Introduction

Primary sclerosing cholangitis (PSC) is a chronic cholestatic syndrome of unknown etiology characterized by fibrosing inflammatory destruction of the intra- and extrahepatic bile ducts. It is frequently progressive and may be fatal in the absence of liver transplantation. Inflammatory bowel disease (IBD) is associated with PSC in approximately 70 % of patients [1•] and frequently precedes clinical recognition of hepatobiliary disease.

Death from cancer occurs in a significant subset of patients with PSC. Although cholangiocarcinoma (CCA) is the cancer most frequently associated with PSC, other cancers are recognized as part of the disease spectrum. In a study of 200 PSC patients followed for over a decade in Belgium, CCA developed in thirteen patients, colorectal cancer developed in ten, hepatocellular carcinoma developed in three, and pancreatic cancer developed in five. In all, 61 patients died, 28 of them from cancer [2•]. In another large cohort of 604 Swedish patients with PSC with a median follow-up of 5.7 years (range 0–27.8 years), 171 patients died, 44 % of them from cancer. The risk of CCA was 160-fold greater than in the general population, and 37 % of CCAs were diagnosed less than 1 year after the diagnosis of PSC [3]. A retrospective analysis of more than 500 patients with PSC in Germany found that cholangiocarcinoma accounted for 75 % of all tumors with 20 % of patients receiving the diagnosis at first presentation or within the first year of diagnosis of PSC [4•]. This study also reported a very low incidence of hepatocellular carcinoma (HCC) in patients with PSC.

A second major concern in patients with PSC is colorectal cancer (CRC). Patients with PSC-IBD have a 4-fold higher risk of CRC than patients with IBD alone and a 10-fold higher risk of colorectal cancer than the general population [5]. A recent analysis of the English literature reported a risk of CRC in up to 30 % of patients followed for 20 years after the diagnosis of PSC-IBD [1•]. The authors also pointed out that the pathogenesis and optimal screening of these patients remain uncertain.

Although many issues remain unresolved, the following review will give an overview of common malignancies seen in this patient population and will suggest reasonable cancer surveillance strategies for PSC patients.

Cholangiocarcinoma

Patients with PSC have a 10 to 15 % lifetime risk of developing CCA, i.e., a 398-fold increased risk for developing CCA compared to the general population [6•]. Unlike hepatocellular carcinoma, CCA may occur in the absence of advanced hepatic fibrosis, and the incidence increases steadily after the first year, with no late rise in frequency [7]. In fact, many patients are diagnosed with CCA early in their fourth decade of life [2•, 6•]. In a series of 161 patients from the Mayo Clinic, 7 % developed CCA during a mean follow-up period of 11.5 years, and no association was found between the duration of PSC and the incidence of CCA [7].

A substantial portion of patients with CCA are diagnosed soon after their PSC diagnosis with some studies reporting 30 % in the first year after PSC diagnosis [2•]. This finding suggests that superimposed CCA may have elicited the symptoms leading to the diagnosis of PSC.

Cholangiocarcinoma has a dismal prognosis with some reporting a median survival as low as 5 months after diagnosis [8]. A large Dutch population-based study identified 590 patients with PSC with a median follow-up of 92 months. CCA occurred in 41 patients (7 %) of whom 33 died (80 %) after a median period of 1 year. The cumulative risk of CCA after 10, 20, and 30 years was 6, 14 and 20 %, respectively, contradicting the notion of time-independent risk reported above [6•].

The diagnosis of cholangiocarcinoma is usually based on the appearance of symptoms, commonly the onset of jaundice. These patients typically have unresectable tumors and a very bad prognosis -only 10 % survive past 2 years [9]. Distinguishing between a benign and a malignant dominant stricture is challenging in patients with PSC. A single tertiary center in London reported their 25-year experience with PSC-related dominant strictures. A total of 128 patients with PSC were followed for a mean of 9.8 years. The mean survival of those with dominant strictures (13.7 years) was worse than those without dominant strictures (23 years). Much of the survival difference was related to a 26 % risk of cholangiocarcinoma developing in those with dominant strictures. Nearly half of CCAs were diagnosed within 4 months of the diagnosis of PSC. The mean CA 19-9 was 8236 IU/L in those with CCA and 97 IU/L with benign dominant strictures [10•].

The efficacy of tumor markers for early diagnosis of CCA in patients with PSC has not been established. A recent study of 208 patients found a sensitivity and specificity of 78 and 98 %, respectively, using a CA 19-9 value of 129 U/mL as a cutoff value [11]. However, more than one third with this cutoff did not have cholangiocarcinoma after an average period of 30 months of follow-up [12•]. CA 19-9 also requires the presence of the Lewis blood group antigen. Individuals with a negative Lewis phenotype (estimated at 5 to 10 % of the population) do not make CA 19-9 and will not benefit from this testing.

For patients requiring endoscopic retrograde cholangiopancreatography (ERCP), routine brush cytology analysis alone detects CCA with low levels of sensitivity (40 % or less), but specificity is nearly 100 % [13]. Fluorescence in situ hybridization (FISH) may be more sensitive and as specific as cytology [14]. Among patients with PSC and a dominant stricture, the presence of polysomy identified those with an eventual diagnosis of CCA with 88 % specificity [15].

The American Association for the Study of Liver Diseases (AASLD) does not recommend routine cholangiocarcinoma surveillance in PSC patients except for those with deteriorating constitutional performance status or liver biochemical tests [13]. This nihilistic approach is based on the absence of proven benefit of early detection and surgical or medical management of patients with CCA. Since late CCA is rapidly fatal, we recommend combining serum CA 19-9 measurements with imaging to improve our ability to detect worrisome extrahepatic strictures that could be cholangiocarcinoma. ERCP with brush cytology and FISH should be reserved for patients with dominant strictures, a rising CA 19-9, or an increase in cholestatic liver tests.

Colorectal Carcinoma

PSC is associated with inflammatory bowel disease (IBD) in approximately 70 % of patients [1•], and patients with ulcerative colitis (UC) and PSC are at increased risk of colorectal dysplasia and carcinomas compared with UC alone, with OR 4.79 (95 % CI 3.458–6.41) [5]. Some data suggest the risk of CRC development can reach up to 30 % at 20 years after a diagnosis of concurrent IBD and PSC [16].

A large population-based study in the Netherlands followed 590 patients with PSC with a median follow-up of over 7 years. Twenty of 590 patients developed colorectal carcinoma (CRC), and all but one had a diagnosis of inflammatory bowel disease. There was a 5-fold increased risk for developing CRC in all PSC patients compared to age and gender-matched general population and a 9-fold increased risk compared to UC patients alone. CRC occurred at a much younger age in PSC-IBD patients compared to IBD controls (median age 39 compared to 59) [6•]. The cumulative risk of CRC or dysplasia was 4.1 % at 5 years, 5.4 % at 10 years, and 9.3 % at 20 years after the detection of PSC [2•]. Given the increased risk of CRC in patients with PSC, the AASLD recommends surveillance colonoscopy with biopsies at 1- to 2-year intervals from the time of diagnosis of PSC in patients with concomitant IBD and an initial full colonoscopy with biopsies for those with a new diagnosis of PSC in the absence of a previous diagnosis of IBD [13].

The risk of active colitis and colon cancer do not decrease after liver transplantation. Patients with PSC-IBD have a high risk of de novo CRC after OLT, approximately ten times the risk of all other patients undergoing OLT [17•] and up to 20-fold higher than risk in the general population [18]. According to an analysis of 439 patients with PSC who underwent OLT between 1984 and 2006 in the Nordic OLT registry, the risk of CRC was higher after OLT than before with a hazard ratio of 1.9, not entirely explained by the duration of IBD [19•]. In contrast, Hanouneh et al. from the Cleveland Clinic performed a case-control analysis of patients with PSC and IBD who underwent OLT (n = 43) compared to those who did not receive a transplant. The rate of colonic neoplasia was similar in the IBD/PSC/OLT group compared to the IBD/PSC/no-OLT group (34 % versus 30 %, p = 0.24) but higher in PSC patients compared to those with other forms of chronic liver disease (34 % versus 0 %, p = 0.018) [20•]. The development of post-OLT colonic neoplasia did not translate to decreased survival post-transplant in patients with IBD and PSC.

The use of ursodeoxycholic acid (UDCA) as a chemopreventive agent in patients with PSC and IBD is controversial with some studies showing benefit [21, 22] and others showing no difference [23]. A meta-analysis of eight studies including 177 cases of colorectal neoplasia in 763 patients with PSC-IBD reported a significant chemopreventive effect of UDCA on the risk of colorectal cancer and/or high-grade dysplasia with OR of 0.35 (95 % CI 0.17–0.73) [24•]. A subgroup analysis showed low-dose UDCA use (8 to 15 mg/kg/day) was also associated with a significant risk reduction of colorectal neoplasia. A nested cohort study from 2011 demonstrated an increased risk of colorectal cancer associated with high-dose UDCA in PSC patients [25•]. Fifty-six subjects with PSC and UC were followed for a total of 235 patient years. Patients who received high-dose UDCA (28–30 mg/kg/day) had a significantly higher risk of developing colorectal neoplasia (dysplasia and cancer) compared to the placebo group (hazard ratio 4.4, CI 1.30–20.1, p = 0.02) [25•]. AASLD recommends against the use of UDCA in all patients with PSC including those with PSC-IBD [13] whereas EASL (European Association for the Study of the Liver) recommends UDCA in high-risk groups such as those with a strong family history of CRC, previous colorectal dysplasia, or long-standing extensive colitis [26].

Given the increased risk of colorectal cancer in patients with PSC and IBD, we recommend surveillance with colonoscopies starting at the time of diagnosis of PSC with repeat colonoscopy every 1 to 2 years as recommended by both EASL and AASLD. Surveillance colonoscopy should continue after liver transplantation because the risk of colorectal cancer appears to be unaffected or worsened by OLT. The data regarding the chemopreventive effects of UDCA are mixed; we do not recommend the routine use of UDCA for chemoprevention in PSC patients.

Gallbladder Carcinoma

The prevalence of gallbladder carcinoma in patients with PSC is estimated at 3 to 14 % [27•]. Gallbladder abnormalities were assessed in 286 patients with PSC treated in Sweden between 1970 and 2005. One or more gallbladder abnormalities were found in 41 % of the patients. A gallbladder mass lesion, mean size of 21 mm, was found in 18 patients (6 %)—gallbladder carcinoma was confirmed in ten (56 %) of these patients. Histological reevaluation disclosed epithelial dysplasia of the gallbladder in nine patients without a gallbladder mass lesion [28].

In another review of 102 patients with PSC who underwent cholecystectomy at the Mayo Clinic between 1977 and 1999, a gallbladder mass was found in 14 patients (14 %). Eight of these had adenocarcinoma (57 %) and six had benign masses. In those with benign masses, 33 % had associated epithelial cell dysplasia. Only seven of the mass lesions were recognized preoperatively, and no significant difference in size was found between malignant or benign lesions.

The high frequency of carcinoma in gallbladder mass lesions indicates that surveillance with regular ultrasound may be of value in these patients. AASLD guidelines from 2010 recommend an annual ultrasound to detect mass lesions in the gallbladder and cholecystectomy for any size polyps in patients with PSC [13]. However, a second retrospective review of PSC patients who underwent cholecystectomy from the Mayo Clinic argues that observation may be appropriate for gallbladder polyps <0.8 cm. The most common indication for cholecystectomy in this 57 patient cohort was gallbladder polyp or mass. Sensitivity and specificity of a gallbladder lesion of 0.8 cm and the presence of gallbladder neoplasia was 100 and 70 %, respectively. Twenty-three patients (40 %) had an early post-operative complication and nearly 35 % of those had more than one complication. Child-Pugh score was the only predictor of post-operative outcomes in the multivariate model [29]. The authors concluded that gallbladder polyps <0.8 cm are unlikely to be malignant, and observation may be appropriate in the absence of other concerning features such as new or enlarging lesions, large masses, adjacent wall thickening, sessile morphology, or an arterial signal on Doppler.

We agree with the AASLD and recommend annual screening for gallbladder cancer with ultrasound.

Hepatocellular Carcinoma

The incidence of HCC in patients with PSC cirrhosis has not been established, but many experts believe it to be below the 1.5 % per year threshold required to institute HCC surveillance. A German retrospective study analyzed the risk of hepatocellular carcinoma (HCC) in 509 patients with PSC, 119 of whom had cirrhosis. Data were collected from periods up to 33 years with a total of 4202 patient-years included in the final analysis. In 292 patient-years under cirrhosis risk, not a single HCC was detected. In reviewing 140 liver explants, no incidental HCC was detected in the series. In contrast to the low numbers of HCC, 35 patients developed cholangiocarcinoma, 3 patients developed gallbladder cancer, and 9 patients developed colorectal cancer [4•]. These data suggest the risk of HCC, even in the presence of cirrhosis, is lower than the limit generally accepted for regular surveillance examinations using ultrasound. We are concerned that a study of 119 patients with PSC and cirrhosis is too small a cohort to change our HCC screening practices. AASLD has no guidelines on HCC screening in patients with PSC and cirrhosis since no supporting evidence exists. We screen cirrhotic patients with PSC with imaging every 6 months—HCC screening is addressed by the same images.

Conclusion

Primary sclerosing cholangitis (PSC) is a chronic, progressive cholestatic syndrome with severe complications that frequently lead to serious health issues, liver transplantation, and/or death. Inflammatory bowel disease (IBD) is associated with PSC in approximately 70 % of patients [1•] and often precedes clinical recognition of hepatobiliary disease. Death from cancer occurs frequently in patients with PSC, and many studies have shown malignancy to be the most common cause of death.

Compared to the general population, the risk of cholangiocarcinoma, gallbladder cancer, and colorectal cancer (specifically in patients with PSC-IBD) is high. Despite this, the value of screening for various cancers has not been adequately studied. As a practical guide, and in the absence of definitive studies, we recommend MRI every 6 months with CA19-9 and AFP to screen for HCC and CCA. We also recommend ultrasound of the gallbladder every year to screen for gallbladder masses since ultrasound is more sensitive than MRI for this diagnosis. Finally, we recommend routine surveillance for CRC in patients with inflammatory bowel disease and a screening colonoscopy to exclude ulcerative colitis in all patients with newly diagnosed PSC.