The Natural History of Small-Duct Primary Sclerosing Cholangitis

Transcription

1 GASTROENTEROLOGY 2008;134: The Natural History of Small-Duct Primary Sclerosing Cholangitis EINAR BJÖRNSSON,* ROLF OLSSON,* ANNIKA BERGQUIST, STEFAN LINDGREN, BARBARA BRADEN, ROGER W. CHAPMAN, KIRSTEN M. BOBERG, and PAUL ANGULO # *Department of Internal Medicine, Section of Gastroenterology and Hepatology, Sahlgrenska University Hospital, Gothenburg, Sweden; Karolinska University Hospital, Stockholm, Sweden; Malmo University Hospital, Malmo, Sweden; John Radcliffe Hospital, Oxford, United Kingdom; Rikshospitalet, Oslo, Norway; and # Department of Medicine, Miles and Shirley Fiterman Center for Digestive Diseases, Division of Gastroenterology & Hepatology, Mayo Clinic, Rochester, Minnesota Background & Aims: The long-term prognosis of patients with small-duct primary sclerosing cholangitis (PSC) remains incompletely characterized. We aimed at determining the natural history and long-term outcomes of a large number of patients with small-duct PSC. Methods: Data from 83 patients with well-characterized small-duct PSC from several medical institutions in Europe and the United States were combined. Each patient with small-duct PSC was randomly matched to 2 patients with large-duct PSC by age, gender, calendar year of diagnosis, and institution. Results: The median age at diagnosis in both groups was 38 years (61% males). Nineteen (22.9%) of the 83 patients with small-duct PSC progressed to large-duct PSC in a median of 7.4 (interquartile range [IQR], ) years. One patient with small-duct PSC who progressed to large-duct PSC was diagnosed with cholangiocarcinoma but after progression to largeduct PSC; 20 patients with large-duct PSC developed cholangiocarcinoma. Patients with small-duct PSC had a significantly longer transplantation-free survival compared with large-duct PSC patients (13 years [IQR, 10 17] vs 10 years [IQR, 6 14], respectively; hazard ratio, 3.04; 95% confidence interval: ; P <.0001). Two patients with small-duct PSC who underwent liver transplantation had recurrence of small-duct PSC in the graft 9 and 13 years, respectively, after transplantation. Conclusions: Smallduct PSC is a disease of progressive potential but associated with a better long-term prognosis as compared with large-duct PSC. Small-duct PSC may recur after liver transplantation. Cholangiocarcinoma does not seem to occur in patients with small-duct PSC, unless the disease has progressed to large-duct PSC. The natural history of classic (ie, cholangiography confirmed) large-duct primary sclerosing cholangitis (PSC) has been well characterized in many long-term follow-up studies. 1 7 In cohorts of patients with PSC diagnosed mostly in the 1970s and 1980s, a median transplantation-free survival of approximately 12 years was observed. 2,3,5 However, a Dutch cohort that also included patients from 1990 to 1999 reported a median transplantation-free survival of 18 years. 7 Cholangiocarcinoma, the most feared complication of PSC, develops in up to one third of patients with PSC. 8 A minority of patients with sclerosing cholangitis of unknown etiology with similar cholestatic and histologic features as those with classic PSC have normal cholangiograms, and they have been referred to as small-duct PSC. 9 This subgroup of PSC has been less well characterized than the classic large-duct PSC. Three studies analyzing the long-term prognosis of patients with small-duct PSC were published in The results from these 3 different studies suggested a better prognosis in patients with small-duct than in large-duct PSC Subsequently, only 1 other series of 6 patients with small-duct PSC with a limited follow-up has been published. 13 All these previous studies, however, have included limited numbers of patients with a relatively short follow-up. Thus, those studies had minimal or no power for an appropriate analysis of long-term survival. In addition, because small-duct PSC is a condition affecting only microscopic bile ducts (ie, bile ducts seen only on liver biopsy), it remains uncertain whether cholangiocarcinoma develops in small-duct PSC and whether liverrelated morbidity and liver-related mortality in patients with small-duct PSC differ from that seen in patients with large-duct PSC. Hence, we aimed at determining the natural history and long-term prognosis of a large number of patients with well-characterized small-duct PSC. In the current investigation, we extended the follow-up of patients with small-duct PSC reported in those 3 series We sought to determine the outcomes and longterm survival of this series of patients with small-duct PSC and compare their outcomes and survival to that seen in a group of appropriately matched patients with classic, large-duct PSC. Materials and Methods The results presented in the current study originate from cohorts of small-duct PSC patients from Nor- Abbreviations used in this paper: IQR, interquartile range; MRC, magnetic resonance cholangiography; PSC, primary sclerosing cholangitis; UDCA, ursodeoxycholic acid by the AGA Institute /08/$34.00 doi: /j.gastro

2 976 BJÖRNSSON ET AL GASTROENTEROLOGY Vol. 134, No. 4 way, Sweden, the United Kingdom, and the United States The criteria for identification of these patients; the diagnostic criteria; and the clinical, biochemical, and histologic variables assessed have previously been reported in detail Briefly, the diagnostic criteria for small-duct PSC included (1) chronic cholestatic liver disease of unknown etiology; (2) percutaneous liver biopsy sample with features suggestive of PSC; (3) normal cholangiogram (ie, endoscopic retrograde cholangiography, percutaneous cholangiography, magnetic resonance cholangiography [MRC]); and (4) appropriate exclusion of other liver or biliary disease using standard clinical, laboratory, imaging, and histologic criteria. None of the patients with small-duct PSC fulfilled the criteria for autoimmune hepatitis according to the International Autoimmune Hepatitis Study Group. 14 Also, none of the patients with small- or large-duct PSC had known risk factors for viral hepatitis, and, in most cases, viral serology was investigated either at baseline or at some point during the follow-up. In the Mayo Clinic cohort, 12 all patients with small-duct PSC were required to have inflammatory bowel disease (IBD), whereas IBD was not a mandatory criteria in the European patients. 10,11 The diagnosis criteria of large-duct PSC were the same as for small-duct PSC, except that a positive cholangiogram showing the typical features of PSC was mandatory. In the summer of 2006, the different centers reexamined their respective cohort of patients with small-duct PSC who were reported in the original publications. A detailed and comprehensive review of the medical files of these patients was performed by the principal investigator in each center, and data on clinical outcomes and survival were recorded. The starting point (ie, time zero) for follow-up was the date of the diagnostic cholangiogram (ie, normal in small-duct PSC and abnormal in large-duct PSC) as previously described ; the follow-up was extended until mid The end points recorded were death, date of liver transplantation, and last date of follow-up. The following information was obtained in all patients: date of diagnosis, date of last follow-up, and total duration of follow-up. Laboratory tests at last follow-up and information about symptoms and signs developing during follow-up were obtained. Liver-related signs recorded included gastroesophageal varices, variceal bleeding, ascites, encephalopathy, and itching. Data on the histologic stage (extent of fibrosis) at diagnosis were recorded and divided into early (stage I II) and advanced (stage III IV) disease. Serum bilirubin values at diagnosis were also recorded and compared between the 2 groups. Presence and type of IBD were recorded in each case. Data on immunosuppressive therapy for IBD were collected and the proportion with long-term immunosuppression (at least 1 year of treatment with azathioprine or methotrexate or treatment with corticosteroids for more than 6 months) was compared between the 2 groups. The proportion of patients treated with ursodeoxycholic acid (UDCA) was analyzed, and the total duration of UDCA treatment in months was recorded. The number of patients undergoing repeated cholangiography (endoscopic retrograde cholangiography, percutaneous cholangiography, MRC) during the follow-up was recorded. The cholangiography images were interpreted by radiologists with experience in biliary tract disease in each participating center. In patients with small-duct PSC, repeated cholangiographies were performed solely for the purpose of this study in the vast majority of patients from Sweden as described. 11 Cholangiographies were only repeated when considered clinically indicated in patients from the Mayo Clinic 12 and in the Norwegian and United Kingdom cohorts. 10 The indication for a repeated cholangiography was based on worsening of liver tests such as increase in alkaline phosphatase and/or serum bilirubin and new symptoms such as cholangitis, itch, or pain considered clinically significant by the treating physician. The development of cholangiocarcinoma was noted, and the number and causes of death were analyzed and divided into liver-related and non-liver related deaths. The study was approved by appropriate regulatory bodies in all participating medical institutions, and all patients had given informed consent for participation in medical research. Matching Process Each patient with small-duct PSC was randomly matched to 2 patients with well-characterized large-duct PSC (1:2 matching design) by gender, age ( 3 years), calendar year of diagnosis ( 3 years), and medical institution. In the 3 original cohorts, a total of 83 patients with small-duct PSC were reported including 32 from Sweden, 22 from Oxford, 18 from the Mayo Clinic, and 11 from Oslo This matching process was accomplished in all except some patients from the Oslo center. Specifically, for the 11 patients with small-duct PSC from Oslo, only 13 controls with large-duct PSC were identified who could be appropriately matched because of large difference in either age or calendar year of diagnosis. Thus, a total of 157 patients with large-duct PSC were compared with the 83 patients with small-duct PSC. Statistical Analysis Continuous variables are presented as medians and interquartile range (IQR). The comparison of continuous variables between the 2 groups was performed by the nonparametric Wilcoxon rank sum test; dichotomous variables were compared using the 2 or the Fischer exact test where appropriate. The Kaplan Meier product limit was used for estimating survival free of liver transplantation. The comparison of survival distribution between the small- and the large-duct PSC patients was performed using the log-rank test. 15

3 April 2008 SMALL-DUCT PRIMARY SCLEROSING CHOLANGITIS 977 Results General Characteristics of Patients With Small- and Large-Duct PSC Among the 83 patients with small-duct PSC, 51 were male (61.5%) and 32 female (38.5%). In the largeduct group, 97 (61.8%) were male and 60 female (38.3%). Median age at diagnosis was 38 (IQR, 28 50) years and 38 (IQR, 28 48) years in the small-duct and the largeduct PSC group, respectively. The median duration of follow-up was 13 (IQR, 10 16) years among the smallduct patients, whereas the large-duct patients had a follow-up of 10 (IQR, ) years (P.002). A total of 72% of the patients in the small-duct group had histologic stage I II disease, and 28% were in stage III IV disease at diagnosis. Corresponding figures in the largeduct group were 66% and 34%, respectively (P.36). Median serum bilirubin at diagnosis was 13 (IQR, 8 19) mol/l (0.76 [IQR, ] mg/dl) vs 14 (IQR, 9 24) mol/l (0.82 [IQR, ] mg/dl) in the small- and the large-duct groups, respectively (P.73). The proportions of patients receiving long-term immunosuppressive therapy were also similar in the 2 groups, ie, 13% in the small-duct group and 15% in the large-duct group (P.70). Clinical Features of Patients With Small-Duct PSC A repeated cholangiography was performed in 68 patients with small-duct PSC, and in 19 (27.9%) of them the cholangiography showed unequivocal evidence of progression to large-duct PSC. Thus, 19 (22.9%) out of the 83 patients with small-duct PSC were shown to have cholangiographic changes compatible with large-duct PSC on repeated cholangiography. The median time from the baseline normal cholangiography to repeated cholangiography showing progression to large-duct PSC was 7.4 (IQR, ) years. Among the other patients who did not undergo repeated cholangiography, none showed any significant clinical or laboratory evidence of worsening of the liver disease indicating the need for a new cholangiography. Sixty-seven (80.7%) patients with small-duct PSC had IBD, either at baseline or at follow-up. Among these, 52 (78%) had ulcerative colitis, 14 (21%) had Crohn s colitis, and 1 patient had collagenous colitis. Twenty patients (30%) underwent colectomy, including 16 patients with ulcerative colitis and 4 with Crohn s disease. The clinical characteristics and outcomes among patients with smallduct PSC with and without IBD are shown in Table 1. The outcome in patients with small-duct PSC and concomitant ulcerative colitis vs those with Crohn s disease was similar in terms of patients dying or undergoing liver transplantation. Four of the 14 (28.6%) patients with Crohn s disease died or underwent liver transplantation vs 11 out of 48 (22.9%) with ulcerative colitis (P.8). Table 1. The Clinical Characteristics and Outcome in Patients With Small-Duct PSC With and Without IBD IBD (n 66) Non-IBD (n 17) Age at diagnosis, y 35 (24 50) 41 (36 50) Sex (males/females) (%) 45/21 (68%/32%) 11/6 (64%/36%) Duration follow-up (y) 14 (10 11) 11 (9 13.5) Death/Tx (%) 16 (24%) 3 (17.6%) Tx (%) 7 (10.6%) 1 (6%) Death (%) 9 (13.6%) 2 (11.7%) NOTE. Medians and interquartile range are given for the age and the follow-up time. No significant differences were observed in any of the variables. Tx, liver transplantation. Esophageal varices, variceal bleeding, ascites, and encephalopathy were diagnosed during follow-up only in those patients who died or underwent a liver transplantation. Other liver-related symptoms such as pruritus was uncommon in those who survived and was present at some period in time during follow-up among 6 (9.4%) out of the 64 survivors, whereas pruritus at some time period was present in 4 (21%) of the 19 patients who died or underwent a liver transplantation. At the end of follow-up, none of the patients with small-duct PSC had developed jaundice, and jaundice had occurred only in those patients who died or underwent liver transplantation. Twenty-three (27.7%) patients with small-duct PSC had been treated with UDCA, with a median duration of 10.5 (IQR, ; range, years) years. Development of Cholangiocarcinoma One patient with small-duct PSC who was diagnosed with large-duct PSC on repeated cholangiography after 10 years was found to have cholangiocarcinoma at the time of diagnosis of large-duct PSC. Otherwise, none of the patients with small-duct PSC developed cholangiocarcinoma or other intestinal malignancies during follow-up. Cholangiocarcinoma developed in 19 out of the 157 patients with large duct PSC (12%), and all these 20 patients with cholangiocarcinoma died, most within a few months from diagnosis. Long-Term Survival and Causes of Death The number of deaths and liver transplantation procedures performed was 11 (13.3%) and 8 (9.6%), respectively, in the small-duct PSC group and 45 (28.7%) and 33 (21%), respectively, in the large-duct PSC group. Among the small-duct patients who died and did not undergo a liver transplantation, liver-related deaths were recorded in 6 of 11 (55%) patients, whereas 5 patients (45%) had a nonliver-related death. Deaths from liverrelated complications were due to complications of cirrhosis (none died from hepatocellular carcinoma). The proportion of patients dying or undergoing liver trans-

4 978 BJÖRNSSON ET AL GASTROENTEROLOGY Vol. 134, No. 4 with UDCA for a median of 8 years, whereas 16 of the 64 (25%) patients who did not progress to large-duct PSC had been treated with UDCA for a median of 9 years (P.9). Five of the 19 (26%) patients with small-duct PSC who died or underwent liver transplantation had been treated with UDCA for a median of 7 years as compared with a median duration of treatment of 9.5 years in 18 of the 64 (28%) patients surviving without liver transplantation (P.9). Figure 1. Kaplan Meier estimated curves for (liver transplantation free) survival in patients with small-duct and large-duct primary sclerosing cholangitis. Survival free of liver transplantation was significantly longer in patients with small-duct PSC (hazard ratio, 3.04; 95% CI: ; P.0001). plantation was significantly higher in the large-duct PSC group than in the small-duct PSC group (49.7% vs 22.9%, respectively; P.0001). In the large-duct PSC group, 35 (78%) out of 45 deaths were due to liver-related complications, whereas 10 (22%) had a nonliver-related death. Patients with small-duct PSC had a significantly longer median survival free of liver transplantation as compared with patients with large-duct PSC (13 years [IQR, 10 17] vs 10 years [IQR, 6 14], respectively; hazard ratio, 3.04; 95% CI: ; P.0001) as illustrated in Figure 1. Among the small-duct patients in stage I II disease, a total of 23% died or underwent liver transplantation, and, similarly, 22.7% of those with fibrosis stage III IV at baseline had this poor outcome. However, in the largeduct group, 73.5% of those with stage III IV died or underwent liver transplantation, whereas this outcome was observed in only 40.9% of those with stage I II disease (P.0004). Among the 19 patients with small-duct PSC who progressed to large-duct PSC during follow-up, 9 (47%) either died (n 4) or underwent a liver transplantation (n 5), whereas in the other 10 patients the disease remained stable. Conversely, only 10 of the 64 (15.6%) patients with small-duct PSC who did not progress to large-duct disease died (n 6) or underwent liver transplantation (n 4). Thus, a significantly higher proportion of patients with small-duct PSC progressing to largeduct PSC reached the outcomes of death or liver transplantation than those who did not progress (47% vs 15.6%, respectively, P.004). Two of the 8 patients in the small-duct group who underwent liver transplantation developed recurrent small-duct PSC in the graft requiring retransplantation 9 and 13 years later, respectively. In patients with small-duct PSC, 7 of the 19 (37%) who subsequently developed large-duct PSC had been treated Discussion The current series of a large number of patients with small-duct PSC and appropriately matched controls with large-duct PSC demonstrates that (1) patients with small-duct PSC follow a significantly better long-term prognosis as compared with large-duct PSC; (2) approximately a fourth of patients with small-duct PSC may progress to large-duct PSC within an average follow-up of 8 years; (3) cholangiocarcinoma does not seem to occur in patients with small-duct PSC; (4) patients with smallduct PSC may progress to end-stage liver disease with the consequent need of liver transplantation without evidence of development of large-duct disease; and (5) small-duct PSC may recur in the allograft. In our study, patients with small-duct PSC had a significantly better long-term survival than patients with large-duct PSC. The natural history of classic large-duct PSC is highly variable. 1 8 Some patients have a rapid progression of symptoms soon after diagnosis, whereas other patients have a long-standing silent disease without signs of progression for many years. 1 8 Thus, to determine the natural history of small-duct PSC, we chose to include a large number of well-characterized patients with large-duct PSC for comparison. Furthermore, the long duration of follow-up of the small-duct group in our series is of major importance to assess the natural history and the different outcomes in this group of patients. The 13 years of a median follow-up in our patients with small-duct PSC is the longest follow-up ever reported in patients with small-duct PSC. Although patients with large-duct PSC were matched by calendar year of diagnosis, they had a significantly shorter follow-up as compared with patients with small-duct PSC because of progression of their liver disease leading to death or liver transplantation faster than in the smallduct PSC group. Despite the prolonged follow-up, only approximately 23% of the small-duct PSC group underwent liver transplantation or died, and the majority of the small-duct PSC patients dying without having liver transplantation had a nonliver-related cause of death. In the large-duct PSC group, almost 50% of patients underwent liver transplantation or died, and in almost 80% of the patients with large-duct PSC who died without liver transplantation, the cause of death was a liver-related complication. Thus, our data suggest that small-duct PSC may be an entity on its own, and in only one fourth

5 April 2008 SMALL-DUCT PRIMARY SCLEROSING CHOLANGITIS 979 of cases the disease represents an earlier stage of classic, large-duct PSC. Although 80% of patients with small-duct PSC underwent repeated cholangiography during follow-up, a limitation of our study is that a number of patients did not have a repeated cholangiography during follow-up. In those who were investigated with a new endoscopic retrograde cholangiography or MRC, a total of 22.9% from the total cohort of 83 (or 27.9% of the 68 patients who underwent repeated cholangiography) were found to have developed large-duct PSC. This represents approximately one fourth of patients showing progression to large-duct disease, which is clearly higher than the 12% 13% progression rate in 2 of the prior series 10,11 and higher than the 16.7% progression rate reported in another series. 12 This suggests that the rate of progression from small-duct disease to large-duct disease will be higher as the follow-up lengthens. The remaining 20% of patients with small-duct PSC who did not undergo repeated cholangiography did not show any clinical or laboratory deterioration of their liver disease to indicate the need for cholangiography. Thus, although it cannot be excluded that more patients had developed large-duct PSC, it seems unlikely that a high proportion of those with apparently clinically silent disease had in fact developed large-duct disease. Certainly, it would be unethical to perform invasive cholangiographies (endoscopic retrograde cholangiography or percutaneous cholangiography) repetitively during follow-up in those patients with small-duct PSC whose disease remains stable over time. However, the development of accurate noninvasive techniques for evaluation of the biliary tract such as MRC can make it possible to prospectively follow patients with small-duct PSC. Further prospective studies in this area are necessary. One patient with small-duct PSC who progressed to large-duct PSC during follow-up was diagnosed with cholangiocarcinoma, but, at the same time, a large-duct PSC was identified. Cholangiocarcinoma developed, however, in 12% of the large-duct PSC patients, and all died during follow-up. In none of the patients with small-duct PSC who did not progress to large-duct PSC, was a cholangiocarcinoma observed during this lengthy followup. To our knowledge, the development of cholangiocarcinoma in a patient with small-duct PSC has not been reported ,16,17 It is conceivable that a diseased largeduct epithelium is a prerequisite for the development of cholangiocarcinoma and that the epithelium of small, microscopic bile ducts is resistant or less vulnerable to the carcinogenicity leading to cholangiocarcinoma found to occur in PSC. It is also possible that differences in the pathophysiology of small-duct and large-duct PSC explain this lack of predisposition to develop cholangiocarcinoma; nevertheless, any theories to explain this difference in malignant potential of both conditions would remain elusive at this time. Understanding the mechanisms leading to small-duct or large-duct disease may help to understand the pathophysiology of cholangiocarcinoma occurring in PSC. In our series, a significantly higher proportion of patients with small-duct PSC progressing to large-duct PSC died or underwent liver transplantation as compared with those who did not progress. This shows that, without having progressed to large-duct disease, patients with small-duct PSC may still die from liver-related complications or undergo transplantation. However, most deaths or liver transplantation procedures performed in patients with small-duct PSC occurred once this condition had progressed to large-duct PSC disease. Interestingly, 2 patients with small-duct PSC originally having undergone transplantation for end-stage, smallduct PSC developed recurrent small-duct PSC in the graft requiring retransplantation 9 and 13 years later, respectively. These 2 patients underline the fact that small-duct PSC is a disease of recurrence potential after liver transplantation just as occurs in liver diseases with an autoimmune or viral etiology. Crohn s disease was present in 21% of the patients with IBD in the small-duct PSC group, which is the highest prevalence of Crohn s disease reported in patients with PSC. 1 8 In most studies, the prevalence of Crohn s disease has been present in 5% 10% of patients. 1 8 This is in line with a study of hepatobiliary dysfunction in patients with Crohn s disease demonstrating small-duct PSC as the most common associated condition. 18 However, the occurrence of IBD in association with small-duct PSC in the current study was not found to influence the prognosis because prognosis was similar in patients with and without IBD. However, this might be due to a type II error given the limited number without IBD among the patients of the current study. No tendency was observed for a different outcome in those with Crohn s disease as compared with those with concomitant ulcerative colitis. We recognize that referral bias concerning the nature of the PSC patients reported in the current study cannot be excluded. This potential referral bias is due to the fact that the medical institutions participating in this study are tertiary referral centers in their countries. However, this might affect both the small- and the large-duct groups, and it is therefore not likely to explain the large differences in outcomes between the 2 groups. In summary, our multicenter series of a large number of patients with small-duct PSC who underwent the longest follow-up ever reported demonstrates that patients with small-duct PSC have a significantly better long-term prognosis than appropriately matched patients with large-duct PSC. In a proportion of patients with small-duct PSC, the disease progresses to large-duct disease, but progression to large-duct disease is not necessary for progressing to end-stage liver disease and requiring liver transplantation. Cholangiocarcinoma does

6 980 BJÖRNSSON ET AL GASTROENTEROLOGY Vol. 134, No. 4 not seem to occur in patients with small-duct PSC unless 11. Broomé U, Glaumann H, Lindstöm E, et al. Natural history and the disease has progressed to large-duct PSC. outcome in 32 Swedish patients with small duct primary sclerosing cholangitis (PSC). J Hepatol 2002;36: Angulo P, Maor-Kendler Y, Lindor KD. Small-duct primary sclerosing cholangitis:prevalence and natural history. Hepatology 2002; References 35: Aadland E, Schrumpf E, Fausa O, et al. Primary sclerosing cholangitis: a long-term follow-up study. Scand J Gastroenterol 1987; 22: Wiesner RH, Grambsch PM, Dickson ER, et al. Primary sclerosing cholangitis: natural history, prognostic factors and survival analysis. Hepatology 1989;10: Farrant JM, Hayllar KM, Wilkinson ML, et al. Natural history and prognostic variables in primary sclerosing cholangitis. Gastroenterology 1991;100: Dickson ER, Murtaugh PA, Wiesner RH, et al. Primary sclerosing cholangitis: refinement and validation of survival models. Gastroenterology 1992;103: Broomé U, Olsson R, Lööf L, et al. Natural history and prognostic factors in 305 patients with primary sclerosing cholangitis. Gut 1996;38: Okolicsanyi L, Fabris L, Viaggi S, et al. Primary sclerosing cholangitis: clinical presentation, natural history and prognostic variables: an Italian multicentre study. The Italian PSC Study Group. Eur J Gastroenterol Hepatol 1996;8: Ponsioen CY, Vrouenraets SM, Prawirodirdjo W, et al. Natural history of primary sclerosing cholangitis and prognostic value of cholangiography in a Dutch population. Gut 2002;51: Bergquist A, Broome U. Clinical features in primary sclerosing cholangitis. Clin Liver Dis 1998;2: Nikolaidis NL, Giouleme OI, Tziomalos KA, et al. Small-duct primary sclerosing cholangitis. A single-center seven-year experience. Dig Dis Sci 2005;50: Alvarez F, Berg PA, Bianchi FR, et al. International autoimmune hepatitis group report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol 1999;31: Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53: Wiesner RH, LaRusso NF. Clinicopathologic features of the syndrome of primary sclerosing cholangitis. Gastroenterology 1980; 79: Riordan SM, Williams R. Risk of cholangiocarcinoma in primary sclerosing cholangitis. In: Manns MP, Chapman RW, Stiehl A, eds. Primary sclerosing cholangitis. London: Kluwer Academic Publishers, 1997: Rasmussen HH, Fallingborg JF, Mortensen PB, et al. Hepatobiliary dysfunction and primary sclerosing cholangitis in patients with Crohn s disease. Scand J Gastroenterol 1997;32: Received September 23, Accepted January 10, Address requests for reprints to: Einar Björnsson, MD, Department of Internal Medicine, Sahlgrenska University Hospital, SE Goth- Sweden. einar.bjornsson@medic.gu.se; fax: (46) 9. Wee A, Ludwig J, Coffey RJ Jr, et al. Hepatobiliary carcinomaenburg, associated with primary sclerosing cholangitis and chronic ulcerative colitis. Hum Pathol 1985;16: The authors thank Hanne Prytz, Hanna Sandberg-Gertzen, Sven Björnsson E, Boberg KM, Cullen S, et al. Patients with small duct Almer, and Per Sangfeldt for their help in gathering some of the primary sclerosing cholangitis have a favorable long-term prognosis. Gut 2002;51: Conflicts of interest: The authors have no conflicts of patients follow-up data. interest.