Primary sclerosing cholangitis is a rare disease with a higher prevalence in Northern countries (e.g. Sweden, Norway) compared with southern Europe and Asia. Disease onset of PSC varies from childhood age below 10 up to elderly patients in the 7th or 8th decade of life; however, PSC mostly affects young males at the age of 30 to 40. In northern Europe, the incidence of PSC ranges from 0.9 to 1.3 per 100.000/year with a prevalence from 8.5 to 14.2 per 100.000. The highest prevalence (16.2/100.000) has been reported within a large population-based study in western Sweden with increasing trends for inflammatory bowel disease (IBD)-associated PSC, large duct PSC among men/women, and PSC without IBD and predominately small-duct PSC among males. Thus, PSC is frequently referred to as 'the disease of the North'. Interestingly, the reasons for the observed pronounced geographical differences in incidence and prevalence of PSC are still undefined but presumably include both genetic and environmental factors. As PSC does not have a distinct International Classification of Diseases (ICD) code, epidemiological data may be misrepresented by using administrative databases.
Current Pathogenetic Concepts in PSC
Characteristic histopathological features in PSC include a concentric (onion skin-type) periductal fibrosis of medium-sized and/or large bile ducts subsequently leading to strictures and obliterations. The primary injury is directed against cholangiocytes leading to overexpression of adhesion molecules, cytokines and growth factors, finally resulting in a fibro-obliterative process. Currently PSC is seen as immune-mediated rather than autoimmune-type liver disease as the evidence for the latter is, at least in part, counterintuitive (e.g. male predominance, lack of therapeutic effects of immunosuppressant agents, lack of highly disease-specific autoantibodies). Surprisingly, despite major methodological progresses in modelling various complex diseases, such as the generation of genetically modified mice, to date, no animal model has been developed that exhibits all of the desirable attributes of an 'ideal PSC model'. Although the precise pathogenesis of PSC is elusive, several different approaches exist covering three main areas: (i) to examine the leaky gut and the aberrant lymphocyte homing hypotheses pooling the close association of PSC with IBD; (ii) to define the possible role for 'toxic bile' in PSC and (iii) to uncover the genetic architecture of PSC by genome-wide association studies (GWAS). Interestingly, the high discrepancy in regard to susceptibility loci for ulcerative colitis (UC) and PSC suggests a specific genetic subtype referred to as 'PSC-IBD' mirroring this particular clinical presentation.
(i) Association with IBD: PSC is strongly associated with IBD, which has stimulated several hypotheses linking the pathophysiology of both diseases. As such, ongoing and exaggerated intestinal inflammation causes disturbed gut barrier function in IBD patients which may lead to increased intestinal permeability, lymphocyte activation and consequently increased number of mononuclear cells in the portal blood. IBD is frequently associated with extraintestinal disorders such as skin (i.e. pyoderma gangrenosum, erythema nodosum) and eye (i.e. anterior uveitis) involvement, which occur in parallel to bowel inflammation. This has been proposed as a result of the recruitment of activated effector cells from the gut. As PSC and associated ankylosing spondylitis are independent from IBD activity and PSC may even develop after total proctocolectomy, it has been suggested that long-living gut-derived memory lymphocyte populations circulate for years and consequently may induce liver inflammatory processes. In addition, perinuclear antineutrophil cytoplasmatic antibodies (pANCA) frequently observed in both PSC and IBD patients, may origin from cross reactions between tubulin beta 5 and intestinal bacterial antigens further supporting the gut–liver concept of PSC pathogenesis.
(ii) The 'toxic bile' concept: Several studies in mice demonstrated that genetic or toxin-induced modulation of bile composition induces sclerosing cholangitis and biliary type of liver fibrosis [e.g. Abcb4/Mdr2 (canalicular phospholipid flippase) knock-out mice; feeding the porphyrinogenic substance 3,5-diethoxycarbonyl-1,4-dihydrocollidine or toxic lithocholic acid]. Abcb4/Mdr2 (the rodent orthologue of ABCB4/MDR3) knock-out mice, which show absence of phosphatidylcholine in bile, develop bile duct injury similar to PSC.
(iii) Genetics: According to recent studies, the high genetic impact in the aetiopathogenesis of PSC becomes apparent. The fact that first degree relatives of PSC patients are at significantly higher risk to develop PSC and have a higher risk of developing IBD without PSC compared with the overall population, indicates genetic associations. A recent study aimed to characterize genetic susceptibility to PSC by means of a genome-wide association analysis of 443.816 single nuclear polymorphisms (SNPs) in 285 Norwegian PSC patients and 298 healthy controls. The results revealed strong associations near the extended HLA complex on chromosome 6p21 and weaker associations at the GPC5/GPC6 (GPC, glypican) region on chromosome 13q31 as well as genetic loci on chromosome 3p21 and 2q35, which have been implicated in genetic susceptibility to IBD. Direct evidence for bile toxicity/abnormal bile composition in PSC patients is lacking and the associations between ABCB4 or PXR (nuclear xenobiotic receptor) variants and PSC are weak [20, 21]. Therefore gene variants of ABCB4 and PXR may represent disease modifiers rather than direct causative factors for PSC. In addition, bile salt export pump (BSEP) and MDR3 variant segregation and haplotype structure in a Caucasian population could not detect differences between healthy individuals, primary biliary cirrhosis (PBC) and PSC patients. Notably, TGR5, the G-protein coupled bile acid receptor involved in bile acid-induced fluid secretion in the biliary epithelium, has recently been identified as a potential candidate gene and significant TGR5 associations have been detected in both PSC and UC. In addition, TGR5 variants may also predispose individuals for inflammation as this receptor is highly expressed on Kupffer and endothelial liver cells. Furthermore, three UC susceptibility loci harbouring the putative candidate genes REL, IL2 and CARD9, have been identified to be associated with PSC, suggesting an important role for innate (REL, CARD9) as well as adaptive immune system (IL2-IL2R pathway). Recently, genetic variants of matrix metalloproteinase 3 (MMP3) have been reported to influence PSC progression. MMP3-related mechanisms may play a role in determining disease severity in PSC but there seems to be lack of association with disease susceptibility.
Clinical Features of PSC
The diagnostic challenges in PSC originate from the lack of specific symptoms and clinical or biochemical findings. The 'typical' PSC patient is a young male suffering from IBD showing up with clinical and/or biochemical evidence of cholestasis. At the time of diagnosis, the majority of patients present with elevated alkaline phosphatase (AP) levels, whereas gamma glutamyl transpeptidase (γGT) and bilirubin levels may be within normal range. Half of the patients describe fatigue, pruritus and right upper quadrant pain. Weight loss, fever and chills may be seen in some patients. In 60–80% of PSC patients IBD either coexists or pre-exists years before PSC manifests. Most of these patients (80%) have UC with particular clinical and endoscopic features (right-sided predominance and mild to moderate clinical course). Crohn's disease (CD) has been reported in up to 13% of PSC patients, usually showing involvement of the colon with rectal sparing. IBD patients can develop PSC even years after total colectomy and IBD may manifest for the first-time after patients have already undergone LT for PSC. Therefore, it is highly recommended that PSC patients should undergo colonoscopy including biopsies, even if no IBD symptoms are present.
Features of autoimmune hepatitis (AIH) have been described in 6–9% of PSC patients. In PSC-AIH overlap syndromes, features of additional autoimmunological liver- and pancreatic diseases may occur. When elevated transaminases are observed in combination with positive anti-nuclear antibodies (ANA) or anti-smooth muscle antibodies (SMA), a liver biopsy should be considered for confirming the diagnosis of PSC-AIH overlap syndrome. Small duct PSC or IgG4 associated SC (IgG4-SC) (see below) represent further differential diagnostic challenges. Notably, elevated IgG4 serum levels and positive immunostaining for IgG4 on liver biopsies have been detected in SC suggested to be 'primary' in its origin. In addition, patients with IgG4-SC may also suffer from IBD, suggesting that classifying IgG4-SC as a form of secondary sclerosing cholangitis (SSC) has to be questioned or vice versa diagnostic criteria of PSC have to be rephrased. This important clinical question will need further meticulous studies as this problem remains to be unresolved.