Abstract
Primary sclerosing cholangitis (PSC) is an idiopathic progressive fibrosing cholangiopathy characterized by chronic inflammation and injury of the intra- and/or extrahepatic bile ducts, leading to fibrosis and cholestasis. Although much work has been done over the last several decades to further our understanding of this disease, the underlying mechanisms of pathogenesis remain elusive. To date, there are no effective therapies, and PSC continues to be a well-recognized cause of end-stage liver disease requiring liver transplantation in adults and children.
Primary sclerosing cholangitis (PSC) is an idiopathic progressive fibrosing cholangiopathy characterized by chronic inflammation and injury of the intra- and/or extrahepatic bile ducts, leading to fibrosis and cholestasis. Although much work has been done over the last several decades to further our understanding of this disease, the underlying mechanisms of pathogenesis remain elusive. To date, there are no effective therapies, and PSC continues to be a well-recognized cause of end-stage liver disease requiring liver transplantation in adults and children.
Epidemiology
Primary sclerosing cholangitis is a rare disorder. A systematic review and meta-analysis of adult studies from North America and Europe reported an incidence of 0.77 per 100,000 person years, with males 1.7 times more likely to be affected than females [1]. The median age of diagnosis was 41 years and approximately 68% of patients had concomitant inflammatory bowel disease (IBD). There is also a differential geographical distribution, with northern European countries having a higher burden of PSC compared to southern European countries. The incidence and prevalence in Sweden was reported to be 1.22 and 16.2 per 100,000 person years respectively, compared to that in Spain (0.7 and 0.22 per 100,000 person years respectively) [2]. The incidence of PSC has been reported to be increasing with time. In Sweden between 1992 and 2005, the incidence was reported to increase by approximately 3% per year [3]. Two other European studies have also reported statistically significant increased incidence of PSC over a similar study period [4, 5].
In children, the incidence and prevalence of PSC are approximately one-third of that reported in adults at 0.2–0.23 per 100,000 person years and 1.5–2.1 per 100,000 person years, respectively [6, 7]. In the largest multicenter cohort studied (781 patients with pediatric onset PSC), the median age of diagnosis was 12 years, with males consisting of 61% of the cohort [8]. IBD was present in 76% of patients, of which 83% had ulcerative colitis (UC) or indeterminate colitis and 17% had Crohn disease (CD). Overall 87% of the cohort had abnormal cholangiograms, large duct PSC, while patients with normal cholangiograms and evidence of biliary injury on liver biopsies, small duct PSC, represented 13% of the cohort [8].
Autoimmune sclerosing cholangitis (ASC), an overlap syndrome of PSC and autoimmune hepatitis (AIH), may be a unique pediatric autoimmune liver disease. The precise diagnostic criteria for ASC have yet to be established. However, reported laboratory features include elevated ALT and AST levels, hypergammaglobinemia, the presence of typical autoantibodies found in AIH, and the histological characteristic of both PSC and AIH. Single center adult studies have reported that approximately 7% of patients with PSC and 10% of patients with AIH may fall into the ASC category [9, 10]. In children, ASC appears to be much more prevalent, affecting 33–50% of children with autoimmune liver disease [7, 11]. There is ongoing research to determine whether PSC, ASC and AIH are distinct entities or represent a continuum of autoimmune liver disease (Figure 21.1).
Figure 21.1 Proposed spectrum of disease and relationship between autoimmune hepatitis (AIH) and primary sclerosing cholangitis (PSC). ASC, autoimmune sclerosing cholangitis.
Pathogenesis
The underlying pathophysiology of PSC remains unknown. The current hypothesis posits that in genetically susceptible individuals, environmental exposures trigger immune-mediated biliary injury. As such, genetic and environmental factors in addition to aberrant host immune homeostasis have been implicated.
Genetics
Hereditary factors likely play a role in the pathogenesis of PSC. Swedish registry data have identified that first degree relatives are 11.5, 11.1 and 2.3 times more likely to develop PSC in subjects with children, siblings and parents, respectively, affected by PSC [12].
Recent genome-wide association studies (GWAS) have identified more than 20 risk alleles associated with PSC [13–19]. The strongest signal has been associated with the human leukocyte antigen (HLA) region of chromosome 6 [15]. Non-HLA genes associated with immune activation (IL2, IL2RA, IL21, and NFKB1) and T cell modulation (TNFRSF4, BCL2L11, CTLA4, and FOXP1) have also been identified, further supporting the role of the adaptive immune system in the pathogenesis of PSC. Not surprisingly then, patients with PSC may also be at risk of developing other autoimmune conditions. In a single center cross sectional study of 245 adults with PSC, 6% had psoriasis, 2% had rheumatoid arthritis/ankylosing spondylitis and 4% with autoimmune thyroid disease. Similarly in children, concomitant autoimmune diseases were found in approximately 7% of patients including thyroiditis, celiac disease, type 1 diabetes, systemic vasculitis, juvenile idiopathic arthritis, epidermolysis bullosa, interstitial nephritis, uveitis, alopecia areata, and psoriasis, myasthenia gravis, and idiopathic thrombocytopenia [8].
On examination of the genetic links between IBD and PSC, the correlation between PSC and both CD and UC was weaker than between UC and CD [18]. This suggests that patients with PSC and IBD may represent a unique phenotype, and that environmental exposures play a critical role in the development of PSC.
Environmental Factors
Diet
A number of environmental exposures in adults with PSC have been examined. A single center study from Norway reported that coffee consumption and smoking were protective against PSC [20]. Patients with PSC, median age 47, were less likely to self-identify as coffee drinkers, had less cups per day at the time of the questionnaire and at the age of 18 compared to healthy controls. PSC patients were also less likely to drink coffee daily. In patients where biochemical data was available at the time of diagnosis, coffee drinkers with PSC had significantly lower ALT, AST and GGT levels than non-coffee drinkers with PSC. Smoking was also identified as a protective factor in patients with and without IBD. In a sub-analysis, women with PSC were less likely to have ever used hormonal contraception. The formulation of hormonal contraception (i.e., estrogen containing or progesterone only) was not available for analysis. There was also a correlation between the number of pregnancies prior to the diagnosis of PSC and an older age of PSC diagnosis, suggesting that hormones may also play a role in the pathogenesis of PSC.
A North American multicenter case-control study of 1,000 PSC patients and 663 controls without liver disease completing self-administered questionnaires identified several life style and dietary risk factors [21]. In contrast to the Norwegian study, smoking was protective against PSC only in patients with IBD; however, in the absence of IBD, smoking was not protective. Patients with PSC irrespective of IBD status were less likely to consume fish and grilled/barbecued meats, but more likely to consume well done steaks and hamburgers.
Microbiome
An increasingly recognized “environmental exposure” is the gut microbiome. Over the last decade, there has been a spate of focused research addressing IBD and its relationship with the intestinal microbiota. Given the strong association between PSC and IBD, a number of studies have also examined the microbiome in subjects with PSC [22–25]. A thorough review of the evidence and concepts of the microbiome in PSC is available [26]. Although studies vary in design and sample size, overall, the microbiota in patients with PSC exhibits significantly decreased diversity compared to those with IBD without PSC and from healthy controls. In patients with PSC and IBD, there is an increased proportion of the genus Veillonella, Escherichia, Lachnospiraceae and Megasphera with a reduction in genus Prevotella, Roseburia, and Bacteroides [25, 27]. In particular the genus Veillonella, a gram-negative anaerobe, has been consistently associated with patients with PSC. However, it is unclear how Veillonella contributes to the pathogenesis of PSC and whether the changes in microbiota drive disease or are a result of disease, or the treatment received, remain unanswered.
The mechanisms by which the microbiome is involved in the pathogenesis of PSC have been postulated to include: (1) direct elicitation of an immune response – the “Leaky Gut” hypothesis (see below); (2) molecular mimicry, whereby bacterial components trigger a sustained immune cross reaction to self-antigens resulting in liver injury; (3) regulating responsiveness and tolerance of the adaptive immune pathway; (4) dysbiosis leading to altered bile acid metabolism, accumulation of toxic bile acids and altered bile acid homeostasis leading to hepatotoxicity [26].
Immune Mediated
Leaky Gut
The “Leaky Gut” hypothesis postulates that impaired intestinal barrier function allows the “leakage” of bacterial products and pathogens into the portal circulation, causing inflammation. In support of this hypothesis, Nakamoto et al. demonstrated in a mouse model inoculated from stools samples from patients with PSC that bacterial pathogens including K. pneumonia, P. mirabilis and E. gallinarum act synergistically to activate the T helper 17 pathway causing hepatobiliary injury [28]. Moreover, particular strains of epithelial pore forming K. pneumonia were critical for epithelial disruption and bacterial translocation causing subsequent immune activation.
Abnormal T Lymphocytes Homing
Aberrant T lymphocyte homing to the liver has also been postulated to play a role in the pathogenesis of PSC [29]. Evidence suggests that lymphocytes are activated in the inflammatory milieu of the diseased bowel and then recruited into the liver through expression of liver-specific adhesion molecules. Circulating lymphocytes require the expression of integrin α4β7 and chemokine receptor CCR9 to localize to the intestine through the interaction with mucosal addressin-cell adhesion molecule 1 (MAdCAM-1) and the chemokine CCL25 respectively. Although not normally expressed in the liver, MAdCAM-1 has been detected in the hepatic endothelium of patients with PSC [30]. Another molecule under study is vascular adhesion protein-1 (VAP-1). Generally expressed at low levels in the normal gut, VAP-1 expression is increased during inflammation, suggesting that lymphocytes may also use VAP-1 to enter the mesenteric endothelium [31]. VAP-1 has also been found to be expressed in lymph nodes and along the endothelial surfaces of the liver. Taken together, these studies have furthered our understanding of the recruitment of gut specific T cells to the liver. Of note, patients may even develop PSC many years after proctocolectomy. It is hypothesized that T lymphocytes activated during colonic inflammation become memory T cells that persist in the circulation for prolonged periods of time. When activated, these memory T cells could cause chronic inflammation leading to PSC.
Cholangiocyte Injury
Bicarbonate Umbrella
The microenvironment adjacent to the canalicular membrane is hostile. Cholangiocytes are exposed to high concentrations of bile acids at the epithelial surface and therefore any disruption to the protective homeostatic mechanisms may result in biliary injury. The bicarbonate umbrella hypothesis posits that an alkaline environment is required to maintain bile acids in a negatively charged state. Due to the electromagnetic force, these negatively charged bile acids are unable to come into direct contact with the negatively charged phospholipid bilayer of the cholangiocyte cell membrane, thus protecting it from the detergent properties of the bile acids which would otherwise cause direct cell injury and death. GWAS studies which have identified TGR5 and FUT2 as risk alleles further support this hypothesis. TGR5 is a bile acid sensing receptor which increases bicarbonate excretion through the cystic fibrosis transmembrane conductance regulator (CFTR) and the chloride-bicarbonate anion exchanger 2 (AE2) in a cAMP-dependent manner in the presence of hydrophobic bile acids [32, 33]. FUT2 is a glycocalyx stabilizing protein which may provide a physical barrier to protect cholangiocytes from the high concentrations of bile acids [13].
Cholangiocyte Senescence
Chronic injury to cholangiocytes may also cause cellular senescence, a state of arrest of the cell cycle, to prevent proliferation of cells at risk of malignant transformation [34]. The senescent cholangiocytes may develop into a senescence-associated secretory phenotype (SASP) and secrete pro-inflammatory cytokines and chemokines triggering further biliary injury. This is currently an active area of research as “senolytics” may prove to be novel therapeutic agents.
The current understanding of the pathophysiology of PSC is summarized in Figure 21.2.
Figure 21.2 Schematic representation of potential factors in the pathogenesis of primary sclerosing cholangitis (PSC). 1) Immunopathogenesis is supported by susceptibility to PSC with genes encoding HLA class I and II I-like proteins, determining adaptive (CD4 T-cell) and innate (natural killer (NK) cell) responses, respectively, both injuring biliary epithelial cells (BECs). 2) The “leaky gut” with bacterial structures, pathogen associated molecular patterns (PAMPs), and components of bacterial cell membranes, lipopolysaccharides (LPS), “leaking” into portal circulation where they are recognized by pattern recognition receptors (PRRs) expressed on hepatic macrophages (MP) or dendritic cells (DC), triggering the production of proinflammatory cytokines (TNFα, IL-12, IL18). 3) The “dual homing” hypothesis is based on findings in patients with PSC–inflammatory bowel disease, showing that effector memory T-lymphocytes from the gut are recruited to the liver via mucosal addressin cell adhesion molecule (MAdCAM-1) and vascular adhesion protein 1 (VAP-1), which is expressed on portal vein/sinusoidal endothelium under inflammatory conditions. 4) Disruption to the protective barrier of the “bicarbonate umbrella” exposes cholangiocytes to toxic bile acids leading to the production of inflammatory cytokines activating nearby myofibroblasts causing biliary fibrosis. Chronic injury to the biliary epithelium may lead to cholangiocyte to develop into a senescence associated secretory phenotype (SASP) in which proinflammatory cytokines are secreted directly from the cholangiocytes. BA, bile acids; IL, interleukin; CCR9, chemokine receptor type 9; PMN, Polymorphonuclear leukocytes (neutrophil); PV, Portal vein; TLR4, Toll-like receptor; TNF, tumor necrosis factor;
Clinical Features and Diagnosis
The progression of PSC can be insidious. Based on large cohort studies in adults, approximately 50% of patients are asymptomatic at time of diagnosis, and liver disease is only suspected in the setting of elevated liver enzymes and cholestasis [35]. In a cross-sectional study 322 adults with IBD and at least 20 years of follow-up were screened with magnetic resonance cholangiopancreatogram (MRCP); 24 patients were found to have lesions consistent with PSC even though only seven patients were known to have PSC [36]. Moreover, progressive disease was noted when the MRCP was repeated after a median of 3.2 years in the patients with subclinical PSC. When signs and symptoms are present, hepatomegaly (44%) was most frequent, followed by splenomegaly (39%), abdominal pain (20%), pruritus (10%) and fatigue (6%) [6, 37].
Children may present with signs and symptoms similar to those in adults (Table 21.1). In pediatric series, reported symptoms included abdominal pain (36–40%), fatigue (23–35%), jaundice (17–19%), fever (13–17%), weight loss/delayed growth (17–19%), pruritus (15–17%), anorexia (10%) [38, 39]. In addition, delayed puberty is unique to the pediatric population and should be documented at the time of PSC diagnosis.
Table 21.1 Clinical features of sclerosing cholangitis in children
| Features | |
| Symptoms | Abdominal Pain |
| Fatigue | |
| Fever | |
| Weight loss | |
| Pruritus | |
| Delayed Puberty | |
| Diarrhea | |
| Gastrointestinal hemorrhage | |
| Sign | Hepatomegaly |
| Splenomegaly | |
| Jaundice | |
| Hepatosplenomegaly | |
| Ascites | |
| Xanthomas | |
| Skin excoriations | |
| Other | Elevated cholestatic markers |
| Presentation as autoimmune hepatitis poorly responsive to therapy | |
Inflammatory Bowel Disease
The relationship between IBD and PSC is well established and there appears to be a unique PSC-IBD phenotype. Multiple studies in the adult PSC population have shown an association with the specific triad of pancolitis, rectal sparing and backwash ileitis [40, 41]. In patients with PSC and CD, ileocolonic CD was more common than ileitis only [41]. There is also evidence that patients with PSC-UC have a milder course of UC compared to those with UC without PSC [42]. Chronic pouchitis appears to be more common in patients with PSC compared to those without [43].
The PSC-UC phenotype characterized by pancolitis, mild inflammation, back wash ileitis and rectal sparing has also been identified in children. In a single center retrospective study of 39 PSC-IBD compared to 95 IBD without PSC patients, the PSC-UC phenotype was characterized by pancolitis (96% vs. 64%, p < 0.001), but the colitis was less severe – patients with IBD-PSC had fewer IBD-related hospital admission (32.4%, 63.2%, p = 0.003) and were less likely to require treatment with steroids (76.5% vs. 91.6%, p = 0.03) or undergo colectomy (5.8% vs. 24.2%, p = 0.02) [44]. Another single center retrospective study of 52 children with PSC found rectal sparing in 27% of patients with PSC-IBD; pouchitis was a common complication of those who required protocolectomy [45]. This study also highlighted that 11% of children had asymptomatic IBD. It should also be noted that three (6%) patients in their cohort had evidence of colorectal dysplasia, reinforcing the need for screening colonoscopies, even in children. Another study from Canada reported that children with PSC-IBD were more likely to have endoscopic and histologic evidence of disease activity despite a lack of clinical symptoms as defined by a Pediatric Ulcerative Colitis Activity Index (PUCAI) score of less than 10 [46]. Fecal calprotectin correlated with endoscopic findings, and may be a useful screening tool in this population.
Laboratory Findings
Elevated ALP levels and mildly elevated aminotransferase values are typical biochemical findings for adult patients with PSC. ALP levels have been linked to prognosis and changes in ALP levels are used as an end point in several clinical trials. In a retrospective review of 87 newly diagnosed patients with PSC, complete normalization of ALP levels was associated with decreased incidence of cholangiocarcinoma, liver transplantation and death [47]. Children with PSC also have mildly elevated aminotransferase levels. However, ALP is neither sensitive nor specific for pediatric PSC [11, 38, 39]. ALP levels in children may be confounded by rapid bone turnover, especially during periods of growth. In children, serum GGT levels are more sensitive and specific than ALP levels for biliary injury. In large pediatric series, 100% of 92 patients had elevated GGT levels while only 88% had elevated ALP levels [8, 38, 39, 48–51]. Serum GGT levels may also have prognostic value, with higher GGT concentrations at time of PSC diagnosis correlating with worse outcomes, and complete normalization of GGT associated with good outcomes [7].
Autoantibodies
A number of autoantibodies have been detected in patients with PSC. Excluding the autoantibodies associated with AIH to diagnose ASC, all other autoantibodies including p-ANCA, ASCA, AMA, anti-biliary epithelial cell antibodies, have not been shown to be helpful clinically. Non-specific elevation of these autoantibodies is thought to be secondary to the non-specific immune dysregulation in PSC. A full review of autoantibodies in PSC is available for the interested reader [52].
Imaging
Historically, endoscopic retrograde cholangiopancreatography (ERCP) with injection of contrast into the biliary system to visualize gross changes was required to confirm the diagnosis (Figure 21.3). There has been a rapid evolution of MRI-based imaging and the advent of MRCP; this has now become the most commonly used imaging modality to detect biliary abnormalities in patients with PSC. A meta-analysis evaluating the role of MRCP in adult patients with PSC showed that MRCP had a summary area under the receiver characteristic (AUROC) of 0.91 sensitivity of 0.86 and specificity of 0.94 [53]. This is consistent with small single center pediatric studies showing a sensitivity of 0.81–0.84 and specificity of 1.0 [54]. In patients with an equivocal MRCP, ERCP may be used to inject contrast to visualize the biliary system under pressure to better visual abnormalities. Otherwise, ERCP is generally reserved for therapeutic interventions such as dilatation of strictures, stent placements, stone removal or brush sampling.
Figure 21.3 Typical endoscopic retrograde cholangiopancreatography changes in a patient with sclerosing cholangitis. There is overall irregularity of the intrahepatic ducts and areas of distinct stenosis, for example the junction of the right and left hepatic ducts (large arrow); the distal intrahepatic ducts are relatively spared (small arrows). Note the unusually long cystic duct (asterisks), which is a normal variant.
In patients with large duct PSC, radiographic abnormalities in the biliary system are characterized by multiple focal strictures and dilatations (i.e., “beaded” appearance), long-segment biliary stricture, dilation of the intrahepatic and/or extrahepatic bile ducts, and the presence of “floating” or “isolated” peripheral intrahepatic bile ducts (i.e., pruning of the biliary tree) (Figure 21.4). The Majoie classification was developed to objectively score biliary changes in PSC [55] (Table 21.2). Correct stratification of patients based on this score is important as it has been shown to have prognostic value. A Dutch study of 174 adults with PSC showed that combining the intra- and extrahepatic Majoie scores strongly predicted survival [56]. Patients with the lowest and highest combined scores had a 15-year survival of approximately 90% and 50%, respectively. In pediatrics, studies have not found any correlation between the Majoie classification and histologic fibrosis nor prognosis [57].
Figure 21.4 3D MRCP maximum intensity projection (MIP) images from 1.5T scanner with. (A) Normal intrahepatic and extrahepatic biliary tree. (B) Diffusely irregular intrahepatic biliary tree with increased number of ducts visualized, multiple narrowings/strictures (arrows). Extrahepatic biliary tree is diffusely dilated (arrowheads). (C) Common bile duct stricture (arrow) and dilated common bile duct. (D) Dilated common bile duct and left hepatic duct with an attenuated appearance of peripheral ducts resembling a pruned appearance.
Table 21.2 Majoie Classification for cholangiographic findings in primary sclerosing cholangitis
| Intrahepatic | |
| I | Multiple strictures; normal caliber of bile ducts or minimal dilatation |
| II | Mulitple strictures, saccular dilatations, decreased arborization |
| III | Only central branches filled despite adequate filling pressure; severe pruning |
| Extrahepatic | |
| I | Slight irregularities of duct contour; no stenosis |
| II | Segmental stenosis |
| III | Stenosis of almost entire length of duct |
| IV | Extremely irregular margin; diverticulum like outpouching |
In small duct PSC, imaging of the biliary system is normal, but evidence of biliary injury is identified on liver biopsy. A nomenclature of disease based on histology and cholangiography has been suggested (Table 21.3).
Table 21.3 Ludwig’s classification of duct disease in primary sclerosing cholangitis
| Suggested terminology | Liver biopsy features | Cholangiographic features |
| Small duct PSC | Typical | Not diagnostic |
| Large duct PSC (extra- and/or intrahepatic) | Not diagnostic | Typical |
| Combined large and small duct PSC (Global or classic PSC) | Typical | Typical |
On gross inspection of the large intrahepatic and extrahepatic bile ducts, the walls are noted to be thickened and fibrous [58]. Areas of narrowing and dilation may be present, corroborating MRCP features of large duct disease.
Histology
The classic histologic features of PSC are biliary obstruction, inflammation and fibrosis. Although the disease may be patchy, histological parameters of biliary obstruction including an increased number of bile duct profiles and bile ductular reaction may be present [59] (Figure 21.5). Biliary inflammation characterized by acute cholangitis, lymphocytic cholangitis, pericholangitis and lymphoplasmacytic infiltration may also be variable given the limited sampling from percutaneous liver biopsy. In the setting of ASC, interface hepatitis and a more impressive lymphoplasmacytic infiltration can be appreciated (Figure 21.6).
Figure 21.5 Typical histologic features of primary sclerosing cholangitis. A) Bile ductular reaction (black arrow) and bile duct proliferation (white arrow) with evidence of peri-cholangitis. B) Periductal sclerosis (onion skinning) Images taken at 100x magnification.
Figure 21.6 Liver biopsy from a patient with ASC. Note prominent lymphoplamsacytic infiltrate along the limiting plate similar to that seen in AIH, and the presence of periductal sclerosis. Image taken at 100x magnification.
In general, the surrounding hepatocytes show non-specific changes. However, as intralobular bile ducts are progressively lost, copper deposition and canalicular cholestasis may be noted. The degree of inflammation in PSC is relatively mild compared to autoimmune hepatitis and chronic viral hepatitis. The characteristic periductal sclerosis (i.e., onion skinning) is seen in only a minority of patients. Significant fibrosis and bile duct paucity are present in later stages of disease [60].
Several histology-based scores have been used to stage disease progression in patients with PSC. The Ludwig score was the first score developed to stage PSC [59] (Figure 21.7). Stage 1 is characterized by inflammatory and fibrotic changes to be confined within the portal tract. A mixed inflammatory infiltrate composed of lymphocytes, eosinophils, and neutrophils are found in the portal tracts, particularly around the bile ducts. In stage 2, the portal tracts are expanded with disruption of the limiting plate. Biliary piecemeal necrosis may be present and bile ductular proliferation is present. Mild portal fibrosis with thin radiating septae may be present. Bridging fibrosis defines stage 3 and cirrhosis is the hallmark of stage 4. In the later stages of disease, inflammation is not a predominate feature.
Figure 21.7 Schematic representation of the four stages of liver disease (chronic hepatitis) associated with primary sclerosing cholangitis. In the highpower view of stage II, clusters of hepatocytes are shown within the enlarged portal tract; cholangitis is also depicted. In stage III, portal-toportal fibrous bridging is present. The low-power view of stage IV shows a garland-shaped regenerative nodule. In stages III and IV, duct obliteration is present; the presence of these duct abnormalities would not be essential for staging.
The Ishak stage was originally developed for chronic hepatitis, and has been widely applied to report hepatic fibrosis in liver disease of various etiologies [61]. Stage 0 represents no fibrosis. There is fibrous expansion of the occasional portal area in stage 1 while most portal areas are affected in stage 2. Fibrous expansion of most portal areas with occasional portal-to-portal bridging is noted in stage 3 and marked portal-to-portal bridging is present in stage 4. Significant bridging fibrosis with occasional nodules is indicative of stage 5, and finally, stage 6 is characterized by cirrhosis.
Nakanuma et al. developed a histologic score for primary biliary cirrhosis (PBC) [62]. Since PBC and PSC share common features such as biliary injury and bile duct loss, a study examining the use of the Nakanuma stage for PSC patients found that it was predictive of transplant-free survival [63]. The Nakanuma stage is calculated by combining three components, each scored from 0–3, consisting of fibrosis, bile duct loss, and deposition of orcein positive granules. Stage I is a score of 0, stage II is a score of 1–3, stage III is score of 4–6 and stage IV is a score of 7–9. Patients with stage 1 had 100% transplant-free survival while patients with stage 4 had 0% transplant-free survival at a median follow-up time of 112 months. A multicenter study of 119 adults validated the prognostic value of the Ludwig, Ishak and Nakanuma histologic scoring systems of PSC [64]. Both the Ludwig score and the Ishak stage correlate with liver transplantation and liver-related events such as gastroesophageal variceal bleeding, development of ascites, splenomegaly, or hepatic encephalopathy. However, only the Nakanuma stage was predictive of PSC-related death and liver transplantation.
Pediatric specific studies are limited; therefore how the above studies translate to children remains uncertain.
Prognosis
The natural history of PSC has been studied in several large adult cohort studies [59, 65–68]. In general, disease progression was slow, with the median duration from diagnosis to liver transplantation or death of 12–18 years. The most common causes of death were cholangiocarcinoma or liver failure. These studies have identified a number of prognostic factors including advanced age at diagnosis, elevated serum bilirubin concentration, and a high histologic stage to be associated with poor prognosis. The presence of IBD as a poor prognostic has been inconsistent between studies. Weismuller et al. studied a large international multicenter retrospective cohort of 7,121 patients with PSC in Europe, North America and Australia [69]. Approximately 37% of patients required liver transplant or died at a median of 14.5 years form diagnosis; 8% developed cholangiocarcinoma with increased age at diagnosis correlating with increased risk of malignancy. The incidence rate was 1.2 per 100 patient-years for patients younger than 20 years old, up to 21.0 per 100 patient-years for patients older than 60 years. The study also identified the presence of CD (HR, 0.62) or no IBD (HR, 0.90) as protective against liver transplant or death compared to the presence of UC. Risk of cholangiocarcinoma was also higher in those with UC compared to CD or no IBD. Women with PSC were also associated with a decreased risk of liver transplant or death (HR, 0.88) and malignancy (HR, 0.68) compared to men.
To determine whether the prognosis of small duct disease differs from large duct disease, a study from two large referral centers in Europe compared 33 patients with small duct PSC to 260 patients with large duct PSC [70]. Patients with small duct PSC had similar demographics for age of diagnosis, sex, presence and type of IBD. Serum bilirubin, ALT, AST and ALP were not significantly different between the two groups. Patients with large duct PSC were more likely to be symptomatic at time of presentation with weight loss, pruritus and jaundice. When followed for a mean of 105 months, 54 (31%) of the patients with large duct disease underwent liver transplantation or died, compared to 4 (12%) of the patients with small duct PSC. During the time of follow-up, four patients with small duct PSC at time of diagnosis developed large duct disease. In this study none of the small duct disease patients developed cholangiocarcinoma, while 11% of patients with large duct disease did. The association between small duct PSC and decreased risk of liver transplant, death and malignancy was also confirmed in the study by Weismuller et al. [69].
The natural history of pediatric PSC was studied in a multicenter international collaboration [8]. For 471 patients, the median individual follow-up time from diagnosis was 4.4 years with a total of 4,277 person years. Complications of portal hypertension were present in 5% of patients at time of diagnosis and in 38% after ten years of disease. Overall, 14% of patients required liver transplantation at a median of four years from time of diagnosis. Survival with native liver was noted in 88% and 70% at five and ten years after diagnosis, respectively; 1.4% of the cohort died from liver disease-related complications at a median of nine years after diagnosis. Cholangiocarcinoma developed in 8 (1%) of the patients aged 15–18 years in a median time of six years after diagnosis. Survival without complications of liver disease, transplantation or death was 70% and 53% at five years and ten years respectively. Similar to the adult studies, children with small duct PSC had a more favorable prognosis (HR, 0.69–0.71) for event-free survival compared to patients with large duct PSC. In contrast to the adult study mentioned above, the presence of IBD appeared to be protective for event-free survival in children (HR, 0.63–0.66), although IBD subtypes were not specifically studied.
The natural history of ASC is poorly characterized. A single center prospective study compared the clinical outcomes of children with AIH and ASC. Although both groups had a similar proportion of patients achieving biochemical remission on immunosuppression, patients with ASC were more likely to require liver transplantation during the follow-up period [71]. Deneau et al. reported that complication of liver disease including portal hypertension, obstructive cholangitis, liver transplantation and death within five years of diagnosis was 25% in ASC, which was higher than the 15% in AIH and less than the 37% in patients with PSC for the same time period [7]. The survival rate with native liver after five years of diagnosis for ASC was similar to AIH at 90% and 87%, respectively, compared to only 78% in children with PSC.
Management
Broadly, the goals of management are to provide symptom relief by improving biliary drainage, managing complications, and screening for malignancies associated with end-stage liver disease secondary to PSC (Box 21.1). A number of medical and surgical approaches have been used to manage patients with PSC (Box 21.2). There are currently no effective treatment strategies for patients with PSC. However, several approaches deserve mention.
Box 21.1 Therapeutic goals in the management of patients with primary sclerosing cholangitis
| Provide symptomatic relief |
|
| Improve biliary drainage |
|
| Prevent/recognize/ameliorate complications |
|
Box 21.2 – Medical and surgical therapies used to manage sclerosing cholangitis and its complications
| Medical |
|
| Surgical/endoscopic |
|
Ursodeoxycholic acid (UDCA) is a hydrophilic bile acid with choleretic, immunomodulatory and anti-apoptotic effects; this agent is a commonly used medication in patients with PSC. In a single center, double-blind study of 105 patients with PSC randomized to either placebo or UDCA (13–15 mg/kg/day), Lindor found that the UDCA group had significantly improved ALP, AST, bilirubin and albumin at one and two years compared to the placebo group [72]. There were no significant differences found in the two treatment arms in regard to time to liver transplantation. A subsequent study of 48 patients with PSC receiving UDCA 10 mg/kg/day for two years again showed significant improvement in serum ALP, GGT and AST levels with minimal side effects [73]. However, there were no differences in symptoms, histology or disease progression between the groups. Given the consistent improvement in liver biochemistries with UCDA, a double-blind randomized control trial was done to determine if high dose UDCA (28–30 mg/kg/day) would be more effective than placebo [74]. Seventy-six patients were enrolled into the UDCA arm and 73 patients into the placebo arm. The primary outcomes were cirrhosis, varices, cholangiocarcinoma, liver transplantation, or death. A liver biopsy and cholangiogram were performed prior to randomization and after five years of follow-up. The study was terminated after six years because of futility. Consistent with previous studies, the UDCA group had improved AST and ALP levels compared to the placebo group. Surprisingly, despite having similar entry characteristics, 30 patients (39%) from the UDCA group had reached one of the primary outcomes compared to only 19 patients (26%) in the placebo arm. It remains unclear how a medication thought to be so well tolerated and associated with biochemical improvement would lead to worse clinical outcomes. Importantly, this study also highlighted that ALP levels may not be the best indicator of disease activity and progression, given that the treatment arm had lower ALP levels despite poorer outcomes. It is worth noting that patients with normalization of ALP levels while receiving high dose UDCA still had good outcomes.
One possible explanation for the unexpected results was that higher doses of UDCA led to unabsorbed medication reaching the colon. In the colon, bacteria could metabolize UDCA into lithocholic acid (LCA), a hydrophobic bile acid with potentially hepatotoxic effects. In the rat model, LCA has been shown to cause injury to the biliary epithelium [75]. In mouse models of biliary obstruction, the addition of UDCA is associated with worsened liver histology and increased mortality. Mice with bile duct ligation fed UDCA developed higher biliary pressures, as a result of its choleretic effect, which led to rupture of biliary cholangioles and marked hepatocyte injury [76]. An alternative explanation for worsened clinical outcomes in the high dose UDCA group draws from its anti-apoptotic effects [77] described in the rat model. By preventing apoptosis of activated hepatic stellate cells, progressive fibrosis could ensue, leading to cirrhosis.
Due to the possible harmful effects of UDCA, Wunsch et al. studied UCDA withdrawal in 26 patients receiving 10–15 mg/kg/day chronically. Three months after the withdrawal of UDCA, ALP, GGT, total bilirubin, ALT and AST levels significantly increased from baseline. Serum bile acid analysis detected a decrease in LCA. While health-related quality of life indicators did not change, 42% of patients also reported increased pruritus.
A meta-analysis of 567 patients in nine randomized control trials reviewing the efficacy of UCDA (>15 mg/kg/day) in patients with PSC found no significant benefit for mortality, pruritus, fatigue, development cholangiocarcinoma, nor histologic progression [78]. Similarly, a Cochrane review of 592 patients enrolled in eight RTCs found no significant reduction in the number reaching liver transplant, complications of portal hypertension, histologic progression or cholangiographic progression despite significant improvements in serum bilirubin, ALP, AST and GGT levels [79].
Ursodeoxycholic acid is used in over 80% of pediatric patients with PSC [8]. In a single center retrospective study, in 47 children with PSC treated with UDCA at doses of 20–30 mg/kg/d, serum ALT, AST, ALP and GGT levels were significantly lower after one year of treatment [39]. Similar results have been replicated in other small single center studies using doses as low as 9 mg/kg/day [80].
In a large retrospective pediatric study, UDCA and non-UDCA treated patients had similar outcomes [81]. Although the UDCA-treated patients had lower GGT at one year compared to untreated patients, the rate of complications of portal hypertension, biliary complications, cholangiocarcinoma, liver transplantation and liver related deaths were similar between groups. Interestingly, this study found that complete normalization of GGT by one year after diagnosis, regardless of UDCA treatment, was associated with good clinical outcomes.
The effects on patient outcomes in relation to UDCA have not been studied prospectively in children.
Oral Vancomycin
Vancomycin is a glycopeptide antibiotic that prevents cell wall cross linking and inhibits cell wall synthesis. Thus it is effective against gram-positive bacteria and is an established treatment for C. difficile-associated diarrhea. Cox et al. reported the potential benefit of oral vancomycin in three children with PSC, IBD and C. difficile-associated diarrhea [82]. Liver enzymes normalized while receiving oral vancomycin, increased after discontinuation, and normalized again after restarting oral vancomycin. One patient had a liver biopsy while receiving oral vancomycin treatment which showed less portal inflammation than an earlier time point without oral vancomycin. The authors went on to report a cohort of 14 patients with PSC and IBD treated with oral vancomycin (50 mg/kg/day) for a median of 54 months [83]. All patients had significant improvements in their ALT, GGT, ESR levels and clinical symptoms. The authors also noted that patients with cirrhosis at time of treatment had a less dramatic improvement. A randomized double-blind study with oral vancomycin or metronidazole in adults showed that vancomycin significantly decreased ALP levels after 12 weeks of treatment while metronidazole did not. Both antibiotics were associated with decreased bilirubin levels, Mayo PSC risk score and pruritus at specific doses [84]. A triple-blind randomized placebo-controlled study of 29 adult Iranian patients, of which 77% had IBD, showed a significant decrease in the Mayo PSC risk score at 12 weeks compared to baseline in the oral vancomycin group; the placebo group did not improve [85]. The ALP, ESR, GGT levels and patient symptoms including improvement in fatigue, diarrhea, and anorexia improved compared to baseline in the oral vancomycin group. Patients in the placebo group had a significant improvement in pruritus but not in the oral vancomycin group.
In addition to its antimicrobial effects on colonic dysbiosis, the therapeutic mechanism of oral vancomycin may be related to its immunomodulatory effects. In children with PSC, circulating regulatory T cells (Tregs) increased during oral vancomycin administration compared to pre-treatment levels, and decreased when the vancomycin was discontinued. Re-initiating oral vancomycin was then associated with a rise in Tregs and normalization of liver enzymes [86]. Given the possible association with immune dysregulation, the increase in Tregs would be expected to down-regulate autoimmune pathways. Oral vancomycin therapy for the medical management of PSC appears promising in children and adults, however, whether all patients with PSC or only the subgroup with PSC-IBD would benefit remain unanswered. Thus, high quality randomized control trials with meaningful clinical endpoints are needed to determine the role of oral vancomycin in the management of PSC.
Related posts:
Chapter 5 – Cirrhosis and Chronic Liver Failure in Children
Chapter 13 – Familial Hepatocellular Cholestasis
Chapter 14 – Alagille Syndrome
Chapter 26 – Cystic Fibrosis Liver Disease in Children
Chapter 32 – Lysosomal Storage Disorders in Children
Chapter 43 – Liver Transplantation in Children: Indications and Surgical Aspects
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
