Primary sclerosing cholangitis: review for radiologists

Abstract

Primary sclerosing cholangitis is a rare chronic inflammatory disease affecting the bile ducts, which can eventually result in bile duct strictures, cholestasis and cirrhosis. Patients are often asymptomatic but may present with clinical features of cholestasis. Imaging plays an important role in the diagnosis and management. This review covers the pathophysiology, clinical features, imaging findings as well as methods of surveillance and post-transplant appearance.

Graphical Abstract

Introduction

Primary sclerosing cholangitis (PSC) is a rare disease in which progressive inflammation and stricturing of the intra- and extrahepatic bile ducts leads to cholestasis and liver deterioration. In addition to liver failure, the chronic inflammation associated with PSC also predisposes to an increased risk of cholangiocarcinoma (CCA). Imaging, in particular MRI, plays an important role in managing these patients and although surveillance strategies for CCA continue to evolve, it seems likely that a combination of non-invasive imaging using magnetic resonance imaging (MRI) and magnetic resonance cholangiopancreatography (MRCP), biomarkers, and strategic endoscopic retrograde cholangiography (ERC) will likely become the surveillance strategy of choice. In this review, we discuss current thinking about the pathophysiology of the disease and imaging findings compatible with the disease, with attention to variants and overlap syndromes. We then discuss current strategies for surveillance for CCA with potential future developments.

Pathophysiology and epidemiology of primary sclerosing cholangitis

PSC is considered an autoimmune disease and, similar to other autoimmune diseases, it is thought to arise out of a combination of genetic predisposition and environmental triggers [1]. The precise mechanism that initiates PSC is not fully known, however. The disease’s association with the gastrointestinal tract and inflammatory bowel disease (IBD) suggests that it originates from a biliary-gastrointestinal axis. Alteration of the colonic microbiome, a particular microbial exposure (“leaky gut” with increased bacterial translocation), and/or activation of immune cells in the colon, or some combination of these, have been hypothesized to trigger the biliary inflammation [2].

After the initial environmental trigger, toxic cycles of worsening biliary injury are thought to occur with T-cell release of profibrogenic cytokines and destruction of an intra- and extracellular bicarbonate buffer released by cholangiocytes to protect them from the injury induced by bile salts [3, 4]. Superinfection in the obstructed bile ducts may also play an important role in disease progression [5]. As the fibrosis, collagen deposition and generation of strictures worsens, the potential for further injury increases. Eventually concentric, obliterative fibrosis of the bile ducts occurs that extends into the liver parenchyma, resulting in “biliary” cirrhosis.

Unlike many other autoimmune diseases, PSC is reported to have a male predilection (2:1) and it has a bimodal mean age of presentation of 15 and 35 years old [6]. The bimodal distribution suggests different subpopulations of the disease, but this has yet to be determined. Overall, PSC is a rare disease, with a pooled incidence ratio of 0.6 per 100,000 patient years [7]. It is thought that there may be significant regional variations in the incidence of the disease, with a higher incidence in Northern European populations, but limited incidence data from outside of North America and Europe limits confidence in this assessment [7].

PSC, IBD, and other associations

One of the most well-known associations with PSC is IBD (up to 80%, although there is regional variation in the incidence [7]). This correlation has suggested a causal role for the bowel in initiating the liver disease. The classic association is with ulcerative colitis (UC) and a smaller percentage are diagnosed with Crohn disease. More recent genome-wide association studies, however, suggest that PSC is less associated with UC and Crohn disease than with a separate entity, “PSC-IBD” [1, 8]. Patients with PSC-IBD have a different microbiome compared with patients with UC without PSC [9].

Patients with PSC-IBD are considered high risk for colorectal cancer (10 times above the general population) [10] and colonoscopy screening every 1–2 years is recommended by most gastrointestinal societies [11]. The risk and surveillance strategy is similar to patients with UC or Crohn colitis who have more than a third of the colon affected, although some societal recommendations recommend an even more vigilant annual surveillance schedule for PSC-IBD [11].

PSC is associated with other autoimmune diseases, including thyroiditis, diabetes mellitus type 1, and rheumatoid arthritis. There is also regional variation in the associations, with the associations more common in northern Europe [12].

Clinical presentation and differential

Although up to 40% of patients with PSC are asymptomatic at the time of diagnosis [13, 14], many present with signs and symptoms of a cholestatic liver disease. Symptoms include right upper quadrant pain and fatigue, with more advanced cases presenting with pruritis. Physical examination is abnormal in nearly half of more advanced cases, with hepatosplenomegaly and jaundice more common findings. Laboratory evaluation may be normal in early disease, but as the disease progresses will eventually show a cholestatic pattern with increased alkaline phosphatase (most common) and elevated bilirubin levels.

Although liver function tests can be useful to initiate a workup, there are no specific lab markers that help to make a diagnosis of PSC. Imaging, therefore, has an important role in making the diagnosis of a sclerosing cholangitis. If characteristic imaging findings are found, accurate clinical correlation is essential to exclude secondary causes of sclerosing cholangitis (Table 1) [15, 16]. If no cause can be found, then the sclerosing cholangitis is considered a “primary” sclerosing cholangitis [17]. Liver biopsy is not typically required, except when there is a suspicion for small duct PSC or an overlap syndrome.

Table 1.

Secondary sclerosing cholangitides

IgG4 cholangitis

Recurrent pyogenic cholangitis

Portal cholangiopathy

Eosinophillic and/or mast cell cholangitis

Hepatic inflammatory pseudotumor

Primary immune deficiency

AIDS-related cholangiopathy

Variants and overlap syndromes

Two PSC variants deserve mention as possible diagnostic confounders when evaluating a patient.

The first is “small duct PSC” (SD-PSC). This is a variant in which the patient has typical findings of PSC on liver biopsy, but since it involves 4th or higher order bile ducts (< 400 nm diameter) it is below the resolution of both MRCP and ERCP. Small duct PSC by itself has a better prognosis than classic “large duct” PSC and it is thought that cholangiocarcinoma does not occur in small duct PSC [18] and annual surveillance is not currently recommended by the American Gastroenterological Association (AGA) [19]. If the disease progresses, it does so by first converting to large duct PSC, which can occur in 33–55% of cases [20, 21]. SD-PSC with IBD is more similar to large duct PSC in its HLA association than small duct PSC without IBD [22]. Small duct PSC has a relatively nonspecific appearance on imaging, but periductal enhancement, heterogeneous parenchymal signal intensity, inhomogeneous liver enhancement, periportal lymphadenopathy, and gallbladder dilatation have been described [20, 21]. MRI/MRCP studies in SD-PSC are often reported as normal and a high degree of clinical suspicion leads to a liver biopsy confirming the diagnosis.

The second potential imaging and clinical confounder are “overlap syndromes,” in particular an overlap with autoimmune hepatitis (PSC-AIH) that shows evidence of both diseases on histology. The HLA haplotype of B8-DR3 is associated with both PSC and AIH [23] and it is thought to be more prevalent in young patients [24]. Estimating the prevalence of overlap is limited by variation in the definition, and estimates range widely between 1 and 54% [25]. Unlike typical PSC, PSC-AIH responds to immunosuppressive therapy and has a better prognosis (although it has a worse prognosis than AIH alone) [26]. The role of imaging in these patients is to raise the possibility of an overlap syndrome if the patient has typical PSC finding on MRCP, but has a clinical and serologic pattern compatible with AIH [27].

The role of imaging—initial detection

Classic imaging findings in PSC include multifocal stricturing of the bile ducts with intervening segments that are relatively normal in caliber or mildly dilated. This results in a “beaded” appearance of the bile ducts on cholangiography. Generally, stricturing in PSC involves both the intrahepatic and extrahepatic bile ducts, though a minority of patients have disease that is confined to the intrahepatic (15–25%) or extrahepatic ducts (5–10%) [28, 29]. Other parts of the biliary system may be affected, including the gallbladder and cystic duct. Pancreatic duct involvement has even been reported [30]. Concentric periductal soft tissue thickening is a common finding in PSC and likely indicates the presence of active cholangitis (Fig. 1).

Fig. 1.

Biliary dilatation and enhancement in PSC. a MRCP showing beaded dilatation of bile ducts in the left hepatic lobe (arrow). b Axial post contrast image demonstrating stricture and bile duct dilatation with wall thickening and intense wall enhancement. c MRCP demonstrates a dilated duct (arrow) in the right hepatic lobe ducts. d Coronal post contrast image, the corresponding duct demonstrates wall thickening and hyperenhancement

Early PSC may be relatively subtle with a few minimally dilated ducts at the periphery of the liver. In more advanced cases, more severe strictures are present. The term “dominant stricture” (DS) is frequently used to describe the most focal severe stricture, but its use is controversial and the International PSC Study Group (IPSCSG) advocates avoiding the term when using MRCP alone [31]. Importantly, there is an overlap between the imaging appearance of a new severe stricture and cholangiocarcinoma and tissue sampling is usually required with ERCP and brush cytology or EUS-guided sampling.

Although patients with PSC may be initially evaluated with ultrasound or CT, cholangiographic evaluation with MRCP or ERCP is generally required to establish the diagnosis with certainty. Advanced cases of PSC may have findings evident on ultrasound such as wall thickening of the central intrahepatic or extrahepatic duct, echogenic portal triads, or areas of segmental biliary duct dilation, though ultrasound may also be normal early in the course of the disease (Fig. 2) [32]. CT may also show biliary duct wall thickening and enhancement with segmental biliary duct dilation, although the strictures themselves are often not well appreciated (Fig. 3) [33].

Fig. 2.

Examples of ultrasound findings in PSC. a In this image, there is biliary ductal dilatation (arrow), but as can be seen in b, there are other areas of the intrahepatic ducts that are not dilated and have echogenic walls (arrow), compatible with a focal stricture. In c the normally anechoic common bile duct has low level echoes (arrow), corresponding to diffuse wall thickening of the duct. Color Doppler evaluation d shows a recanalized umbilical vein, compatible with underlying portal hypertension from cirrhosis

Fig. 3.

Appearance of PSC on CT. a–c There is irregular dilatation of the central bile ducts with intervening strictures (example at ovals) that would produce a “beaded” appearance on cholangiography. There is also wall thickening and enhancement of the ducts, compatible with active cholangitis (most noticeable in c at the arrow). A wider field of field of view of the liver shows some peripheral ductal dilatation (d). Splenomegaly can also be seen

MRI with MRCP is considered the noninvasive imaging method of choice, and it shows the multifocal strictures and segmental dilatation to better advantage than CT or ultrasound. Indeed, early cases of PSC may have findings that are only apparent on 3D MRCP maximum intensity projection (MIP) images (Fig. 4). Unlike ERCP, MRCP avoids contamination of the biliary tree and can evaluate the entirely of the biliary system upstream from a severe stricture, which may not be opacified on ERCP. However, ERCP is often needed and is complementary to MRCP, as it allows for cytologic sampling or stenting of a severe stricture.

Fig. 4.

A 42-year-old male with PSC. a An MRCP MIP image shows multifocal segmental strictures with mild dilatation involving both the intrahepatic and extrahepatic bile ducts. The liver parenchyma shows no signal abnormality on b T2W, c DWI (b = 600), and d pre-contrast T1W, and no abnormal parenchymal enhancement in e the arterial phase, f portal venous phase and g 5-min delayed phase. h MR elastography (MRE) shows normal liver stiffness

On MRI, biliary ductal wall thickening and enhancement is compatible with ongoing inflammation and/or fibrosis [34, 35]. The liver parenchyma often shows heterogeneous signal intensities on T2W, T1W and DWI sequences (Figs. 4 and 5). These heterogeneous signal intensities are usually distributed in a patchy fashion, often located at the periphery of the liver. Areas of increased parenchymal T2W hyperintensity also likely represent inflammation and/or fibrosis. Heterogeneous postcontrast enhancement of the liver parenchyma with early arterial phase enhancement and delayed enhancement (fibrosis) usually correspond to the locations of T2W and/or DWI signal abnormality [34]. These regions represent regions of inflammation and/or fibrosis and correspond to regions of increased stiffness on MR Elastography (MRE).

Fig. 5.

A 50-year-old male with PSC. MRI demonstrates heterogeneous patchy signal abnormalities in the liver parenchyma. a MRCP shows multifocal strictures and “pruning” of the peripheral bile ducts (oval and circle). b Axial fat-suppressed T2W image with corresponding c DWI (b = 600), d pre-contrast T1W, e post contrast enhanced arterial phase, f portal venous phase, g 5-min delayed phase images and h MR elastography (MRE). Note patchy peripheral hyperintensities (arrows) on the T2W images with corresponding hyperintensities on DWI, and hyperenhancement in the arterial and delayed phases. These regions also show increased stiffness on MRE

Focal, eccentric and/or nodular thickening of a bile duct should raise a suspicion for developing CCA. Chronic biliary stasis may lead to formation of biliary stones (hepatolithiasis) which are usually seen as T1 bright foci within the dilated ducts.

Macroregenerative nodules (MRN) may also be observed. These tend to develop in the central liver parenchyma (Fig. 6) and appear hypointense on T2W relative to inflamed and/or fibrotic liver parenchyma. They do not show significant restricted diffusion or any post-contrast hyperenhancement relative to background liver. Vessels course through these regions without any displacement and the regions show normal or mildly elevated liver stiffness on MRE confirming the regenerative normal parenchyma. In advanced PSC, fibrotic changes also occur in the nodules.

Fig. 6.

Macronodular regeneration. A 43-year-old male with PSC on follow-up. MRI images showing a central T2 hypointense region (arrows, a) that is iso- to hypointense on DWI (b = 600) (b), mildly hypointense on pre-contrast T1W (c), hypoenhancing in the arterial phase (d), portal venous phase (e), and delayed phase (f). Note the enhancing portal vein branch (black arrow) within the region. MRN shows lower stiffness compared to fibrotic parenchyma in the periphery on MRE (g). Follow-up MRI after 5 years showing stable MRN (arrow, h)

As PSC progresses, fibrosis and obliteration of more peripheral ducts leads to a “pruned” appearance of the biliary tree on cholangiography [29]. The extent of peribiliary fibrosis, which limits the ability of the ducts to dilate, explains why there is a relative lack of biliary ductal dilation proximal to even high-grade strictures. Chronic PSC eventually leads to a biliary cirrhosis and characteristic morphologic changes in the liver including a “rounded” or “box-like” appearance from caudate hypertrophy, left lateral section hypertrophy, and right hemiliver atrophy. However, the appearance of liver morphology in long-standing PSC is variable with variable atrophy and compensatory hypertrophy of different liver segments.

The appearance of the gallbladder in the diagnosis of PSC is not well understood, although a distended gallbladder has been reported to be associated with PSC [36]. Gallbladder polyps in a setting of PSC or potential PSC are at higher risk for adenocarcinoma and should be viewed with a high degree of suspicion.

The role of imaging—surveillance

Surveillance of a patient with PSC is an important part of their care. It mostly involves monitoring for the development of two outcomes: cirrhosis and cancer (cholangiocarcinoma and gallbladder cancer). The mortality rate median is at 15–20 years, with the majority of the deaths due to cancer (40–50%) but with liver failure almost as common [37].

The course of the disease is progressive, but there is wide variation in the rate of progression among individuals. Clinically, monitoring alkaline phosphatase is thought to be helpful and a level that remains only mildly elevated (< 1.5 × normal) is considered reassuring [38].

Cholangiocarcinoma

The annual incidence of cholangiocarcinoma (CCA) is 1.2 per 100,000 (0.0012%) in the US general population [39]. In contrast, the annual incidence of CCA in PSC is 0.6–1.5%, a several 100-fold increase [10, 40]. The cumulative incidence of CCA is 6–11% after 10 years and 20% after 30 years [19, 41, 42]. Although cumulative incidence increases over time, there does not appear to be a direct relationship between the duration of the disease and the occurrence of cholangiocarcinoma. In one study, the incidence of CCA was similar across three time periods: 26.8% of CCA occurred within the first year after diagnosis of PSC, 36.2% occurred between 1 and 10 years after PSC diagnosis, and 36.2% developed > 10 years after PSC diagnosis [41].

Risk factors for CCA in PSC are older age at PSC diagnosis and IBD subtype (UC higher risk than Crohn disease). The incidence of CCA is nearly 20 times higher in those who are 60 years or older compared to those under the age of 20 (21 vs 1.2 per 100 person-years, respectively).

Cholangiocarcinoma is a heterogeneous disease and is classified into three main types: intrahepatic CCA (iCCA), perihilar CCA (pCCA), and distal CCA (dCCA). Typically, cancers arising from the GB and ampulla of Vater are not considered cholangiocarcinoma. Most studies that record the incidence of CCA in PSC do not state the relative proportions of CCA subtypes [43]. As a result, frequency of occurrence of subtypes in prior studies is not very helpful as a diagnostic aid.

Surveillance for CCA in PSC

Although prospective data is absent, a single-center retrospective study of 830 PSC patients, 51% of whom underwent surveillance for hepatobiliary cancers, suggested improvement in survival with surveillance strategies in PSC. The 5-year survival rate was 68% in those who underwent surveillance vs 20% in those with no surveillance [44]. Annual screening with imaging for adult patients is recommended by major liver societies (American Association for the Study of Liver Diseases (AASLD), European Association for the Study of the Liver (EASL), AGA, IPSCSG) [19, 45-47].

In a PSC patient with no concerning signs or symptoms of CCA, surveillance consists of US or MRI/MRCP along with CA 19-9 every 6 to 12 months. MRI/MRCP has a higher sensitivity and specificity compared to US [48]. If a “dominant stricture,” mass, or increasing CA 19-9 levels are encountered during surveillance, a directed evaluation for CCA is initiated. [19, 49].

Directed evaluation for CCA involves ERCP, cholangioscopy, or endoscopic ultrasound with tissue sampling of suspicious areas (“dominant stricture” or mass). A high-grade stenosis involving the major ducts (common duct < 1.5 mm diameter; or right or left hepatic ducts within 2 cm of confluence < 1.0 mm diameter) is generally considered a severe or “dominant” stricture [49, 50] and associated with a higher risk of CCA [51].

Although MRI/MRCP is the best noninvasive imaging modality for detecting developing CCA in PSC, it can still be a challenge and high-quality MRCP and reader experience are important to improve accuracy. The MRI/MRCP should ideally be reviewed side-by-side with prior imaging, looking for any changes in the known stricture (severity, associated thickness), development of new stricture(s), and for any associated vascular changes.

Recently Eaton et al. proposed two imaging categories for diagnosis of CCA in PSC: Definite CCA -typical imaging features of CCA; and possible CCA- not definitive features but should evaluate further for confirmation [52].

According to this classification system, MRI features of definite/typical CCA include: (i) a perihilar mass (with or without intrahepatic or intraductal mass lesions) with progressive enhancement on delayed phase imaging or (ii) a malignant stricture (periductal focal soft tissue thickening with vascular narrowing or encasement, or periductal focal soft tissue thickening that enhances on delayed phase imaging) (Fig. 7) [52].

Fig. 7.

Extrahepatic cholangiocarcinoma presenting as a new “dominant” stricture in a 49-year-old male with ulcerative colitis and long-standing primary sclerosing cholangitis. 3D MRCP MIP image a shows a severe stricture involving the mid to distal common bile duct (arrows) which was new from the prior study. Upstream biliary duct dilation and mild multifocal stricturing was evident in the intrahepatic ducts (arrowheads). Coronal T1-weighted image with IV contrast (b) and coronal thin slice T2-weighted image with fat saturation (c) show circumferential enhancing soft tissue (oval in b) with a shouldered appearance (arrow in c) corresponding to the severes stricture on MRCP. ERCP biopsy revealed adenocarcinoma

MRI features of possible CCA are (i) stricture with nonfocal periductal thickening but without delayed phase enhancement or vascular narrowing (or encasement) or (ii) progressive lobar atrophy secondary to worsening perihilar stricture or (iii) ill-defined, irregular delayed enhancement of ductal wall without distinct mass [52].

The “definite” criteria had an accuracy of only 0.78 with low sensitivity of 58% but a high positive predictive value of 96%. However, when both “definite” and “possible” criteria were combined, the accuracy improved to 0.87 with higher sensitivity of 90% but a drop in positive predictive value to 86%. MRI was particularly useful in detecting CCA in asymptomatic PSC patients.

Gallbladder carcinoma

The lifetime risk for gallbladder carcinoma is lower than for cholangiocarcinoma, accounting for 2%, but with a risk of malignancy as high as 50% in a gallbladder polyp [53]. The sequence of inflammation, dysplasia, and carcinoma similar to UC is seen in the gallbladder epithelium [54]. The result of the sequence can appear as eccentric wall thickening, polyp, or a mass accounting for 20–30%, 12–25%, and 45–60% respectively of the imaging findings for gallbladder carcinoma [55]. Ultrasound is the imaging modality of choice for the initial diagnosis and follow-up of gallbladder pathology. Data supporting the effectiveness of MRCP in detecting polyps in these patients is sparse and needs further evaluation, so often these patients receive ultrasound follow up for the gallbladder in addition to a surveillance MRI/MRCP. On ultrasound, a polyp is seen as echogenic soft tissue with well-defined margins protruding from the gallbladder wall, often with a small stalk attached. On color Doppler, a stalk may show linear blood flow. Masses tend to have mixed echogenicity with irregular margins and nonlinear vascular core on color Doppler [56]. Wall thickening and enhancement can be detected on MRI as well (Fig. 8).

Fig. 8.

Gallbladder adenocarcinoma in a 57-year-old woman with newly diagnosed PSC who presented with jaundice. Axial (a) and coronal (b) contrast-enhanced CT images shows an intraluminal polypoid lesion in the GB (arrows) with associated soft tissue thickening involving the gallbladder neck (arrowhead), cystic duct, and common bile duct (not shown). ERCP (c) showed mild diffuse beading and structuring of the intrahepatic bile ducts, compatible with PSC. Percutaneous and endoscopic biopsies were nondiagnostic, though subsequent operative biopsy confirmed the presence of adenocarcinoma

The risk of cancer development is associated with the size of the polyp but the management of gallbladder polyps in the general population is controversial. Various professional organizations have their recommendation criteria for annual gallbladder surveillance. However, all agree that PSC is a high-risk underlying condition, and intervention or shorter interval follow-up is recommended. Both AASLD and EASL recommend annual gallbladder surveillance with ultrasound to identify the development of cancer at an early stage and cholecystectomy for all gallbladder lesions [45, 56]. The American College of Radiology recommends evaluation and follow-up for the abovementioned findings depending on the mass size and additional clinical factors with surgical consult recommended for a polyp ≥ 10 mm [57]. The recent joint guidelines between European societies recommend cholecystectomy for polypoid lesions of 6–9 mm and the American College of Gastroenterology advises cholecystectomy for polyps 8 mm in size [19, 58]. The most recent Society of Radiologists in Ultrasound (SRU) gallbladder polyp consensus conference recommendations acknowledges the increased risk in PSC patients and defers polyp follow-up recommendations and management to specialty guidelines [59].

Even though the risk for developing gallbladder cancer in PSC is low, the prognosis is poor with a survival of 5–10% [60]. A multidisciplinary approach should be considered for the management of gallbladder findings in high-risk patients given the varied recommendations by different societal guidelines. If the surgery is not performed, follow-up imaging is recommended [61].

Cirrhosis and hepatocellular carcinoma

Other than development of malignancies, progression of liver fibrosis is the most important prognostic factor for liver-related outcomes in PSC. The liver fibrosis stage is also predictive of transplant-free survival and the time to liver transplant [62]. It is important to note that the degree of parenchymal fibrosis does not always correlate with number or severity of biliary strictures and/or dilatation.

Liver biopsy is the gold standard for evaluation of fibrosis, but it is invasive, therefore surrogate markers are very useful. This is particularly true for a chronic and slowly progressive disease like PSC that requires long term surveillance. Elastography techniques measures liver stiffness (LS) and these have been shown to have excellent correlation with liver fibrosis stage acquired with a percutaneous biopsy [63, 64]. Ultrasound elastography plays an important role in non-invasively monitoring for fibrosis; liver stiffness measurements by transient elastography above 9.5 kPa are compatible with advanced fibrosis [46]. Similarly, LS evaluation with MR elastography can help distinguish mild from severe liver fibrosis at the time of diagnosis. Because the fibrosis that occurs in the liver from PSC is heterogeneous, the larger area evaluated in MRE is better for getting a sense of the true fibrosis burden than with the narrower field of view of ultrasound-based techniques.

MR elastography has a high specificity for detecting cirrhosis and the LS value is predictive of liver function deterioration in PSC. An LS of 4.93 kPa was the optimal point to detect cirrhosis with 100% sensitivity and 94% specificity [63]. Low, medium and high risk for hepatic decompensation were < 4.5 kPa, 4.5–6.0 kPa, and > 6.0 kPa (respectively) [63]. The trend of LS values is also important and the rate of progression of LS is strongly associated with clinical outcome in PSC [64].

A strong association between number/severity of strictures or dilatation has not been shown with liver stiffness measurements. MRE-determined segmental LS has been shown to be associated with segmental bile duct strictures and parenchymal signal abnormalities, however no strong association was found with liver stiffness and extrahepatic strictures, particularly common bile duct strictures, suggesting that segmental strictures are more likely to be associated with parenchymal fibrosis [33]. Sequential MRE studies are useful in showing progression of liver fibrosis (Fig. 9). These can be performed with MRCP studies for assessment of strictures and for screening for CCA.

Fig. 9.

A 41-year-old male with PSC. MRCP images (a–d) and corresponding MRE images (e–h) at initial presentation and during follow-up at 2 years, 4 years, and 5 years. The initial liver stiffness was normal at 2.4 kPa and there were mild segmental strictures throughout the biliary tree on MRCP. During follow up there was progression of strictures with associated intrahepatic biliary dilatation. The liver stiffness also showed progression increasing to 4.1 kPa at 2 years, 5.2 kPa at 3 years and 6 kPa at 5 years

Similarly, biliary dilatation or its severity alone does not strongly correlate with parenchymal fibrosis. Magnetic resonance risk score such as the Anali score which includes features such as dilatation of intrahepatic ducts, dysmorphic appearance and portal hypertension, has been shown to be associated with prognostic outcomes in PSC [65, 66]. Liver function can also be qualitatively assessed on MRI with hepatobiliary contrast agents, such as gadoxetate. Heterogeneous parenchymal enhancement correlates with biliary obstruction. Further studies are needed for assessing a specific role of gadoxetate-enhanced MRI studies in the evaluation of parenchyma disease and prognostic outcomes.

When PSC develops into cirrhosis, it is generally considered appropriate to monitor these patients for the development of hepatocellular carcinoma with ultrasound screening. This is complementary to annual MRI screening for cholangiocarcinoma. PSC is not felt to be an extra risk factor for the development of HCC over and above cirrhosis.

Liver transplantation

Because PSC is a chronic progressive disease without a definitive medical therapy, liver transplant (LT) remains the only definitive treatment shown to prolong survival [10]. Approximately 15% of PSC patients will require LT, either for liver failure or malignancy. The indications for LT in patients with PSC are similar to other liver diseases and qualifying MELD scores (for those with cirrhosis). The revised Mayo risk score is the most widely used for PSC prognosis [67].

Pre-transplant imaging

MR imaging of the donor and recipient prior to transplant is valuable for assessment of arterial and vascular anatomy, volume measurements, and differentiating benign lesions from any malignant pathologies that would preclude transplantation. Additionally, imaging markers have been explored for predicting LT-free survival since prognostic indices such as Child–Pugh score and MELD scores have limitations in PSC patients due to fluctuations of biomarkers. Liver fibrosis and stiffness can predict disease severity as mentioned. Liver volumes, including right and left hepatic lobe ratios, and spleen volumes have been explored for predicting outcomes [68]. A more recent study suggests the presence of arterial peribiliary hyperenhancement in PSC patients on MRI is associated with higher Mayo risk scores and may suggest a poorer prognosis [35]. For those with underlying IBD, pre-operative remission of colitis is recommended since active disease at transplant has been reported to predict subsequent graft failure [69]. However, the role of MR or CT enterography for pre-operative evaluation has not been fully explored.

Post-transplant recurrence

Over a 10-year period, recurrence of PSC (rPSC) can occur in 20–25% of cases leading to the need for re-transplantation, significant morbidity, and increased mortality risk. Both live-donor LTs (LDLT) and deceased-donor LTs (DDLT) are performed. Retrospective studies have been performed on whether the risk in LDLTs compared to DDLTs vary, especially considering that genotype/haplotype-matching in related donors may be associated with increased risk of rPSC. Although there has been a general trend for increased risk of rPSC in LDLTs, findings are not statistically significant. For example, a single-center experience demonstrated rPSC in 16% of DDLTs and 28% of LDLTs at approximately 5-year follow-up but differences were insignificant. However, risk of rPSC increased to 37% for LDLTs for biologically related donors. A more recent analysis demonstrated a 21.1% of rPSC at 10 years for LDLT; no differences between LDLT (regardless of relationship) and DDLT recipients [70]. Overall graft and patient survival have been shown to be similar between the LDLT and DDLT groups [70, 71].

Although extensively explored, no consistent single risk factor has been reported for rPSC. However, several studies have found active IBD is a risk factor for recurrent PSC and led to the suggestion that a colectomy could be protective [72, 73]. Just as in the diagnosis of PSC, the diagnosis of rPSC is one of exclusion and should be made in combination of biochemical, imaging, and occasionally histological findings. In the presence of radiological features that suggest rPSC, other causes of biliary pathology must be excluded, including those more commonly associated with LT such as acute rejection, hepatic artery stenosis and subsequent ischemic cholangitis, anastomotic strictures, and non-anastomotic strictures. Recurrent PSC presents as multifocal strictures and dilatations of the biliary tree as seen in cholangiography (Fig. 10). MRI/MRCP remains the first choice of imaging modality for assessing ducts after transplant due to the non-invasiveness. However, ERCP and PTC can be performed.

Fig. 10.

63-year-old female with history ulcerative colitis and PSC cirrhosis. The patient is status post living donor liver transplant and Roux-en-Y hepaticojejunostomy in 2005. a Axial T2 weighted images after transplant in 2005. b Axial T2 weighted images in 2022 demonstrate ductal dilatation in the posterior right hepatic lobe with segmental atrophy (arrow). c Coronal 3D MRCP maximum intensity projection images in 2022 demonstrate a severe stricture of the common bile duct at the hepaticojejunostomy anastomosis (white arrow). Beaded appearance of recurrent PSC in the upstream ducts can also be seen

Summary

Primary sclerosing cholangitis is an uncommon disease, but because of the importance of imaging in its diagnosis and management, particularly MRI/MRCP, it is a disease that commonly finds its way to an abdominal radiologist’s worklist. The heterogeneity of the process and the subtlety of some of the adverse outcomes, in particular the development of cholangiocarcinoma within a stricture, can be challenging and high-quality MRI/MRCP imaging is necessary for accuracy. Monitoring the progressive fibrosis of the disease is also important for management and elastographic techniques have been shown to be useful for this. Further work remains to be done regarding optimal multidisciplinary surveillance strategies of the liver and gallbladder, as well as the role of hepatobiliary-specific contrast agents.