Two to four percentage of individuals with PSC will suffer from coexisting autoimmune hepatitis, termed as a PSC–autoimmune hepatitis overlap syndrome. Unlike individuals with isolated PSC, these individuals will fulfil the criteria for autoimmune hepatitis diagnosis. These individuals typically have transaminase levels greater than five times the upper limit of normal, IgG levels 1.5–2 times the upper limit of normal, high titre antibodies (antinuclear antibodies, anti-smooth muscle antibodies) and on liver biopsy demonstrate an excess of plasma cell infiltration with interface hepatitis. This should be treated by standard immunosuppressive therapies for autoimmune hepatitis such as induction with corticosteroids and maintenance with azathioprine or mycophenolate mofetil. However, unlike isolated autoimmune hepatitis, these individuals never normalise transaminase levels completely which is indicative of underlying PSC. Therefore, the treatment aim should be normalisation of IgG levels and an improvement of interface hepatitis on liver biopsy. This overlap of PSC would be distinct from the variant syndromes that are seen in paediatric populations where autoimmune sclerosing cholangitis (ASC) and autoimmune hepatitis are associated with significantly better outcomes when compared to PSC. This is in contrast to adult populations where PSC–autoimmune hepatitis is associated with accelerated disease, and earlier need for transplantation and higher rates of recurrent disease post-liver transplantation [7].

A proportion of individuals will present with features consistent with PSC and extensive biliary changes; however, they will suffer from a closely related condition termed IgG4 related disease which is a distinct entity characterised by dense lymphoplasmacytic infiltrates that are rich in IgG positive plasma cells. This disease can affect virtually every organ system and lead to intense fibrosis, autoimmune destruction and at times, the formation of inflammatory pseudotumours that can mimic cholangiocarcinoma. Individuals with IgG4 disease typically suffer not only from biliary disease but often autoimmune pancreatitis which often dominates their presentation and a relative absence of coexisting IBD although some cases have been reported. Infiltration with IgG4-secreting plasma cells is pathognomonic of this disease, and it can occur not only in the biliary tree and pancreas but also in the salivary glands, periorbital tissues, kidneys, lungs, lymph nodes, meninges, aorta, breast, prostate, thyroid, pericardium and the skin and has some similarity to sarcoidosis and its disease distribution. Although this disease can present with multiple liver masses and biliary changes suggestive of cholangiocarcinoma, these are mostly inflammatory pseudotumours that respond and disappear with immunosuppressive therapies. The mainstay of treatment for this disease is immunosuppression with corticosteroids and azathioprine with a high risk of relapse when discontinued [8].

The biliary tree is limited in its ability to respond to dithering insults and as such, a wide range of diseases affecting the biliary tree can lead to cholestasis and eventual fibrostenotic changes (Table 4.1). All of these diseases can have typical radiological changes associated with PSC and include recurrent gallstone disease, congenital ductal plate malformations such as Caroli’s disease, biliary tract trauma following cholecystectomy, cholangiocarcinoma, recurrent bacterial biliary infection, AIDS related cholangiopathy secondary to Cryptosporidium infection, portal vein thrombosis with portal venous cholangiopathy, hepatic arterial insufficiency with biliary structuring, immunodeficiency (most notably X-linked Hyper IgM syndrome) with recurrent Cryptosporidium infection and progressive cholangiopathy and associated cholangiocarcinoma. In addition to these causes, direct drug-induced effects and prolonged intensive care unit stay has also been associated with biliary stricturing as have a variety of micro-organisms which include HIV and cytomegalovirus [9].

Up to 20 % of patients with PSC will suffer with autoimmune diseases other than IBD. A recent study found significant association with coeliac disease, psoriasis, rheumatoid arthritis, sarcoidosis, sacroiliitis, Sjogren’s syndrome, type 1 diabetes and thyroid disease. This data is summarised in Table 4.2. The mere presence of these diseases is associated with increased mortality in patients with PSC. IBD is however the most dominant associated disease which occurs in roughly 70 % of PSC sufferers at a ratio of 6:1 ulcerative colitis to Crohn’s disease. IBD in PSC often precedes the diagnosis of PSC by a mean of 10–14 years and typically presents as a pancolitis. This can be asymptomatic with relative rectal sparing and predominantly right-sided inflammation. PSC-IBD has an increase in backwash ileitis but isolated small bowel IBD is uncommon [10]. A proportion of individuals with PSC will proceed to end stage liver disease and liver transplant only to develop de novo IBD following transplantation. This disease can often be aggressive and unresponsive to immunosuppressive drugs as it already occurs in the presence of potent immunosuppressive drugs required for the liver allograft. Thirty percentage of individuals with PSC appear not to suffer from IBD. It is unclear whether these individuals truly have no gut inflammation or whether this is a failure to detect subclinical inflammation by standard means as virtually all PSC patients will describe some degree of changes in their bowel habit and diarrhoea at some stage. Surgery for IBD in PSC is complicated by higher than expected rates of pouchitis when compared to patients with IBD alone. The rates of pouchitis in this population vary tremendously from 60 % to almost 100 % in some case series. There is accumulation of risk at 1, 2, 5 and 10 years following ileal pouch anal anastomosis of 22, 43, 61 and 79 % respectively for patients with PSC. Pouchitis in the setting of PSC is associated with poorer treatment responses compared to individuals without PSC and higher rate of subsequent pouch removal [11].

There is very little data on environmental risk factors for PSC. Possibly the strongest data is a negative association with cigarette smoking. In one study of 184 patients with PSC and age/sex matched controls, only 4.9 % of PSC cases smoked compared to 26.1 % of controls. 69.65 % had never smoked compared to 46.7 % of controls. These differences have since been replicated and although the absolute smoking percentages vary based on the population prevalence of smoking, these striking differences between PSC and controls remains. Some data would suggest that those with PSC that have smoked but subsequently stopped have a later onset of disease.

Less pronounced but equally significant is the negative association with coffee consumption. Seventy-six percentage of PSC cases versus 86 % of population controls (odds ratio 0.52) consumed coffee by age 18. These effects were largely restricted to male patients. Female patients with PSC reported less use of hormonal contraception use than controls. This potential interaction with female sex hormones is further underlined by the observation that female patients with multiple pregnancies develop PSC at a later age [12].

The majority of the genetic susceptibility to PSC is strongly clustered to chromosome 6 and the human leukocyte antigen (HLA) complex. The first association with this HLA complex was reported in 1982 where it was shown that HLA-B8 and HLA-DR3 were found in 80 % of patients with PSC and ulcerative colitis. The association with HLA-B8 was subsequently replicated in a further cohort where 60 % of PSC patients compared to 25 % in controls carried this risk allele. This provided the first evidence of a heritable component to PSC which was subsequently confirmed in a Swedish study of 678 PSC patients and 6,347 healthy controls which demonstrated a substantially increased risk of PSC in first-degree relatives of PSC patients with hazard ratios of 11.5 for the offspring. Despite this compelling data, familial patterns of PSC inheritance remain rare with only five patients in this cohort having first degree relatives with PSC (3.4 %). Thus the prevalence of PSC amongst first-degree relatives was 0.7 % with a prevalence in siblings of 1.5 %.

Genetic studies in PSC have largely focused on risk alleles that increase genetic susceptibility to the disease and have expanded the HLA haplotypes associated with PSC. The initial haplotype associated with PSC were B8, DR3 and D2 as well as A1, B8 and DR3. This was expanded to include HLA-DRW52A (DRB3*0101) which has an association with PSC between 55 and 100 % depending on the cohort. Subsequently, DR6 was added as a risk haplotype and strong association reported with DRB1*1301, DQA1*1301 and DQB1*0603. Much of the difficulty in deciphering the HLA has come from the relative homogenous genetic background of the study populations which in most studies only include Caucasians from Scandinavia and North Western Europe. The recent Immunochip experiment in PSC sequenced 3,789 PSC cases using a high resolution SNP chip. This experiment replicated the previous HLA alleles and demonstrated that the majority of PSC risk resides in HLA-B and HLA-DRB1. The study showed strong linkage disequilibrium with the HLA-B*08:01 haplotype which is part of the ancestral HLAB*08:01-DRB1*03:01 haplotype (AH8.1) which is an ancestral haplotype associated with multiple autoimmune diseases. This study added HLA-B*08:01, HLA-DQA*01:03, HLA-DQA*05:01, DRB1*15:01 and DQA*01:01 to the current list of risk haplotypes for PSC. It is clear that certain HLA haplotypes carry a risk of PSC whereas others afford protection to this disease and is summarised in Table 4.3 [13, 14].

The relationship between the HLA and disease cause has been explored with HLA-DR52A, patients reportedly developing more severe disease and earlier mortality. Similarly, HLA-DR4, DR3, DQ2 heterozygosity has been associated with rapidly progressive disease. Much more work is needed to decipher the HLA complex in PSC and also the potential association with disease progression. This is, however, a key element of PSC as the HLA class 1 and class 2 genes determine the ability of the immune system to recognise both self and non-self antigens and could be a critical part in understanding potential antigens that drive PSC.

Gut inflammation is associated with increased gut permeability and hence it has been postulated that PSC occurs because of persistent portal venous bacterial translocation. This has, however, not been borne out in several studies that have failed to show significant bacterial populations within the portal venous blood or the liver. A further complicating factor is that gut permeability in IBD is dependent on the severity of the IBD which does not translate to severity of PSC. This makes a direct pathogen translocation theory less likely. This does not, however, exclude the possibility of bacterial products from the gut being involved in PSC nor does it account for potential changes in the gut microbiota which is increasingly being linked with human disease. As described above, the FUT-2 mutation does directly affect the microbiota composition within the biliary tree with similar changes expected in the gut where FUT-2 polymorphisms are also associated with Crohn’s disease. IBD is associated with very significant alterations in the gut microbiota with loss of diversity and the emergence of dominant microbial flora.

Similarly, immunodeficiency syndromes can give rise to PSC-like syndromes. Both HIV as well as X-linked hyper IgM syndrome (CD40L deficiency) lead to biliary colonisation with cryptosporidium and a progressive cholangiopathy than can lead to cholangiocarcinoma. These specific defects have not been demonstrated in PSC but might well have accelerated the phenotype as recurrent biliary infections as are seen in chronic gallstone disease also lead to a sclerosing cholangitis type liver picture [9].

Viral infections are yet to be associated with PSC. Cytomegalovirus (CMV) can give rise to severe bile duct destruction and mimic PSC but does not lead to extended fibrosis and large duct disease. The effects on the biliary tree and HIV are thought not to be directly related to the HIV virus but rather persistent cryptosporidium or microsporidial infection of the biliary system.

Fungal infections are often seen in advanced large duct PSC as secondary infections. Evidence for a primary role of fungi in the pathogenesis of the PSC remains controversial. A large proportion of PSC patients will have high levels of anti-saccharomyces cerevisiae antibodies (ASCA). This appears to be independent of the stage of PSC or their IBD phenotype. Biliary cultures in these individuals have so far failed to reliably culture this organism, casting some doubt as its role in the pathogenesis of PSC.

One theory to explain PSC comes from the observation that mice deficient in the MDR2 (ABCB4) gene which is involved in cannulicular phospholipid generation develop a phenotype similar to PSC with intense peribiliary fibrosis. This has led to the hypothesis of toxic bile contributing to the development of PSC. To date, the MDR2 mutation has not been reported in humans with no evidence of polymorphisms in this locus in PSC. It is conceivable that protective mechanisms such as bicarbonate secretion is altered by inflammation or that inflammation leads to dysfunction of biliary transporters which leads to the accumulation of toxic bile acids which in turn directly interact with the cholangiocytes to drive apoptosis, fibrosis and possibly carcinogenesis. This might be a direct effect of bile acid accumulation or a secondary effect due to changes in the gut microbiota which in turn affects the enterohepatic recirculation bile acid. There is no doubt that chronic cholestasis is toxic to the biliary tree whether in advanced PSC or in secondary sclerosing cholangitis due to obstructive cholangiopathy, but evidence for a primary defect to explain PSC remains lacking.