A 70-year-old male nonsmoker with a past medical history significant for coronary artery disease and hypertension is admitted with a 1-month history of fatigue and new-onset jaundice. He denies any abdominal pain, fevers, or weight loss. He is taking no new medications and denies alcohol or drug use. There is no personal or family history of inflammatory bowel disease or gastrointestinal cancers. He underwent a remote cholecystectomy in the past and denies recent travel. Vital signs are stable, and examination reveals scleral icterus with mild jaundice, in the absence of cutaneous stigmata of chronic liver disease. Initial laboratory evaluation reveals the following: total bilirubin 7.2, ALT 229, AST 190, alkaline phosphatase 622, and lipase 27. CBC and coagulation profile are both within normal limits. Hepatitis serologies (A/B/C) are negative and an HIV test is nonreactive. CA 19-9 is moderately elevated at 118 and anti-nuclear antibody is nonreactive. Contrast-enhanced CT scan demonstrates mild intra- and extrahepatic bile duct dilatation in the presence of an ill-defined 4.5 cm lesion in the porta hepatis (Fig. 9.1). There is no evidence of choledocholithiasis or cholelithiasis. A MRCP is ordered and a gastroenterology consult is subsequently placed.

In Western countries, an isolated benign biliary stricture is frequently iatrogenic, with the majority of cases following operative trauma or cholecystectomy [8]. These stenoses are most often localized and result from scarring at the site of prior surgical resection or anastomosis. While this is an important consideration for our patient, the presence of a porta hepatis lesion and concomitant diffuse ductal dilation argues for an alternate etiology. As such, post-surgical biliary strictures (Chap. 8) shall be largely excluded from this case discussion.

One possible etiology for this patient’s presentation is primary sclerosing cholangitis (PSC) . PSC is a chronic fibroinflammatory condition that can lead to progressive and diffuse stenoses of both the intra and extrahepatic biliary tree [9]. Up to 80 % of PSC cases are associated with inflammatory bowel disease, most frequently ulcerative colitis, although increasingly Crohn’s disease as well [10]. Small-duct PSC has also been described [11] and requires liver biopsy for establishing the diagnosis given the paucity of findings on imaging.

PSC more commonly affects men (2:1) and usually presents between ages 30 and 40. If present, initial symptoms are frequently nonspecific and can include malaise, fatigue, and weight loss. Jaundice typically occurs later in the disease process, once stricturing has progressed to cause significant cholestasis. Laboratory evaluation most often reveals an elevated alkaline phosphatase in early stages with hyperbilirubinemia reflecting more advanced disease. Liver biopsy can help support the diagnosis of PSC, especially in cases limited to smaller ducts, yet is not regularly required in clinical practice. Chronic biliary inflammation and cholestasis can progress to cirrhosis, and a significant number of patients will eventually require liver transplantation. The estimated 10-year survival in PSC patients is approximately 65 % amidst great variability [12] with a reported incidence of cholangiocarcinoma of 7–9 % over the same time period [13, 14].

Aside from variants of cholangiocarcinoma, other conditions may mimic PSC. These include post-operative biliary stricture as mentioned earlier, chronic choledocholithiasis , sequelae of intra-arterial chemotherapy, and even metastatic disease. Mast cell cholangiopathy and eosinophilic cholangitis have also been reported [15, 16]. In these two conditions, where underlying systemic disease is commonly observed, respective mast cell and eosinophilic infiltration of ductal epithelium leads to a fibroinflammatory response that can progress to biliary stricturing. Radiographically, both may appear similar to PSC, thus stressing the need for histology. Another similar benign condition that may mimic both extrahepatic PSC and cholangiocarcinoma is fibroinflammatory biliary stricture (FIBS). FIBS is a rare myofibroblastic lesion with a distinct profile on immunohistochemistry and is frequently diagnosed following extensive resection for presumed cholangiocarcinoma. An increasingly recognized cause of secondary sclerosing cholangitis is IgG4-related disease , referred to here as IgG4-associated cholangitis (IAC). In a 2001 paper by Hamano et al. high serum levels of IgG4 were detected in patients with sclerosing pancreatitis and thus emerged a new classification for autoimmune pancreaticobiliary disease [17]. There have been subsequent reports of IgG4-related disease in a variety of other organs, including both lacrimal and salivary gland involvement [18, 19].

The majority of IAC cases are associated with autoimmune pancreatitis (AIP) (92.5 %), although isolated cases of cholangiopathy have been observed [20]. In IAC, parenchymal infiltration by IgG4-positive plasma cells can progress to tissue fibrosis and obliterative phlebitis. Patients are typically older on presentation than those with PSC and most often demonstrate obstructive jaundice. Other features can include weight loss, abdominal pain, and new onset diabetes. Laboratory evaluation typically reveals elevated liver enzymes in a cholestatic pattern with or without abnormal lipase/amylase. Unlike PSC, bilirubin is frequently markedly elevated on diagnosis . The hallmark feature of IAC is a significantly elevated serum IgG4 level ( 300 mg/dL), which occurs in 50–80 % of patients and may portend a favorable response to corticosteroid therapy [17, 21]. It is important to recognize that serum IgG4 is also moderately elevated in 9 % of PSC patients [22]. PSC cases with IgG4 plasma cell infiltration have been described with one study reporting greater than 10 IgG4 + cells per HPF in 23 % of explanted livers with PSC [23]. Inflammation was rarely observed surrounding large ducts and other histologic features were more consistent with PSC, despite higher-than-expected levels of serum IgG4. This suggests that serum IgG4 level must be interpreted with caution, particularly when only mildly-to-moderately elevated. Elevated levels should be analyzed in the context of other clinical features and findings in order to more accurately distinguish IAC from both PSC and cholangiocarcinoma. Perinuclear anti-neutrophil cytoplasmic antibody (p-ANCA) positivity may help differentiate PSC from IAC as p-ANCA occurs in up to 80 % of PSC patients, even in the absence of underlying inflammatory bowel disease [11]. At this time, other autoimmune markers have little role in diagnosing PSC or IAC. In the setting of cholestasis, mild-to-moderate elevations in CA 19-9 can be nonspecific; thus, further evaluation is warranted to exclude underlying malignancy , particularly if levels remain persistently or markedly elevated.

Abdominal ultrasound and contrast-enhanced CT are often the initial modalities for hepatobiliary imaging. Findings of PSC can be variable and inherently depend on the disease state, ranging from minimal radiologic findings to significant biliary stricturing/dilatation and even cirrhosis. Typically demonstrated on MRCP or ERCP are multifocal stenoses of both intra and extrahepatic bile ducts with intervening normal-appearing or mildly dilated segments to produce a “beaded” ductal appearance. Although high-quality MRCP has good performance characteristics for diagnosing or excluding PSC, ERCP remains the gold standard for the diagnosis of PSC. Various studies on MRCP report a sensitivity of 80–88 % in experienced centers [24–27]. Specificity was initially reported in the range of 92–98 % [17–19], although recent studies have produced more modest results [20]. While both modalities have demonstrated good interobserver agreement on the degree of intrahepatic stricturing, ERCP exhibits better agreement for extrahepatic strictures [20], particularly those that may warrant endotherapy. At present, MRCP remains suboptimal for predicting need for endotherapy.

Due to the high sensitivity and specificity of MRCP, noninvasive imaging may be sufficient for diagnosis in patients with little concern for underlying malignancy or need for endotherapy. ERCP is generally reserved for patients who cannot undergo MRCP, have nondiagnostic MRCP, or demonstrate dominant stenoses with upstream ductal dilatation. In the latter situation, excluding underlying cholangiocarcinoma with targeted sampling and providing biliary decompression remain the principal reasons for ERCP.

An important radiologic finding in PSC with considerable implication on disease management is the dominant stricture (Fig. 9.2). The definition of dominant strictures in PSC within the literature is variable. Repeated attempts have been made to define dominant strictures in terms of their location within the biliary tree and their respective ductal diameters. At our center, we feel dominant strictures can occur anywhere in the biliary tree and that the status of the upstream bile duct is equally important as the nature of the stricture in determining which strictures warrant therapy. Longitudinal studies have suggested that 33–50 % of PSC patients develop a dominant stricture over 8–13 years, stressing the need for improved detection and management [28, 29]. While data have yet to establish specific risk factors for the development of dominant strictures, an association with stage 2–4 fibrosing inflammation on liver biopsy has been reported [29].

A major concern with all dominant strictures is the potential for underlying cholangiocarcinoma. Excluding malignancy can be challenging with MRI and ERCP, especially given the aforementioned limited sensitivity of ERCP biopsies/brushings and unreliable tumor markers in cholestatic disease [30–32]. Greater sensitivity (~ 80 %) has been demonstrated with higher cut-off points for CA 19−9 (≥ 130 U/mL) in PSC patients [30]. A study of 333 patients with PSC performed at our center reported an even higher sensitivity (100 %) and nearly 80 % specificity for the diagnosis of cholangiocarcinoma when standard brush cytology was combined with abnormal CA 19-9 ( 180 U/mL) and CEA ( 5.2 ng/mL) levels [32]. More data are required to evaluate the cost-effectiveness of such a strategy and proper screening intervals for symptomatic and asymptomatic PSC. Newer modalities, including 3-dimensional MRI with bile duct reconstruction and direct peroral cholangioscopy with or without confocal endomicroscopy, are being investigated to improve detection of malignancy in dominant strictures.

Similar to PSC, IAC typically affects large bile ducts and can demonstrate diffuse areas of stricturing with either normal or dilated intervening segments on cholangiography. Differentiating between PSC and IAC with current imaging techniques can be challenging (Fig. 9.3). Nishino et al. report a higher incidence of distal bile duct strictures and a more segmental (less beaded) appearance in IAC compared to PSC [33]. Significant intrahepatic stricturing is rarer (less than 10 %), yet when present, can be diffuse in IAC. Concomitant pancreatic duct strictures are frequently encountered. There have even been reports of IgG4-related disease causing inflammatory mass-like lesions and subsequent biliary obstruction [34, 35].

Histology obtained during ERCP may help diagnose IAC. One study reports modest sensitivity (52 %), despite excellent specificity (97 %), for diagnosing IAC from ampullary and intraductal biopsies [36]. When diagnostic, histopathology is most frequently characterized by the typical lymphoplasmacytic infiltrate and transmural fibrosis. Positive stains for IgG4 are further suggestive of IAC. Unlike in PSC, onion-skinned concentric fibrosis is not observed within peri-ductal regions.

Newer techniques, such as cholangioscopy and intraductal ultrasonography (IDUS), have also been investigated in IgG4-related disease, yet their precise roles remain to be defined. IDUS has demonstrated more concentrically thickened bile duct walls with smoother margins in IAC compared to cholangiocarcinoma [37]. In the presence of concomitant pancreatic disease, a core pancreatic biopsy under endoscopic ultrasound guidance can be useful.

MRCP reveals moderate dilation of intra and extrahepatic bile ducts with an area of extrahepatic abnormal soft tissue surrounding the main portal vein. A common hepatic duct stricture is noted with associated upstream dilatation. There is no obvious pancreatic head mass. No choledocholithiasis or cholelithiasis is noted. An ERCP is subsequently performed without complication (Fig. 9.4). The patient demonstrates serologic improvement in liver function tests following stent placement. Intraductal biopsies and brushings of the common hepatic duct stricture return negative for malignancy. Serum IgG4 is normal at 87. An endoscopic ultrasound is subsequently performed to obtain tissue from the porta hepatis lesion. Cytopathology results demonstrate mildly atypical epithelial cells without evidence of malignancy.

The interpretation of findings in the context of suspected autoimmune pancreaticobiliary disease has led to the development of various consensus criteria for its diagnosis. Using the HISORt (histology, imaging, serology, organ involvement, response to therapy) criteria, the diagnosis of IAC requires intrahepatic and/or proximal extrahepatic bile duct stricturing in the context of (1) typical radiologic findings of AIP (encapsulated, sausage-shaped pancreas), (2) supportive radiologic findings of AIP (focal pancreatic enlargement) with abnormal pancreatic ductal imaging and elevated serum IgG4 or other organ involvement (retroperitoneal fibrosis or 10 IgG4+ cells per hpf from an intraductal or ampullary biopsy obtained during ERCP), (3) clinical and serologic response to steroid therapy with supportive radiologic findings, or (4) histology consistent with AIP from core pancreatic biopsy during EUS [38]. A more recent Japanese consensus criterion has since emerged [39]. Validation of both guidelines will be warranted in future studies.

Given that differentiating between PSC and IAC can be difficult by cholangiography alone, diagnosis largely rests on the patient’s history and clinical picture. As previously mentioned, PSC patients tend to be younger males with evidence of mild cholestasis and concomitant inflammatory bowel disease. Rarely do they demonstrate evidence of associated pancreatic disease with the most common cause for this is medication-related sequelae from azathioprine. In older patients without underlying inflammatory bowel disease , IAC receives greater consideration, particularly when evaluation for suspected cholangiocarcinoma is negative on repeat endoscopic examinations. At our center, serum IgG4 levels are not routinely checked in patients with suspected PSC because mild elevations can be observed and may not be clinically relevant. When concomitant autoimmune disease or a family history thereof is observed, serum IgG4 levels and biopsies for IgG4 staining are pursued. In the small subset of patients with concomitant pancreatic disease and/or mass-like lesions that are persistently negative for malignancy on histology, IAC becomes more plausible and thus an empiric trial of steroids is considered prior to advocating for surgical resection.

Management options in PSC have included both medical and endoscopic therapies in an effort to prolong (and hopefully obviate) the need for liver transplantation. The American Association for the Study of Liver Diseases (AASLD) currently recommends against the use of ursodiol in managing PSC patients given unclear benefit on transplant-free survival [40]. More recent data from Scandinavia recommend ursodiol (17–23 mg/kg/day) in patients who respond with normalization or significant decreases (≥ 40 % reduction) in alkaline phosphatase, which was associated with increased survival [41]. Although associated with improvement in liver function testing, the effect of ursodiol on disease progression remains uncertain. Large multi-center studies have investigated both low-dose (10–15 mg/kg/d) and high-dose (17–30 mg/kg/d) ursodiol in PSC, yet neither has been shown to improve histology or lengthen time to liver transplantation [42–44]. In fact, one multicenter trial found high-dose (28–30 mg/kg/d) ursodiol associated with increased mortality and shorter time to transplantation [44]. Recent studies have suggested a possible chemoprotective effect of ursodiol in reducing the risk of colon cancer in patients with concomitant inflammatory bowel disease [45, 46]. Such an effect has yet to be demonstrated for cholangiocarcinoma. Validation of the limited evidence is necessary prior to recommending ursodiol for chemoprevention. Immunomodulator and steroid therapy have also been trialed without demonstrated benefit [47], except in overlap syndromes of PSC and autoimmune hepatitis and/or IgG4+ infiltrate on liver biopsy [48, 49].

Acute episodes of cholangitis are treated with usual antibiotic therapy targeted at enteric pathogens, particularly gram-negative and anaerobic organisms. In patients who suffer from repeated episodes of cholangitis, suppressive antibiotic therapy may be considered as a bridge to liver transplantation. In these patients, a Model for End Stage Liver Disease (MELD) score exception may be granted and more urgent transplantation indicated.

Endoscopic management of PSC usually occurs later in the disease and relies heavily on the endoscopist to accurately recognize a dominant stricture. Endotherapy with balloon dilation and/or stenting is indicated to alleviate biliary outflow obstruction and possible resultant hepatocellular injury. Since the original report of endoscopic decompression in PSC in 1982 [50], multiple studies (mostly retrospective) have investigated the role of balloon dilation and stenting in this patient population [51–54]. One prospective study by Stiehl et al demonstrated improved survival following endoscopic management of dominant strictures compared to the predicted survival derived from the Mayo Clinic survival model prior to ERCP in 106 PSC patients on ursodiol [29]. However, the Mayo Clinic model was not validated in highly selected PSC patients with acute obstructive jaundice. Because it is heavily weighted to the serum bilirubin level, it should be applied to patients immediately after endoscopic interventions and then followed long term. Regardless of these limitations, improved patient outcomes can be seen in selected patients with obstructive jaundice and PSC following endoscopic intervention [29, 55, 56].

In order to access the biliary tree and facilitate techniques for biliary evaluation and decompression, a small sphincterotomy is often performed. Following cannulation of the bile duct , recognition of potentially treatable strictures becomes of paramount importance. Short-segment (≤ 2 cm long) dominant strictures of the common bile duct (CBD) were the first to be successfully treated endoscopically. Subsequent efforts over the years have investigated the role of endotherapy in long-segment CBD strictures and strictures within distal hepatic ducts. In a prospective study of 171 patients on ursodiol followed for 20 years, 96 patients underwent endotherapy for either dominant strictures or cholangitis [57]. Both CBD strictures greater than 2 cm in length and short strictures within 2 cm of the bifurcation may be effectively treated with dilation and/or stenting. Significant stricturing of the intrahepatic system proximal to this 2-cm mark may be difficult to treat endoscopically and thus may warrant more prompt transplant referral.

The dilemma of balloon dilation alone versus dilation with placement of a “short-interval” biliary stent continues to be better defined. Longer-term stent placement has been associated with a high complication rate around 20 % [56]. Debris can occlude biliary stents and the stents can block opposing peripheral bile ducts. These complications can inevitably lead to cholestasis and sepsis. For these reasons, a “short-interval” biliary stent is generally recommended. In a retrospective study of 32 patients treated with biliary stents removed after a mean of 11 days, clinical improvement was observed in 80 % of the population [58]. Perhaps even more significant, 60 % of patients did not require repeat intervention at 3-year follow-up. This suggests that short-term stent placement may yield the desired clinical and biochemical response.

Balloon dilation remains a mainstay of endotherapy in PSC, as it allows for graded dilation of the biliary tree. The balloon diameter must match the smaller diameter of the downstream bile duct to avoid iatrogenic injury . Once the balloon is inflated, pressure is maintained for a period of 30–60 s in order to achieve adequate response. Some advocate the use of short-term stents following balloon dilation while others try to avoid stent placement all together. At our center, 10 French plastic stents are uniformly placed for patients with episodes of acute bacterial cholangitis following therapy with balloon dilation. Patients are instructed to return for reevaluation and stent removal in 2 weeks. In patients with evidence of recurrent cholangitis ( 3 episodes) and/or progressive stricturing that becomes dependent on endotherapy to alleviate persistent obstruction, transplant referral is once again the most appropriate next step in management. Surgical reconstruction prior to liver transplantation leads to worse outcomes and should be avoided if possible.

Peri-procedural antibiotic prophylaxis is universally recommended, as mere injection of contrast during the procedure can result in sepsis and/or infection within more peripheral bile ducts. Patients are generally given ciprofloxacin 500 mg twice daily for 3–5 days following the procedure with longer durations (7–14 days) advocated in those with systemic constitutional symptoms or evidence of pus on endoscopic evaluation. The complication rate of ERCP in PSC patients is suspected to be slightly higher than in patients with CBD stones or solitary malignant strictures . Clinically significant complications have been observed in 7–14 % of PSC patients [29, 53, 59, 60], mostly due to increased rates of cholangitis. Other series have reported even higher complication rates, although these findings may be attributable to longer and/or greater number of stent placement prior to removal.

Unlike PSC, where steroids have no reported benefit, the first line of therapy in IAC is an extended taper of prednisone starting at 40 mg daily. This is usually tapered by 5 mg weekly, after which a maintenance dose of 5 mg daily for 3–6 months is recommended. Azathioprine, mycophenolate mofetil, and rituximab have been utilized as maintenance regimens in steroid-resistant disease [20, 61, 62]. Lab parameters (specifically liver function tests, glucose, and IgG4 levels) should be followed throughout the treatment period. In patients who demonstrate clinical worsening and/or laboratory deterioration while on corticosteroids, therapy should be tapered and further evaluation for malignancy should ensue. Biliary stents are frequently placed for interval decompression during initial steroid treatment and often removed after 1–2 months. Repeated dilations and stenting may be warranted in refractory disease. Peri-procedural antibiotic prophylaxis is generally recommended, particularly in the setting of hilar strictures and intrahepatic involvement.

Given suspicion for IAC and lack of evidence of malignancy, the patient is started on an extended course of steroids. He is followed closely and continued to demonstrate both clinical and serologic improvement while on steroid therapy. Repeat imaging is performed 1 month following initiation of steroid therapy (Fig. 9.5). Repeat ERCP reassesses the disease process following a course of steroids for possible stent removal/exchange (Fig. 9.6).

A 55-year-old male with a past medical history significant for hepatitis C cirrhosis status post orthotopic liver transplantation 2 years ago presents for evaluation of abnormal liver function tests. His operative course was uncomplicated and the patient has had no issues following his transplant. He denies jaundice, abdominal pain, and fever. Vital signs are stable and physical exam is largely unremarkable. Laboratory evaluation reveals the following: total bilirubin 2.9, ALT 202, AST 121, and alkaline phosphatase 1000. CBC and coagulation profile are within normal limits. An abdominal ultrasound with Doppler shows patent vessels and no evidence of ductal dilation. An MRCP is subsequently ordered.

Liver transplantation has become a principal treatment modality in eligible patients for a variety of conditions; the most common remain end-stage liver disease , hepatocellular carcinoma (HCC), and acute liver failure. A donor-to-recipient choledocho-choledochostomy is created in 75–88 % of liver transplants in the United States [63–65], with the remaining cases being a Roux-en-Y choledochojejunostomy (primarily for PSC and pediatric cases). Biliary complications are frequently encountered following liver transplantation, affecting 5–25 % of the post-transplant population [66–68]. Early complications occur within 30 days of transplant and are often attributed to surgical manipulation proper or vascular compromise. The most common early complications include bile leaks and extrahepatic biliary obstruction. While the latter is frequently a consequence of post-surgical edema/inflammation, other etiologies include anastomotic narrowing, vascular insufficiency, and even torsion. Late complications generally occur after 1 month following liver transplantation. Ductopenic rejection and hepatic artery thrombosis are two important diagnostic possibilities to consider in this population; both may be suggested by cholestasis alone. Other possible etiologies for late cholestasis include recurrent liver disease, choledocholithiasis , biliary cast syndrome, and/or post-transplant ampullary dysfunction. Similar to ischemic cholangiopathy that is associated with the development of hepatic artery thrombosis following transplant, biliary cast syndrome is mostly observed in patients with some degree of vascular insufficiency and is more frequent in those with hilar strictures. This uncommon phenomenon, which may be related to both intra-operative ischemia times and/or the presence of concomitant acute cellular rejection, often requires repeat attempts at endoscopic therapy (sphincterotomy and balloon sweeping) to achieve ductal clearance. Another late etiology for cholestasis is post-transplant ampullary dysfunction. These patients typically lack the pain associated with sphincter of Oddi dysfunction , presumably due to a denervated biliary tree. Filling defects are not observed on cholangiography and the degree of ductal dilation is typically uniform upstream of the papillary orifice. Efficient treatment is achieved with biliary sphincterotomy . Given the observed variability in the onset of biliary diseases following liver transplantation, recognition of temporal relationships can have a profound impact on patient morbidity and mortality.