a normal AP activity should not preclude further steps to diagnose PSC if suspected on clinical basis. Serum aminotransferase levels are elevated at diagnosis in the majority of patients, typically to levels 2–3 times upper limits of normal, but normal levels are also observed. Serum bilirubin levels are normal at diagnosis in up to 70% of patients. Elevated levels of IgG have been noted in 61% of patients, most often to a level up to 1.5 times upper limit of normal [80]. In one retrospectively studied cohort, 9% of PSC patients were reported with slightly elevated IgG4 levels, but total IgG was not reported in these patients. It remains unclear whether some of these patients suffered from IgG4-associated cholangitis (IAC) rather than PSC [81]. Increased IgM levels have been reported in up to 45% of PSC cases [78].

Autoantibodies

A variety of autoantibodies have been detected in PSC [82]. The autoantibodies most frequently reported are perinuclear antineutrophil cytoplasmic antibodies (pANCA) (26–94%), antinuclear antibodies (ANA) (8–77%), and smooth muscle antibodies (SMA) (0–83%) [82]. The pANCA pattern in PSC is “atypical”, as the putative antigen is located in the nucleus rather than in the cytoplasm. Atypical pANCA is frequently present in UC and AIH, and specificity in the diagnosis of PSC is low. Positive titres of ANA and SMA also are unspecific. A routine autoantibody screening is not required to establish a diagnosis of PSC. Analysis of ANA and SMA may be relevant in a subgroup of patients to support a suspicion of “autoimmune” features that may have therapeutic implications (see “PSC–AIH overlap syndrome”).

Liver biopsy

Liver histological findings may support a diagnosis of PSC, but they are non-specific and may show considerable variation. PSC has been described to progress through four stages. The initial changes (stage 1, portal stage) are limited to the portal tracts with features including portal oedema, mild portal hepatitis, a non-destructive cholangitis with infiltration of lymphocytes in the bile ducts, and ductular proliferation. Periductal fibrosis and fibrous-obliterative cholangitis may be present. In stage 2 (periportal stage), the lesion extends to involve periportal fibrosis, sometimes with interphase hepatitis. Portal tracts are often enlarged. In stage 3 (septal stage) there is development of bridging fibrous septa, while bile ducts degenerate and disappear. Stage 4 is characterized by cirrhosis [83]. Periductal concentric fibrosis is considered highly suggestive of PSC, but this finding is relatively infrequent in needle biopsies in PSC and may also be associated with other conditions. Histological changes can be very subtle, and a liver biopsy may even appear normal because of sampling variability and since the liver is not uniformely involved. In PSC patients with relatively high serum aminotransferase levels, particularly in combination with positive ANA and/or SMA titres and markedly elevated IgG levels, a liver biopsy may be indicated to disclose features of a PSC–AIH overlap syndrome.

Imaging

Ultrasonography (US): In PSC, US is not diagnostic and often normal, but bile duct wall thickening and/or focal bile duct dilatations may be observed by experts. One or more gallbladder abnormalities, including wall thickening, gallbladder enlargement [84], gallstones, cholecystitis, and mass lesions, have been reported on the basis of US or cholangiography in up to 41% of PSC patients [85].

Cholangiography: A detailed cholangiographic assessment of the biliary tree is essential in making a diagnosis of PSC [86]. Efforts should be made to adequately visualize also the intrahepatic ducts to avoid false-negative results by overlooking subtle changes. The characteristic cholangiographic findings of PSC include mural irregularities and diffusely distributed multifocal, short, annular strictures alternating with normal or slightly dilated segments producing a “beaded” pattern [87]. Sometimes outpouchings have a diverticular appearance [87]. With more advanced disease, long, confluent strictures may be seen [87]. In the majority of cases, both the intra- and extrahepatic bile ducts are involved. A variable proportion of patients (<25%) is described to have isolated intrahepatic disease, whereas lesions confined to the extrahepatic ducts are rarely observed (usually <5%) and should only be diagnosed in the presence of adequate filling of the intrahepatic ducts. Since intrahepatic bile duct abnormalities can also be seen in other chronic liver diseases, one must be cautious when diagnosing PSC in the presence of intrahepatic changes only. The gallbladder and cystic duct are involved in some cases, and abnormalities of the pancreatic duct resembling those of chronic pancreatitis have been noted in a variable number of PSC patients [87].

Endoscopic retrograde cholangiopancreatography (ERCP) has been the gold standard in diagnosing PSC [[86], [87]], but ERCP is associated with complications such as pancreatitis and sepsis [88]. Clinicians may be reluctant to proceed with an ERCP in the assessment of cholestasis, and therefore, PSC most likely has been an underdiagnosed condition. Magnetic resonance cholangiopancreatography (MRCP) is a non-invasive method that in experienced centres is now generally accepted as a primary diagnostic modality in cases of suspected PSC. Studies comparing ERCP and MRCP have shown similar diagnostic accuracy, although the depiction of bile ducts may be poorer with MRCP than with ERCP [89]. Sensitivity and specificity of MRCP has been ⩾80% and ⩾87%, respectively, for the diagnosis of PSC [[89], [90]]. MRCP is superior in visualizing bile ducts proximal to duct obstructions. The method can also reveal changes within the bile duct walls and pathologies in the liver parenchyma as well as in other organs. However, cases with mild PSC changes without bile duct dilatation may be missed by MRCP and one should therefore be cautious to exclude early PSC on the basis of a normal MRCP. Thus, diagnostic ERCP still has a role in equivocal cases. The main role of ERCP, however, lies in therapeutic procedures and in diagnostic purposes like cytology sampling in PSC.

Small duct PSC

The term small duct PSC refers to a disease entity which is characterized by clinical, biochemical, and histological features compatible with PSC, but having a normal cholangiogram [91]. One report has restricted the diagnosis of small duct PSC to patients with concomitant IBD [92], whereas IBD has only been present in a proportion (50–88%) of cases in other studies [[93], [94]]. These studies carry the risk to include patients with other cholangiopathies such as ABCB4 deficiency which cause histological features compatible with small duct PSC [95]. A high-quality cholangiogram is mandatory in order to exclude PSC with isolated intrahepatic distribution. One future approach for the diagnosis of small duct PSC is to accept a negative MRC in patients with concomitant IBD, but require a normal ERCP and a negative mutation analysis of ABCB4 in patients without IBD. Diagnostic criteria in small duct PSC are however still being discussed.

PSC in children

Criteria for diagnosis of PSC in adults also apply to children. Of note, levels of serum AP activity were observed within the normal range for the age group in up to 47% of cases [[96], [97]]. Patients with normal AP usually had elevated γGT activity [[96], [97]]. Presentation of PSC in children is frequently reported with features similar to those of autoimmune hepatitis, including high IgG concentrations, positive ANA and/or SMA titers and interphase hepatitis in the liver biopsy [[96], [97], [98]].

Differential diagnosis of PSC versus secondary forms of sclerosing cholangitis

Before the diagnosis of PSC can be settled, causes of secondary sclerosing cholangitis such as previous biliary surgery, cholangiolithiasis and disorders mimicking PSC such as carcinoma of the bile ducts have to be excluded although cholangiolithiasis and cholangiocarcinoma may also be the consequence of PSC [76]. Clinical and cholangiographic findings resembling those of PSC have most commonly been described in relation to intraductal stone disease, surgical trauma from cholecystectomy, abdominal injury, intra-arterial chemotherapy, and recurrent pancreatitis [76]. A variety of other conditions have also been associated with features imitating those of PSC, including IgG4-associated cholangitis/autoimmune pancreatitis (see below), hepatic inflammatory pseudotumor, eosinophilic cholangitis, mast cell cholangiopathy, portal hypertensive biliopathy, AIDS cholangiopathy, recurrent pyogenic cholangitis, ischemic cholangitis, as well as others [76]. Differentiating between primary and secondary sclerosing cholangitis may be particularly difficult since PSC patients themselves may have undergone bile duct surgery or have concomitant intraductal stone disease or even cholangiocarcinoma (CCA). Factors like clinical history, the distribution of the cholangiographic abnormalities, as well as the presence of concomitant IBD, have to be taken into account when determining whether a pathological cholangiogram is due to PSC or secondary to a benign or malignant bile duct stricture without PSC [76].

Recommendations

  1. A diagnosis of PSC is made in patients with biochemical markers of cholestasis not otherwise explained, when MRCP shows typical findings and causes of secondary sclerosing cholangitis are excluded (II-2/B1). A liver biopsy is not essential for the diagnosis of PSC in these patients, but allows activity and staging of the disease to be assessed.
  2. A liver biopsy should be performed to diagnose small duct PSC if high-quality MRCP is normal, (III/C2). A liver biopsy may also be helpful in the presence of disproportionally elevated serum transaminases and/or serum IgG levels to identify additional or alternative processes (III/C1).
  3. ERCP can be considered
    • If high-quality MRCP is uncertain (III/C2): the diagnosis of PSC is made in the case of typical ERCP findings.
    • In patients with IBD with normal high-quality MRCP but high suspicion for PSC (III/C2).

Follow-up of PSC

Inflammatory bowel disease and risk of colon cancer

PSC is strongly associated with IBD, with a prevalence of IBD in Western countries commonly in the range of 60–80% [[77], [78]] whereas in a recent report on 391 Japanese patients only 125 had a history of concomitant IBD [99]. UC accounts for the majority (80%) of IBD cases in PSC, while around 10% have Crohn's disease and another 10% are classified as indeterminate colitis [100]. IBD can be diagnosed at any time during the course of PSC, but in a majority of cases IBD precedes PSC. Since the colitis in PSC characteristically is mild and sometimes even asymptomatic, colonoscopy with biopsies is recommended as part of the routine work-up in a patient diagnosed with PSC. A diagnosis of IBD has implications for follow-up and dysplasia/cancer surveillance as patients with UC and PSC have a higher risk of dysplasia and colon cancer than patients with UC only [[101], [102]]. Compared to UC patients without PSC, the colitis in PSC more frequently is a pancolitis (87% vs. 54%), with backwash ileitis (51% vs. 7%), and rectal sparing (52% vs. 6%) [100]. Patients with PSC and Crohn's disease characteristically only have colonic involvement. We recommend that PSC patients with colitis are enrolled in a surveillance program with annual colonoscopy with biopsies from the time of diagnosis of PSC [102].

Hepatobiliary malignancies in PSC

PSC is associated with an increased risk of hepatobiliary malignancies, in particular cholangiocarcinoma (CCA). In a large cohort of 604 Swedish PSC patients followed for (median) 5.7 years, hepatobiliary malignancies (CCA, hepatocellular carcinoma (HCC), and gallbladder carcinoma) were observed in 13.3%, corresponding to a risk 161 times that of the general population [103]. CCA is by far the most common hepatobiliary malignancy in PSC, with a cumulative life-time incidence of 10–15% [104], whereas gallbladder carcinoma [85] and HCC [105] are observed in up to 2% of PSC patients, each. Up to 50% of CCA are diagnosed within the first year of diagnosis of PSC. After the first year, the yearly incidence rate is 0.5–1.5% [104]. Although factors like older age, alcohol consumption and smoking, long duration of IBD before diagnosis of PSC, and a history of colorectal malignancy, have been associated with an increased risk of CCA in PSC, no clinically useful prognostic variables have been identified so far. Possible genetic markers should be further explored [75]. The symptoms of CCA complicating PSC may be very difficult to differentiate from those of PSC without concomitant malignancy, but awareness of CCA must in particular be raised in cases of rapid clinical deterioration.

Median levels of the serum tumour marker carbohydrate antigen 19-9 (CA 19-9) are significantly higher in PSC patients with CCA than in those without [104], but in the individual case CA 19-9 cannot be relied upon in the differential diagnosis between PSC with and without CCA [104]. Distinguishing benign from malignant changes in PSC by imaging modalities like US, CT, MRCP/MRI as well as ERCP, is equally difficult [[104], [106]]. Serum CA 19-9 combined with cross-sectional liver imaging may be useful as a screening strategy [107], but further validation is needed. Whether dynamic (18F)-fluoro-deoxy-d-glucose positron emission tomography (FDG-PET) [108] is more effective when combined with CT or MRI, needs to be shown. Brush cytology sampling, and biopsy when feasible, during ERCP adds to the diagnostic accuracy of CCA in PSC [[104], [107], [109]], but methodological refinement including validation of digital image analysis (DIA) and fluorescence in situ hybridization (FISH) of cell samples [107] is needed.

Gallbladder mass lesions in PSC frequently (>50%) represent adenocarcinomas independently of their size [85]. Cholecystectomy is recommended in PSC patients with a gallbladder mass even <1 cm in diameter [85]. The risk for pancreatic carcinoma was 14-fold increased in a Swedish cohort of PSC patients in comparison to a matched-control population [103], but its incidence in PSC is markedly lower than that of hepatobiliary malignancies, and regular screening strategies are, therefore, not recommended at present.

Recommendations

  1. Total colonoscopy with biopsies should be performed in patients in whom the diagnosis of PSC has been established without known IBD (III/C1) and should be repeated annually (or every 1–2 years in individualized patients) in PSC patients with colitis from the time of diagnosis of PSC (III/C1).
  2. Annual abdominal ultrasonography should be considered for gallbladder abnormalities (III/C2).
  3. There is at present no biochemical marker or imaging modality which can be recommended for early detection of cholangiocarcinoma. ERCP with brush cytology (and/or biopsy) sampling should be carried out when clinically indicated (III/C2).

Treatment of PSC

Ursodeoxycholic acid (UDCA)

UDCA and disease progression: UDCA is an effective treatment of primary biliary cirrhosis (PBC) as outlined above (2.2.1). UDCA has, therefore, also been investigated as a potential candidate for the treatment of PSC. Small pilot trials of UDCA in the early 1990's demonstrated biochemical and in some cases histological improvement in PSC patients using doses of 10–15 mg/kg/day [[110], [111], [112], [113]]. A more substantial trial was published by Lindor in 1997 [114], recruiting 105 patients in a double-blind placebo-controlled trial of 13–15 mg/kg of UDCA for 2 years. The results indicated improvement in serum liver tests but not in symptoms and, most importantly, no improvement in liver histology as evaluated by disease stage [114]. Higher doses of UDCA were then studied on the grounds that larger doses might be necessary to provide sufficient enrichment of the bile acid pool in the context of cholestasis, and that these doses might also enhance the potential immunomodulatory effect of the drug. Studies using 20–25 mg/kg/day demonstrated significant improvements in the histological grade of liver fibrosis and the cholangiographic appearances of PSC, as well as the expected biochemical improvement [115]. A shorter, open-label trial using 25–30 mg/kg/day showed a significant improvement in projected survival using the Mayo risk score, but no direct measurement of the progression of the disease, such as liver biopsy or cholangiography was undertaken. Confirmatory results were obtained in a 2-year dose ranging pilot study of 30 patients in which the low dose (10 mg/kg/d) and the standard dose (20 mg/kg/d) tended to improve and the high dose (30 mg/kg/d) significantly improved projected survival [116].

The Scandinavian UDCA trial deserves major credit for recruiting the largest group of PSC patients (n = 219) for the longest treatment period (5 years) ever studied using a dose of 17–23 mg/kg/day. It demonstrated a trend towards increased survival in the UDCA-treated group when compared with placebo [117]. But despite the relatively large number of patients recruited, it was still insufficiently powered to produce a statistically significant result. In comparison to other studies, the biochemical response was unexpectedly poor in this trial which prompted questions about adequate compliance in a part of the study population. Recently, a multicentre study using high doses of 28–30 mg/kg/d of UDCA in 150 PSC patients over 5 years has been aborted because of an enhanced risk in the UDCA treatment group to reach primary endpoints such as liver transplantation or development of varices in more advanced disease while biochemical features improved in the whole UDCA group [118]. Thus, the role for UDCA in slowing the progression of PSC-related liver disease is as yet unclear and high dose UDCA may be harmful in late-stage disease.

UDCA and chemoprevention: Recent work has suggested that UDCA may have a role in the prevention of colonic neoplasia in patients with PSC associated with underlying IBD. Experimental studies in vitro and in vivo had suggested that UDCA might prevent development of colon carcinoma. A cross-sectional study of 59 PSC patients with ulcerative colitis (UC) undergoing colonoscopic surveillance found a significantly reduced risk of colonic dysplasia in patients taking UDCA although in comparison to an exceptionally high rate of dysplasia in the control group [119]. A historical cohort study compared 28 PSC patients under UDCA treatment with UC to 92 PSC patients with UC not treated with UDCA [120] and found a trend towards a lower risk of colonic dysplasia and neoplasia under UDCA treatment (adjusted relative risk 0.59, 95% CI 0.26–1.36, p = 0.17) and a lower mortality (adjusted relative risk 0.44, 95% CI 0.22–0.90, p = 0.02) [120]. A third study followed 52 patients with PSC and UC for 355 patient-years who participated in a randomized, placebo-controlled UDCA trial showing a significant reduction to 0.26 (95% CI 0.06–0.92, p = 003) in UDCA-treated patients in the relative risk of developing colorectal dysplasia or carcinoma [121].

Limited evidence for a beneficial effect of UDCA on the risk to develop CCA comes from observational studies. The Scandinavian and American randomized, placebo-controlled UDCA trials with 219 and 150 PSC patients, respectively, did not observe a difference between UDCA- and placebo-treated patients regarding CCA development [117]. A German cohort study including 150 patients followed for a median of 6.4 years under UDCA treatment found CCA in 5 patients (3.3%), which represents about half the expected incidence of CCA in PSC [122]. A Scandinavian study of 255 PSC patients listed for liver transplantation over a period of 11 years revealed lack of ursodeoxycholic acid treatment as an independent risk factor for the development of hepatobiliary malignancy [123].

Immunosuppressive and other agents

Corticosteroids and other immunosuppressants have not demonstrated improvement in disease activity or outcome of PSC. Small randomized, placebo-controlled or pilot trials have investigated the role of agents with immunosuppressive potency like prednisolone, budesonide, azathioprine, cyclosporine, methotrexate, mycophenolate, and tacrolimus, agents with TNFα antagonizing effects like pentoxifyllin, etanercept and anti-TNF monocolonal antibodies and anti-fibrotic agents like colchicine, penicillamine, or pirfenidone. There is no evidence that these drugs are effective and, therefore, none can be recommended for classic PSC. These drugs may well have a role in the context of a PSC–AIH overlap syndrome (see below) since pediatric patients and those with evidence of a PSC–AIH overlap syndrome are more likely to respond to immunosuppressive treatment [[59], [60], [98]]. A retrospective study in adults also suggested a beneficial role of steroids in a subgroup with AIH overlap features [124].

ERCP and endoscopic therapy

Diagnostic ERCP has been the procedure of choice for suspected PSC in the past, but is associated with significant risks including pancreatitis and cholangitis [[125], [126]]. Whilst a low complication rate was found in patients undergoing 'diagnostic' ERCP, the complication rate increased up to 14% when interventions such as balloon dilatation, endoscopic sphincterotomy and stenting were performed [[4], [127]].

Dominant bile duct strictures have been defined as stenoses <1.5 mm in diameter in the common bile duct and <1 mm in the right and left hepatic duct [128]. The prevalence of dominant bile duct strictures in large duct PSC is variously reported as being 10–50%. Studies in animals and humans have suggested that decompression of biliary obstruction may prevent further damage and can reverse fibrotic liver disease [129]. Endoscopic treatment of biliary strictures often improves liver biochemistry and pruritus and may reduce the risk of recurrent cholangitis. Therefore, repeated endoscopic dilatation of dominant biliary strictures has been carried out in symptomatic patients [[130], [131], [132]]. Non-randomized studies comparing jaundice, cholangitis, transplantation and actuarial survival rates with estimates from prognostic models have suggested a trend towards a benefit of endoscopic intervention for dominant biliary strictures although patients also received UDCA [[131], [133]]. In contrast, a Swedish study which compared liver biochemistry in those with and without dominant strictures suggested that variations in cholestasis and jaundice are a feature of PSC liver disease and not related to dilatation of dominant strictures [128]. The optimum method and frequency of dilatation of dominant strictures is unclear. The most widely used technique to facilitate biliary drainage has been plastic stent insertion with or without prior dilatation. The problem with this approach is that further ERCP's are required to remove or replace the stent and there is a high rate of stent occlusion and/or cholangitis within 3 months of insertion. One study assessed the effectiveness and safety of short-term stenting (mean 9 days) resulting in improved outcome, particularly with regard to cholangitis and stent occlusion rates [134]. The strategy of short-term stenting for 2–3 weeks is followed by some experienced centers. Other studies have compared the role of stenting with balloon dilatation, with similar efficacy and lower rates of complications such as cholangitis (18% vs. 50%) associated with balloon dilatation alone [135]. Multiple dilatations are usually required over months or years in order to maintain patency once dominant strictures are identified and treated, and not all strictures are amenable to endoscopic intervention. In these patients, careful consideration should be made regarding a conservative, radiological or surgical (including liver transplantation) approach to treatment.

Liver transplantation

Liver transplantation is the only therapy of late-stage PSC that can cure advanced disease. One and ten-year survival after liver transplantation has lately been above 90% and 80%, respectively, in experienced centers. Resection of the extrahepatic biliary tree and Roux-en Y choledochojejunostomy are widely regarded as the method of choice for biliary reconstruction after liver transplantation in PSC [136]. Recurrence of PSC after liver transplantation has been reported at various rates up to a third of patients transplanted, but is difficult to define due to similarities in bile duct damage with ischemic type biliary lesions, infections, medication-induced injury, preservation injury, or chronic rejection [137]. In different cohorts, PSC recurrence was associated with steroid-resistant rejection, OKT3 use, preservation injury, ABO incompatibility, cytomegalovirus infection, male sex, or donor-recipient gender mismatch [138]. Colectomy prior to liver transplantation for advanced colitis or colon dysplasia protected against PSC recurrence as did the absence of ulcerative colitis [139].

Recommendations

  1. The available data base shows that UDCA (15–20 mg/d) improves serum liver tests and surrogate markers of prognosis (I/B1), but does not reveal a proven benefit on survival (III/C2). The limited data base does not yet allow a specific recommendation for the general use of UDCA in PSC.
  2. Currently there is suggestive but limited evidence for the use of UDCA for chemoprevention of colorectal cancer in PSC (II-2/C2). UDCA may be particularly considered in high-risk groups such as those with a strong family history of colorectal cancer, previous colorectal neoplasia or longstanding extensive colitis (III/C2).
  3. Corticosteroids and other immunosuppressants are not indicated for treatment of PSC in adults unless there is evidence of an overlap syndrome (III/C2).
  4. Dominant bile duct strictures with significant cholestasis should be treated with biliary dilatation (II-2/B1). Biliary stent insertion should be reserved for cases where stricture dilatation and biliary drainage are unsatisfactory (III/C2). Prophylactic antibiotic coverage is recommended in this setting (III/C1).
  5. Liver transplantation is recommended in patients with late-stage PSC (II-2/A1) and may be considered in patients with evidence of cholangiocyte dysplasia or severe recurrent bacterial cholangitis (III/C2).