Prevalence of Sclerosing Cholangitis Detected by Magnetic Resonance Cholangiography in Patients With Long-term Inflammatory Bowel Disease

Accepted Manuscript Prevalence of Sclerosing Cholangitis, Detected by Magnetic Resonance Cholangiography, in Patients with Long-term Inflammatory Bowel Disease Aida Kapic Lunder, Johannes Roksund Hov, Arne Borthne, Jostein Gleditsch, Glenn Johannesen, Knut Tveit, Ellen Viktil, Magne Henriksen, Øistein Hovde, Gert Huppertz-Hauss, Ole Høie, Marte Lie Høivik, Iril Monstad, Inger Camilla Solberg, Jørgen Jahnsen, Tom Hemming Karlsen, Bjørn Moum, Morten Vatn, Anne Negård PII: DOI: Reference:

S0016-5085(16)34678-9 10.1053/j.gastro.2016.06.021 YGAST 60528

To appear in: Gastroenterology Accepted Date: 13 June 2016 Please cite this article as: Lunder AK, Hov JR, Borthne A, Gleditsch J, Johannesen G, Tveit K, Viktil E, Henriksen M, Hovde Ø, Huppertz-Hauss G, Høie O, Lie Høivik M, Monstad I, Solberg IC, Jahnsen J, Karlsen TH, Moum B, Vatn M, Negård A, Prevalence of Sclerosing Cholangitis, Detected by Magnetic Resonance Cholangiography, in Patients with Long-term Inflammatory Bowel Disease, Gastroenterology (2016), doi: 10.1053/j.gastro.2016.06.021. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Prevalence of Sclerosing Cholangitis, Detected by Magnetic Resonance Cholangiography, in Patients with Long-term Inflammatory Bowel Disease

RI PT

Aida Kapic Lunder1,2*, Johannes Roksund Hov2,3,4,5,6, Arne Borthne1,2,

Jostein Gleditsch7, Glenn Johannesen8, Knut Tveit9, Ellen Viktil10, Magne Henriksen11, Øistein Hovde2,12, Gert Huppertz-Hauss13, Ole Høie14, Marte

SC

Lie Høivik15, Iril Monstad2,15, Inger Camilla Solberg15, Jørgen Jahnsen2,16, Tom Hemming Karlsen2,3,4,5,6, Bjørn Moum2,15, Morten Vatn2,17, Anne

1

M AN U

Negård1,3

Department of Radiology, Akershus University Hospital, Lørenskog; 2

Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo; 3

TE D

Norwegian PSC Research Center, Department of Transplantation Medicine, Division of Surgery, Inflammatory Medicine and Transplantation, Oslo University Hospital, Rikshospitalet, Oslo; 4 Section for Gastroenterology,

EP

Department of Transplantation Medicine, Division of Surgery, Inflammatory

5

AC C

Medicine and Transplantation, Oslo University Hospital, Rikshospitalet, Oslo; K. G. Jebsen Inflammation Research Centre, Institute of Clinical Medicine,

University of Oslo, Oslo, 6 Research Institute of Internal Medicine, Division of Cancer Medicine, Surgery and Transplantation, Oslo University Hospital, Rikshospitalet, Oslo; 7 Department of Radiology, Østfold Hospital, Fredrikstad; 8

Department of Radiology, Telemark Hospital, Skien; 9 Røntgensenteret,

Kristiansand; 10 Department of Radiology, Oslo University Hospital, Ullevål, Oslo; 11 Department of Gastroenterology, Østfold Hospital, Fredrikstad; 12

1

ACCEPTED MANUSCRIPT Department of Gastroenterology, Innlandet Hospital, Gjøvik; 13 Department of Gastroenterology, Telemark Hospital, Skien; 14 Department of Gastroenterology, Sørlandet Hospital, Arendal; 15 Department of Gastroenterology, Division of Medicine, Oslo University Hospital, Ullevål,

RI PT

Oslo; 16 Department of Gastroenterology, Akershus University Hospital,

*Corresponding author Address: PB 1000, 1478 Lørenskog, Norway E-mail address: [email protected]

M AN U

Telephone number: 00 47 952 97 027

SC

Lørenskog; 17 EpiGen Institute, Akershus University Hospital, Lørenskog.

Aida Kapic Lunder: acquisition of data; analysis and interpretation of data; drafting of the manuscript; critical revision of the manuscript for important

TE D

intellectual content; statistical analysis.

Johannes Roksund Hov: study concept and design; analysis and interpretation of data; drafting of the manuscript; critical revision of the

EP

manuscript for important intellectual content; statistical analysis.

AC C

Arne Borthne: study concept and design; acquisition of data; analysis and interpretation of data; critical revision of the manuscript for important intellectual content.

Jostein Gleditsch: acquisition of data; critical revision of the manuscript for important intellectual content. Glenn Johannesen: acquisition of data; critical revision of the manuscript for important intellectual content. Knut Tveit: acquisition of data; critical revision of the manuscript for important

2

ACCEPTED MANUSCRIPT intellectual content. Ellen Viktil: acquisition of data; critical revision of the manuscript for important intellectual content. Magne Henriksen: acquisition of data; critical revision of the manuscript for

RI PT

important intellectual content.

Øistein Hovde: acquisition of data; critical revision of the manuscript for important intellectual content.

SC

Gert Huppertz-Hauss: acquisition of data; critical revision of the manuscript for important intellectual content.

M AN U

Ole Høie: acquisition of data; critical revision of the manuscript for important intellectual content.

Marte Lie Høivik: analysis and interpretation of data; critical revision of the manuscript for important intellectual content.

TE D

Iril Monstad: acquisition of data; critical revision of the manuscript for important intellectual content.

Inger Camilla Solberg: acquisition of data; critical revision of the manuscript

EP

for important intellectual content.

AC C

Jørgen Jahnsen: study concept and design; acquisition of data; critical revision of the manuscript for important intellectual content. Tom Hemming Karlsen: study concept and design; critical revision of the manuscript for important intellectual content; obtained funding. Bjørn Moum: study concept and design; critical revision of the manuscript for important intellectual content; study supervision. Morten Vatn: study concept and design; analysis and interpretation of data; drafting of the manuscript; critical revision of the manuscript for important

3

ACCEPTED MANUSCRIPT intellectual content; obtained funding. Anne Negård: study concept and design; acquisition of data; analysis and interpretation of data; drafting of the manuscript; critical revision of the

funding; study supervision. The authors have no conflicts of interest to declare.

South-East Norway Regional Health Authority

SC

The study was supported by:

RI PT

manuscript for important intellectual content; statistical analysis; obtained

PSC Research Center

M AN U

Akershus University Hospital Research Foundation

The Norwegian Society of Radiology

TE D

Writing assistance provided by: American Journal Experts, LLC

601 West Main Street, Suite 102

EP

Durham, NC 27701 USA

AC C

Southeast Norway Regional Health Authority funded the writing support.

Acknowledgements: We would like to thank the South-Eastern Norway Regional Health Authority, Akershus University Hospital, Norwegian PSC Research Center and Norwegian Society of Radiology for funding. We thank the IBSEN study group for their meticulous work during the past two decades, Tryggve Holck Storås for technical MRI support and Jonas Lindstrøm for statistical support.

4

ACCEPTED MANUSCRIPT ABSTRACT

Background & Aims: The prevalence of primary sclerosing cholangitis (PSC) among patients with inflammatory bowel disease (IBD) is unclear. Patients

RI PT

with IBD might be screened for PSC using magnetic resonance

cholangiography (MRC). We aimed to estimate the frequency and distribution of MRC-detected lesions that indicate PSC in patients with IBD 20 years after

SC

their initial diagnosis and to identify clinical characteristics associated with

M AN U

these findings.

Methods: We performed a follow-up analysis of a population-based cohort of 756 patients in South-Eastern Norway diagnosed with IBD from January 1, 1990 through December 31, 1993. Of these subjects, 470 attended a follow-

TE D

up evaluation 20 years later in which they were offered routine clinical blood testing and ileocolonoscopy; 322 were screened by MRC (222 with ulcerative colitis and 100 with Crohn’s disease). Two radiologists independently

AC C

EP

evaluated results from the MRC examinations.

Results: In the MRC examination, 24 patients (7.4%) were found to have PSC-like lesions; only 7 of these patients (2.2%) were known to have PSC. One patient was initially missed and 1 had small-duct PSC, so the final prevalence of PSC was 8.1%. Extensive colitis, a high prevalence of colectomy, and chronic and continuous symptoms of IBD occurred in significantly more patients with suspected PSC than without PSC (P=.029, P=.002, and P=.012, respectively). Among patients with subclinical features of

5

ACCEPTED MANUSCRIPT PSC, the MRC progression score for PSC increased when they were reexamined after a median 3.2 years (P=.046).

Conclusions: Using MRC analysis of patients with long-term IBD, we found

RI PT

the prevalence of PSC to be almost 4-fold higher than that detected based on symptoms. Sixty five percent of patients had subclinical PSC associated with progressive IBD, with no biochemical abnormalities and mild disease, based

M AN U

disease, but long-term outcomes are not known.

SC

on radiology findings. PSC appears to progress in patients with subclinical

KEY WORDS: biliary tract disease; liver disease; magnetic resonance imaging; epidemiology

TE D

INTRODUCTION

Primary sclerosing cholangitis (PSC) is a chronic inflammatory disease of the bile ducts causing biliary strictures and recurrent cholangitis 1 that can

EP

ultimately lead to liver failure and cancer. Currently, there is no established

AC C

medical treatment and liver transplantation is the only potentially curative therapy.

PSC is highly correlated with inflammatory bowel disease (IBD) and approximately 50-80 % of PSC patients have concomitant IBD 2. The reported prevalence of PSC ranges from 0.8 to 5.6 % in ulcerative colitis (UC) patients 3-5

and 0.4 to 6.4 % 6-9 in patients with Crohn´s disease (CD). The true

prevalence remains unclear, and some studies indicate prevalence as high as

6

ACCEPTED MANUSCRIPT 18 % 3, 10. These discrepancies in reported prevalence may be due to methodological differences, selected patient populations and non-uniform diagnostic criteria. However, determining the exact prevalence of PSC in IBD is clinically important because coexistent PSC warrants an extensive follow-

RI PT

up due to an increased risk of cholangiocarcinoma and colon cancer 11, 12.

PSC-IBD is a distinct phenotype that is almost universally colonic, with a right-

SC

sided predominance, high frequency of “backwash” ileitis and rectal sparing 13

. Large-scale genetic studies 14 suggest that PSC only partly overlaps with

M AN U

IBD, promoting the notion of a separate PSC-IBD subtype. The true prevalence of PSC in IBD is therefore likely to complicate all studies of etiopathogenesis in IBD and PSC, and a more precise differentiation of the IBD subgroups would be highly valuable. Recently, the use of MRC screening

TE D

in IBD patients without clinical suspicion of PSC has been studied 15, 16 because PSC may have a subclinical course. Magnetic resonance cholangiography (MRC) has largely replaced endoscopic retrograde

EP

cholangiography (ERC) in the initial diagnosis of PSC because it is non-

AC C

invasive, comparably accurate 17, and a less expensive method.

The Inflammatory Bowel Disease South-East Norway (IBSEN) study represents a prospective, population-based cohort of IBD patients recruited at the time of primary diagnosis (1990-1994) and later examined at 1, 5, and 10 years 18, 19. As part of the prescheduled 20-year follow-up, all eligible patients were offered an MRC. The aim of the present study was to estimate the frequency and distribution of MRC lesions indicating PSC in IBD patients two

7

ACCEPTED MANUSCRIPT decades after the initial diagnosis and to identify clinical characteristics associated with these findings.

MATERIALS AND METHODS

RI PT

Study population

From January 1, 1990 to December 31, 1993, all newly diagnosed IBD

patients in South-Eastern Norway were prospectively recruited. The design

SC

and inclusion criteria are described elsewhere 20. The prescheduled follow-up was performed 1, 5, and 10 years after the diagnosis. Figure 1 shows a

M AN U

flowchart of the IBSEN cohort in which 843 patients suspected of having IBD were initially recruited. Of these, 756 received a final IBD diagnosis that formed the basis for further analysis. According to data from the Norwegian causes of death registry, the mortality rates for this cohort were comparable to

TE D

the general Norwegian population 21, 22.

The IBSEN study invited all patients to a 20-year follow-up visit that included a

EP

structured interview, a clinical examination, routine clinical blood tests,

AC C

ileocolonoscopy and an MRC examination. Of 470 patients who attended the 20-year follow-up, 327 (69.6 %) completed the MRC study. The somewhat low participation rate could be due to the fact that these patients were asymptomatic and in their opinion, MRC offered no personal health benefit. Four patients were excluded due to the presence of foreign bodies (2), pregnancy (1) or dementia (1), and 5 MRCs were discarded due to poor image quality.

8

ACCEPTED MANUSCRIPT Whenever available, earlier MRCs were collected retrospectively for the subset of patients with clinically known PSC in whom an MRC had not been performed as part of the study. Clinical data were retrieved and together with previous imaging data used as a reference standard for assessing MRC

RI PT

sensitivity in diagnosing PSC. The results of the serological analysis for perinuclear Anti-Neutrophil Cytoplasmic Antibodies (pANCA) and Anti-

Saccharomyces cerevisiae antibodies (ASCA) presence from the 10-year

SC

follow-up were available for analysis 23.

M AN U

Imaging

The MRC examinations were performed at 5 locations with 1,5 T scanners (Achieva/Ingenia, Philips; Avanto/Aera, Siemens; Excelert Vantage, Toshiba) between March 9th 2011 and March 26th 2015. A shared magnetic resonance

TE D

imaging (MRI) protocol was developed, consisting of a coronal and axial turbo spin echo (TSE) T2 weighted sequence, an axial T1 weighted gradient echo sequence in and opposed of phase, a 3D single shot (SS) TSE MRC

EP

sequence with maximum intensity projection (MIP) reconstruction and one of the following two sequences: a 2D SS TSE MRC or a thick slab SS TSE

AC C

MRC. More details on the MRI protocol are provided in the supplementary Appendix 1. As part of the clinical follow-up the MRC examination was repeated in patients with PSC-like lesions on the initial MRC (May 28th 2014 April 11th 2016).

Image analysis The images were individually analyzed by two radiologists with 10 to 15 years of experience in abdominal MR imaging. The reviewers were blinded to the

9

ACCEPTED MANUSCRIPT patients' previous imaging findings, other test results and clinical data. The MRC findings were recorded as published in a previous study 24. The reading procedure was initially rehearsed with 10 MRC examinations not included in

RI PT

the study.

MRC findings like bile duct strictures, wall irregularities, focal dilatation and

beading at the level of extrahepatic, first and second order intrahepatic and

SC

peripheral biliary ducts, were graded as being present or absent. Ductal stenosis was considered present when the extrahepatic duct was not

M AN U

depicted, the dilated ducts were partially narrowed or dilated peripheral ducts with no connection to central ducts were seen. Extrahepatic ductal dilatation was considered present when the common bile duct was ≥ 10 mm in diameter. Intrahepatic bile ducts were evaluated with regard to their central

TE D

portion. Beading was recorded in cases of closely alternating duct dilatation and stenosis. The presence of PSC lesions was graded from 1-5 (1 – present, 2 – probably present, 3 – uncertain, 4 – probably not present, 5 – not present) 24

EP

. Disagreements were resolved by consensus reading, when the final

AC C

diagnosis was made. ‘Definitely present’ and ‘probably present’ ratings were considered positive.

An MRI progression risk score, as proposed by Ruiz et al. 25, was assessed in all patients with a positive MRC. Images were reviewed in consensus, and the score, based on the presence of predictive radiologic features, was calculated as follows: 1 x dilatation of intrahepatic bile + 2 x dysmorphy + 1 x portal hypertension.

10

ACCEPTED MANUSCRIPT

Statistical analysis Categorical variables were compared using the chi-square or Fisher’s exact test and continuous variables using Student’s t-test, analysis of variance

RI PT

(ANOVA) and Wilcoxon's signed rank test, as appropriate. The criterion of

statistical significance was P<.05. Interobserver agreement was calculated

using kappa statistics. The level of agreement was given as the percentage of

SC

negative and positive agreement, kappa and prevalence adjusted bias

adjusted kappa (PABAK) statistics 26. A k value of less than .20 represented

M AN U

poor agreement, .21–.40 was deemed fair, .41–.60 moderate, .61–.80 good, and .81 or greater represented excellent agreement. The sensitivity of MRC in diagnosing PSC was calculated. Statistical analyses were performed using

TE D

SPSS v. 23.0 software (SPSS, Inc., Chicago, IL).

Ethical Requirements

The Regional Committee for Medical and Health Research Ethics South East

AC C

consent.

EP

Norway approved the study protocol. All subjects provided informed written

RESULTS

Accumulated prevalence of clinically known PSC in long-term IBD

Prior to the MRC study, a total 17 of 756 IBD patients were diagnosed with PSC, yielding a cumulative prevalence of 2.2 %. Among these, two patients were diagnosed with small duct PSC. The mean age at IBD and PSC diagnosis was 34.5 (SD 14.5) and 39.1 (SD 13.8) years, respectively. Five

11

ACCEPTED MANUSCRIPT PSC patients (29.4 %) were female. Nine (52.9 %) had UC and eight had CD (47.1 %), and all experienced colonic involvement. Four patients (23.5 %) were liver-transplanted while 3 (17.6 %) were deceased at the time of follow-

RI PT

up.

Thirty-six of 756 included patients died of cancer during the 20 years follow-up 21, 22

. There was only one case of cholangiocarcinoma in a patient with known

SC

PSC. One patient died of cancer in the papilla Vateri region, which could have been possibly related to undiagnosed PSC but overall, unrecognized PSC

MRC study population

M AN U

was not a major cause of death in the cohort.

As shown in Figure 1, 599 (79.2 %) of 756 initially included IBD patients were

TE D

alive at the time of follow-up, of which 470 (78.5 %) completed the 20-year follow-up (mean follow-up time, 20.2 years, SD: 1.3). MRC was performed in 322 of 470 patients (68.5 %). The comparison of the demographic and clinical

EP

characteristics of the cohort showed that all patients alive at 20 years had a similar distribution of age, gender, IBD type and disease extension at baseline

AC C

(Appendix 2). Among the 470 patients completing the follow-up, there were no significant differences at 20 years between the individuals examined by MRC and those not (Table 1), suggesting that the MRC group was comparable to all living patients 20 years after the diagnosis.

Interobserver agreement and diagnostic sensitivity of MRC The interobserver agreement was moderate, with a kappa value of 0.55 for the presence of PSC-like lesions. The positive agreement was 0.57 and the

12

ACCEPTED MANUSCRIPT negative agreement was 0.98. Prevalence adjusted bias adjusted kappa (PABAK) was excellent, with a value of 0.93.

Within the subgroup of patients with known large duct PSC (n=15), 8 were

RI PT

examined as part of the IBSEN cohort follow-up and 3 had MRC examinations that could be retrieved for review. The blinded MRC review correctly identified 10 of 11 patients with a final diagnosis of large duct PSC (Figure 2a). This

SC

yielded a sensitivity of 90.9 %. Repeated investigation of the patient who was

M AN U

missed initially revealed sparse intrahepatic bile duct involvement (Figure 2b).

Prevalence and distribution of PSC-like bile duct lesions on MRC Twenty-four patients presented with a positive MRC screening (7.4 %) (Figure 2c - e). Intrahepatic involvement was found in almost all cases, and the

TE D

majority of patients had isolated intrahepatic involvement (Appendix 3). Extrahepatic bile duct strictures were generally observed less commonly than intrahepatic strictures (p=0.009). Isolated extrahepatic bile duct involvement

AC C

EP

was seen in only one patient.

Clinical characteristics of patients with PSC and PSC-like bile duct lesions

Of 24 patients with a positive MRC screening, 7 (2.2 %) were previously diagnosed with PSC (Figure 3). The prevalence of PSC-like bile duct lesions varied by IBD type, with 9.0 % in CD and 6.8 % in UC patients. The condition was more common in women than in men, 9.2 % and 5.9 % respectively, but neither the gender nor the IBD type differences were statistically significant.

13

ACCEPTED MANUSCRIPT One PSC patient was missed in the initial assessment and one had small duct PSC, giving an ultimate PSC prevalence of 8.1 % (26/322). Table 2 shows the clinical characteristics of patients according to the presence of PSC. There were relatively more women in the PSC group than in the non-PSC group.

PSC-like lesions (P= 0.038, Table 3).

RI PT

This difference was caused by the female predominance in the group with

SC

In UC, the presence of PSC was associated with extensive colitis in 80 % compared with 48.3 % in the non-PSC group (P=0.029), irrespective of

M AN U

whether PSC was previously known or not. Ten of eleven CD patients had colonic involvement, which was not significantly different from patients without PSC. The one patient without colonic involvement had intrahepatic lesions

TE D

evaluated as possible PSC (Figure 2f).

Regarding patterns of IBD disease, in terms of self-reported severity of bowel symptoms (Appendix 4) 18, patients with PSC reported chronic, continuous

EP

symptoms more often than the remainder (P=0.012). The presence of PSC

AC C

was also associated with a higher use of steroids in the first 5 years after IBD diagnosis (P=0.018). There were no associations between MRC findings and the use of 5-ASA, azathioprine or biological treatment (Appendix 5). Furthermore, no statistically significant differences in the occurrence of skin, eye and rheumatologic manifestations were found between the groups (Appendix 6).

The accumulated frequency of colectomy was significantly higher in the PSC

14

ACCEPTED MANUSCRIPT group compared to other IBD patients (P<0.001). Three patients in the PSC group were colectomized due to cancer or dysplasia of the colon, compared to one patient in the non-PSC group (P=0.016).

RI PT

Regarding the severity of liver disease, the mean alkaline phosphatase (ALP) and alanine aminotransferase (ALT) levels were significantly elevated

(Appendix 7) at all time points in patients with PSC, but this was confined to

SC

the group with known PSC (Figure 4). The prevalence of pANCA positivity was significantly higher in the PSC-group (P=0.013). The prevalence was

M AN U

highest in the group with established PSC, but there was also a trend towards increased pANCA prevalence in patients with previously unrecognized PSC. There were no significant differences in levels of C-reactive protein (CRP) and Hemoglobin at any time points between the groups. The MRI progression

TE D

score 25 was higher in the group of patients with clinically known PSC than in the group with PSC-like lesions (P=0.076, Table 3).

EP

Clinical and radiological follow-up of MRC positive patients

AC C

Over a median follow-up time of 3.2 years (0.6 – 4.6 years), the second MRC of patients with PSC-like lesions (n= 17) confirmed the cholangiographic findings. The mean MRI progression score of 1.18 (SD 0.95) was significantly higher compared to the initial score of 0.94 (SD 0.75) (P=0.046). Four patients (23.5 %) showed increasing dilatation of the intrahepatic bile ducts. The clinical follow-up revealed 3 patients (17.6 %) with transient liver enzymes elevations. These were not the same patients that exhibited progression on MRC.

15

ACCEPTED MANUSCRIPT

DISCUSSION In this prospective, population-based MRC screening of patients with longterm IBD, probable PSC was found in 8.1 % of patients, which is almost 4

RI PT

times the prevalence of previously known PSC in the same population. Liver biochemistry levels were not elevated and the MRC changes were mild, suggesting early stage or mild PSC. The extended follow-up of newly

SC

detected cases revealed significant, yet still subtle radiological progression.

PSC-like lesions were associated with characteristics of progressive IBD such

M AN U

as extensive colitis, chronic continuous bowel symptoms and increased colectomy rates.

Numerous studies have estimated the prevalence of PSC among patients with

TE D

IBD. After two decades of IBD, the prevalence of clinically recognized PSC in the entire IBSEN cohort was 2.2 %. This is within the range of 0.8-2.4 % previously reported in population-based cohorts 8, 27. However, the present

EP

MRC study revealed possible PSC in 8.1 %, which is in the upper range of

AC C

reported prevalence rates in IBD cohorts from referral centers 5, 9, where higher prevalence rates would be expected. There are few data available for comparison. In a pediatric IBD cohort, MRC identified 15 % with PSC-like lesions and less than 50 % with elevated liver biochemistry levels 15. An unpublished study by Bungay et al. 10 reported that 18 % of IBD patients with normal liver enzymes and colitis had MRC appearances consistent with PSC, suggesting that PSC is more prevalent in IBD than previously acknowledged. One can even argue that 8.1 % in the present study represents an

16

ACCEPTED MANUSCRIPT underestimation of PSC prevalence, taking into consideration the relatively low spatial resolution and slightly lower sensitivity of MRC in detecting PSC compared to ERC 17.

RI PT

Two observations in particular deviate from the current literature: a higher

prevalence of PSC-like lesions in CD than in UC and a higher prevalence in women than in men. Although these differences were not statistically

SC

significant, they differ from the typical categorization of PSC as a disease mainly affecting men with UC 1. Still, the literature does suggest that the

M AN U

prevalence in CD is at least as high in UC. Ronnblom et al. 27 recently reported PSC in 3.0 % of CD patients compared to 1.7 % in UC patients over a 4-year period in a Swedish population-based study. Other studies confirm comparable frequencies, up to 5.6 % in UC patients 5 and 6.4 % in CD

TE D

patients 9. The commonly accepted notion of PSC being primarily associated with UC could therefore be caused by a higher frequency of UC than CD or by challenges in the classification of the colonic inflammation typically observed

EP

in PSC. Almost all PSC patients, regardless of having clinically known or

AC C

subclinical PSC, had extensive colitis, chronic, continuous symptoms and higher accumulated colectomy rates than the remaining population. Their clinical profile was similar across the IBD subtypes, suggesting that they represent a homogeneous subgroup. In fact, several reports suggest that PSC-IBD represents a separate IBD phenotype altogether 13, making the distinction between CD and UC less relevant.

17

ACCEPTED MANUSCRIPT The female predominance was driven by newly discovered subclinical PSC cases. Interestingly, the preliminary data from the MRC screening study by Bungay et al. in a UK IBD cohort reported a trend towards higher prevalence in women than men 10. Overall, this suggests that women and men with IBD

RI PT

have a similar predisposition to PSC, but women may acquire a milder

phenotype, perhaps caused by hormonal factors. The observations of a later onset of PSC and a lower risk of liver transplantation and death in women

SC

than in men in a large international study support this notion 28, as does a

recent study showing that pregnancies were associated with later onset of

M AN U

PSC 29. A better molecular understanding of the milder PSC phenotype in women could potentially lead to new treatment modalities in PSC.

Another significant observation in the present study is increased prevalence of

TE D

the pANCA autoantibodies in PSC at 10 years. This is the first report of a biomarker (other than liver biochemistry) suggestive of PSC in an IBD population, indicating a possible link between pANCA and PSC development.

EP

One important issue is the lack of elevated liver biochemistry levels, which is normally required for a PSC diagnosis. However, liver biochemistry levels

AC C

within the normal range are not uncommon among PSC patients 30, and normalization of liver biochemistry in PSC is associated with a favorable prognosis 31. The normal liver biochemistry could be explained in part by mild liver involvement observed in subclinical PSC patients, supported by a low MRI progression score 25, which indicates a mild, non- or slowly progressive form of PSC. The extended radiological and clinical follow-up of the new PSC group supports the notion that this is a slowly progressing condition. The main

18

ACCEPTED MANUSCRIPT interpretation of these observations is that even a rare disease like PSC is not a dichotomous condition but exists on a continuum, characterized by a gradual increase in dysfunction 32. PSC is known to have a long subclinical

RI PT

phase of up to 38 years (median, 4.3) 33.

This raises the issue of the benefits of screening for PSC. Diagnosing subclinical PSC in IBD patients may improve understanding of the

SC

pathogenesis of IBD and therefore provide potentially better treatment for these patients. Although no medical therapy has been shown to alter the

M AN U

prognosis of PSC, prompt diagnosis might benefit IBD patients with PSC with regard to monitoring for colon cancer and cholangiocarcinoma, the risk of which has been known to increase in the presence of PSC 11, 12. Whether subclinical PSC is associated with similar risks is unknown. MRC is easy to

TE D

perform, with no side effects and high patient acceptance, making MRC screening for PSC in patients with IBD tempting. Still, based on the current evidence, it is highly questionable whether detection of subclinical PSC is of

EP

any benefit to patients, going against the notion of screening programs for

AC C

PSC in IBD. This could change, however, if effective treatment or preventive measures became available.

The strength of the present study is that it prospectively applies a uniform methodology in a population-based cohort. One essential question concerns the validity and generalizability of the prevalence estimate, because only 54 % of the patients still living 20 years after the initial diagnosis were examined. Given the fact that only 42.6 % of the original cohort was MRC examined, we

19

ACCEPTED MANUSCRIPT believe that the prevalence estimate of 8.1 % represents a minimum number. However, the study of demographic data revealed no statistically significant differences between the initial patient population and the subgroup of patients examined by MRC. This suggests that our results may be valid for the entire

study should be interpreted with some caution.

RI PT

population. Still, the results from a non-random subgroup in an observational

SC

One critical question is whether subclinical MRC lesions should be classified as PSC and receive clinical follow-up accordingly. Available literature

M AN U

suggests that in presumably healthy individuals, the upper limit of extrahepatic, right hepatic and the left hepatic bile duct is 6.0, 3.8 and 3.0 mm, respectively 34, 35. We have adhered to these width limits in the present study and believe that a positive MRC examination indicates pathological bile

TE D

ducts. A meta analysis 17 on MRC performance in diagnosing PSC, supports this perception as a negative MRC would be correct in 95% of the cases and there would be a 5 % risk of a false negative test. Conversely, posttest

EP

probability given a positive MRC would be 84 %. These results indicate that in

AC C

a low prevalence cohort as ours, MRC is an acceptable method to diagnose PSC.

Another limitation is the lack of an objective reference standard for the diagnosis of PSC. The clinical diagnosis is often made with a combination of clinical findings and test results. In this study we used standard cholangiographic features to diagnose PSC 36. Although not pathognomonic, they are highly indicative of sclerosing cholangitis. However, MRC imaging

20

ACCEPTED MANUSCRIPT patterns in patients with PSC are highly variable, probably related to the stage and type of the disease. Easily overlooked minor irregularities and diameter variations are in some cases in fact early signs of PSC. Some patients exhibit rapid radiographic progression and apparent liver damage. This range of

RI PT

cholangiographic findings from subtle to obvious can partly explain the

moderate kappa agreement of 0.55. Additionally, given the mathematical

background for kappa calculation, combined with the low PSC prevalence rate

SC

in the present study, the maximum achievable kappa with perfect agreement would not exceed 0.84. When the readers disagreed on the proportion of

M AN U

positive or negative findings, both an increase in bias index and decrease in kappa appeared. Consequently, prevalence adjusted and bias adjusted kappa was applied 26. Invasive tests, such as liver biopsy or ERC, were not considered appropriate to perform in a screening context due to their invasive

TE D

nature and the consensus of experienced radiologists was the only reasonable basis for a final diagnostic decision.

EP

In summary, in this population-based study of long-term IBD, the overall

AC C

prevalence of MRC lesions indicating PSC in long-term IBD was almost 4 times higher than the prevalence of clinically known PSC prior to MRC screening. Sixty five percent of the detected cases had subclinical disease, seen only on MRC. Extensive colitis and more clinically severe IBD were associated with PSC-like lesions. Despite the low MRI progression score and favorable short-term follow-up outcome, the prognosis of subclinical PSC is uncertain and further research will be needed to clarify the significance of MRC screening in IBD.

21

ACCEPTED MANUSCRIPT

Figure legends: Figure 1. Flowchart of the study population. Abbreviations: IBD = Inflammatory bowel disease; UC = Ulcerative colitis; CD

RI PT

= Crohns disease; LTF = Lost to follow-up; MRC = Magnetic resonance cholangiography.

SC

Figure 2. MRC images of the bile duct lesions. (A) Cholangiogram in a 57years-old male patient with CD and known PSC shows extensive stricturing in

M AN U

the liver hilum (long straight arrow) and marked dilatation in the periphery (arrowheads). (B) Cholangiogram of a 38-year-old male CD patient with known PSC, that was the one false negative exam in the study, shows hilar and intrahepatic strictures (long straight arrow) at second glance. (C)

TE D

Stricturing of the common bile duct and intrahepatic ductules (long straight arrow), intrahepatic bile duct dilatation (arrowheads), mural irregularities (short arrow) and beading (curved arrow) in a 53-year-old male UC patient (D)

EP

Stricturing at the confluence of the left bile duct and the common bile duct as

AC C

well as intrahepatically (long straight arrow) in a 73-year-old female CD patient. The arrowhead points at the dilatation of the peripheral bile ducts in the left liver lobe. (E) Stricturing at the confluence of the left bile duct and the common bile duct (long straight arrow) in a 44-year-old female patient with UC. The arrowheads demonstrate dilatation of the peripheral bile ducts. (F) Stricture (long straight arrow) and dilatation (arrowhead) of right intrahepatic bile ducts in a 43-year-old male CD patient with ileitis only.

22

ACCEPTED MANUSCRIPT Abbreviations: MRC = Magnetic resonance cholangiography; PSC = Primary sclerosing Cholangitis; UC = Ulcerative colitis; CD = Crohn´s disease.

Figure 3. MRC Flowchart of different PSC subgroups.

RI PT

Abbreviations: IBD = Inflammatory bowel disease; UC = Ulcerative colitis; CD = Crohns disease; PSC = Primary sclerosing cholangitis; MRC = Magnetic

SC

resonance cholangiography.

Figure 4. Mean liver biochemistry levels during the follow-up period in

M AN U

patients with known PSC (n=9), new PSC (n=17) and no PSC (n=296). A. Levels of alanine transaminase (ALAT). B. Levels of alkaline phosphatase (AFOS) at 0, 1, 5 and 10 years. C. Levels of AFOS at 20 years. The AFOS graphs are split in two due to the introduction of a new method for measuring ALP in 2003

2003: <105 U/L.

TE D

(i.e. between 10 and 20 years follow-up). Reference values before 2003: 70-230 U/L and after

Table 1. Demographic and clinical characteristics of the study

EP

population at 20-year follow-up. Comparison between the MRC-examined

AC C

patients and the non MRC-examined patients. Abbreviations: MRC = Magnetic resonance cholangiography; SD = Standard deviation; IBD = Inflammatory bowel disease.

Table 2. Demographic and clinical characteristics of the non-PSC group and the PSC group. The PSC group consists of cases with evident PSC-like lesions on MRC (n=24) and known PSC cases (n=2) that were either missed (1) or had small-duct PSC (1). *significant P-values

23

ACCEPTED MANUSCRIPT Abbreviations: PSC = primary sclerosing cholangitis; SD = Standard deviation; IBD = Inflammatory bowel disease; ANCA = Anti-neutrophil cytoplasmic antibodies.

RI PT

Table 3. Demographic and clinical characteristics of the PSC group

according to time and mode of diagnosis. New PSC group represents

patients with bile duct lesions on MRC, i.e. putative PSC. Known PSC group

PSC patient. *significant P-values

SC

consists of patients with clinically recognized PSC, including one small-duct

M AN U

Abbreviations: PSC = primary sclerosing cholangitis; SD = Standard deviation; IBD = Inflammatory bowel disease; ANCA = Anti-neutrophil cytoplasmic antibodies; MRI = Magnetic resonance imaging.

3. 4.

5.

6.

7.

EP

2.

Hirschfield GM, Karlsen TH, Lindor KD, et al. Primary sclerosing cholangitis. The Lancet 2013;382:1587-1599. Fausa O, Schrumpf E, Elgjo K. Relationship of inflammatory bowel disease and primary sclerosing cholangitis. Semin Liver Dis 1991;11:31-9. Broome U, Bergquist A. Primary sclerosing cholangitis, inflammatory bowel disease, and colon cancer. Semin Liver Dis 2006;26:31-41. Wewer V, Gluud C, Schlichting P, et al. Prevalence of hepatobiliary dysfunction in a regional group of patients with chronic inflammatory bowel disease. Scand J Gastroenterol 1991;26:97-102. Schrumpf E, Fausa O, Kolmannskog F, et al. Sclerosing cholangitis in ulcerative colitis. A follow-up study. Scand J Gastroenterol 1982;17:339. Parlak E, Kosar Y, Ulker A, et al. Primary sclerosing cholangitis in patients with inflammatory bowel disease in Turkey. J Clin Gastroenterol 2001;33:299-301. Tobias R, Wright JP, Kottler RE, et al. Primary sclerosing cholangitis associated with inflammatory bowel disease in Cape Town, 1975 1981. S Afr Med J 1983;63:229-35.

AC C

1.

TE D

References:

24

ACCEPTED MANUSCRIPT

13.

14.

15.

16. 17.

18.

AC C

19.

RI PT

12.

SC

11.

M AN U

10.

TE D

9.

Bernstein CN, Blanchard JF, Rawsthorne P, et al. The prevalence of extraintestinal diseases in inflammatory bowel disease: a populationbased study. Am J Gastroenterol 2001;96:1116-22. McGarity B BD, Robertson DA, Millward-Sadler GH, Shepherd HA. Primary sclerosing cholangitis: an important and prevalent complication of Crohn's disease. Eur J Gastroenterol Hepatol 1991:361-364. Bungay H, Buchel O, Cummings F, et al. Prevalence and determinants of primary sclerosing cholangitis in a cohort of patients with inflammatory bowel disease and normal liver function tests. Gut: John Radcliffe Hospital, Oxford, UK, 2008:(Suppl I):A1–A172. Torres J, Pineton de Chambrun G, Itzkowitz S, et al. Review article: colorectal neoplasia in patients with primary sclerosing cholangitis and inflammatory bowel disease. Aliment Pharmacol Ther 2011;34:497508. Burak K, Angulo P, Pasha TM, et al. Incidence and risk factors for cholangiocarcinoma in primary sclerosing cholangitis. Am J Gastroenterol 2004;99:523-6. Loftus EV, Jr., Harewood GC, Loftus CG, et al. PSC-IBD: a unique form of inflammatory bowel disease associated with primary sclerosing cholangitis. Gut 2005;54:91-6. Liu JZ, Hov JR, Folseraas T, Ellinghaus E, et al. Dense genotyping of immune-related disease regions identifies nine new risk loci for primary sclerosing cholangitis. Nat Genet 2013;45:670-5. Alexopoulou E, Xenophontos PE, Economopoulos N, et al. Investigative MRI cholangiopancreatography for primary sclerosing cholangitis-type lesions in children with IBD. J Pediatr Gastroenterol Nutr 2012;55:308-13. Engelmann G. Does screening for subclinical PSC contribute to the management of IBD? J Pediatr Gastroenterol Nutr 2012;55:238. Dave M, Elmunzer BJ, Dwamena BA, et al. Primary sclerosing cholangitis: meta-analysis of diagnostic performance of MR cholangiopancreatography. Radiology 2010;256:387-96. Solberg IC, Vatn MH, Hoie O, et al. Clinical course in Crohn's disease: results of a Norwegian population-based ten-year follow-up study. Clin Gastroenterol Hepatol 2007;5:1430-8. Solberg IC, Lygren I, Jahnsen J, et al. Clinical course during the first 10 years of ulcerative colitis: results from a population-based inception cohort (IBSEN Study). Scand J Gastroenterol 2009;44:431-40. Moum B, Vatn MH, Ekbom A, et al. Incidence of inflammatory bowel disease in southeastern Norway: evaluation of methods after 1 year of registration. Southeastern Norway IBD Study Group of Gastroenterologists. Digestion 1995;56:377-81. Hovde O, Kempski-Monstad I, Smastuen MC, et al. Mortality and causes of death in Crohn's disease: results from 20 years of follow-up in the IBSEN study. Gut 2014;63:771-5. Hovde O, Smastuen MC, Hoivik ML, et al. Mortality and Causes of Death in Ulcerative Colitis: Results from 20 Years of Follow-up in the IBSEN Study. Inflamm Bowel Dis 2016;22:141-5.

EP

8.

20.

21.

22.

25

ACCEPTED MANUSCRIPT

28.

29.

30. 31.

32. 33.

AC C

34.

RI PT

27.

SC

26.

M AN U

25.

TE D

24.

Solberg IC, Lygren I, Cvancarova M, et al. Predictive value of serologic markers in a population-based Norwegian cohort with inflammatory bowel disease. Inflamm Bowel Dis 2009;15:406-14. Fulcher AS, Turner MA, Franklin KJ, et al. Primary sclerosing cholangitis: evaluation with MR cholangiography-a case-control study. Radiology 2000;215:71-80. Ruiz A, Lemoinne S, Carrat F, et al. Radiologic course of primary sclerosing cholangitis: assessment by three-dimensional magnetic resonance cholangiography and predictive features of progression. Hepatology 2014;59:242-50. Byrt T, Bishop J, Carlin JB. Bias, prevalence and kappa. J Clin Epidemiol 1993;46:423-9. Ronnblom A, Holmstrom T, Tanghoj H, et al. Appearance of hepatobiliary diseases in a population based cohort with inflammatory bowel diseases (ICURE). J Gastroenterol Hepatol 2015. Weismuller TJ, Hansen BE, Trivedi PJ, et al. Gender and IBD phenotype are independent predictors of death or transplantation and of malignancy in Primary Sclerosing Cholangitis – a multicenter retrospective study of the International PSC Study Group (IPSCSG). AASLD Abstracts. Hepatology, 2015:246A. Andersen IM, Tengesdal G, Lie BA, et al. Effects of coffee consumption, smoking, and hormones on risk for primary sclerosing cholangitis. Clin Gastroenterol Hepatol 2014;12:1019-28. Aadland E, Schrumpf E, Fausa O, et al. Primary sclerosing cholangitis: a long-term follow-up study. Scand J Gastroenterol 1987;22:655-64. Lindstrom L, Hultcrantz R, Boberg KM, et al. Association between reduced levels of alkaline phosphatase and survival times of patients with primary sclerosing cholangitis. Clin Gastroenterol Hepatol 2013;11:841-6. Plomin R, Haworth CM, Davis OS. Common disorders are quantitative traits. Nat Rev Genet 2009;10:872-8. Broome U, Olsson R, Loof L, et al. Natural history and prognostic factors in 305 Swedish patients with primary sclerosing cholangitis. Gut 1996;38:610-5. Peng R, Zhang L, Zhang XM, et al. Common bile duct diameter in an asymptomatic population: A magnetic resonance imaging study. World J Radiol 2015;7:501-8. Matcuk GR, Jr., Grant EG, Ralls PW. Ultrasound measurements of the bile ducts and gallbladder: normal ranges and effects of age, sex, cholecystectomy, and pathologic states. Ultrasound Q 2014;30:41-8. Ito K, Mitchell DG, Outwater EK, et al. Primary sclerosing cholangitis: MR imaging features. AJR Am J Roentgenol 1999;172:1527-33.

EP

23.

35.

36.

Author names in bold designate shared co-first authorship.

26

ACCEPTED MANUSCRIPT Table 1.

Non MRC-examined,

MRC-examined

N=470

N=148

N=322

Female, n (%)

82 (55.4)

153 (47.5)

.14

.51

Age (years) Mean (SD)

54.1 (14.5)

53.2 (11.3)

-

Median (range)

51.2 (27.4-94.0)

51.7 (29.4-84.5)

222 (68.9)

-

Ulcerative colitis

92 (62.2)

-

Crohn´s disease

56 (37.8)

Ulcerative colitis – location n (%) Proctitis

15 (16.3)

Left-sided

32 (34.8)

100 (31.1)

69 (31.1)

45 (48.9)

112 (50.5)

11 (19.6)

17 (17.0)

22 (39.3)

29 (29.0)

23 (41.4)

54 (54.0)

penetrating

26 (46.4)

50 (50.0)

-

B2, stricturing

26 (46.4)

41 (41.0)

-

B3, penetrating

4 (7.1)

9 (9.0)

-

Extensive

.17

41 (18.5)

M AN U

-

SC

-

IBD type, n (%)

P value

RI PT

Characteristics/Groups

.80

Crohn´s disease – location, n (%) -

L1, terminal ileum

-

L2, isolated colon

-

L3, ileocolon

.29

Crohn´s disease – behavior, n (%)

TE D

B1, non-stricturing non-

AC C

EP

-

.82

ACCEPTED MANUSCRIPT Table 2.

Variable/Group

Non-PSC group,

PSC-group,

n=296

n=26

P-value

Gender: Female (%)

139 (47.0)

14 (53.8)

-

Male (%)

157 (53.0)

12 (46.2)

.54

53.0 (11.1)

55.1 (12.5)

.36

Mean age (SD) IBD type:

RI PT

-

Ulcerative colitis (%)

207 (69.9)

15 (57.7)

-

Crohn´s disease (%)

89 (30.1)

11 (42.3)

Ulcerative colitis – location -

Proctitis and left-sided

107 (51.7)

-

Extensive

100 (48.3)

-

L1, terminal ileum

-

L2, isolated colon

-

L3, ileocolon

Patterns of IBD disease -

Remission

.27

3 (20.0)

12 (80.0)

M AN U

Crohn´s disease – location

SC

-

16 (18.0)

1 (9.1)

26 (29.2)

5 (45.5)

47 (52.8)

5 (45.5)

198 (68.0)

13 (50.0)

.029*

.57

Increase in severity

8 (2.7)

2 (7.7)

Chronic continous symptoms

24 (8.2)

7 (26.9)*

-

Chronic intermittent symptoms

61 (21.0)

4 (15.4)

54 (18.7)

6 (23.1)

0.60

Prevalence of colectomy

23 (7.8)

8 (30.8)

.001*

Colectomized due to malignancy

1 (4.3)

3 (37.5)

.016*

History of cholecystectomy

8 (2.7)

1 (3.8)

.73

Positive ANCA at 10 years

68 (25.6)

12 (57.1)

.013*

Use of steroids 0-5 years

144 (49.0)

18 (75.0)

.018*

AC C

EP

Current smoker

TE D

-

.012*

ACCEPTED MANUSCRIPT Table 3.

Variable /Group

New PSC, n=17

Known PSC, n=9

P-value

Gender: Female (%)

12 (70.6)

2 (22.2)

-

Male (%)

5 (29.4)

7 (77.8)

.038*

56.9 (12.9)

51.8 (11.8)

.34

Mean age (SD) IBD type:

RI PT

-

Ulcerative colitis (%)

12 (70.6)

3 (33.3)

-

Crohn´s disease (%)

5 (29.4)

6 (66.7)

0 (0.0)

Ulcerative colitis – location -

Proctitis and left-sided

3 (25.0)

-

Extensive

9 (75.0)

Crohn´s disease – location L1, terminal ileum

-

L2, isolated colon

-

L3, ileocolon

Patterns of IBD disease -

Remission

1 (20.0)

3 (100.0)

.10

1.00

0 (0.0)

M AN U

-

SC

-

2 (40.0)

3 (50.0)

2 (40.0)

3 (50.0)

8 (47.1)

5 (55.6)

1.00

-

Increase in severity

2 (11.8)

0 (0.0)

-

Chronic continuous symptoms

4 (23.5)

3 (33.3)

Chronic intermittent symptoms

3 (17.6)

1 (11.1)

.87

5 (29.4)

1 (11.1)

.38

-

TE D

Current smoker

5 (29.4)

3 (33.3)

1.00

Positive pANCA at 10 years

5 (41.7)

7 (77.8)

.18

Use if steroids 0-5 years

12 (75.0)

6 (75.0)

1.00

Mean MRI progression risk score (SD)

0.94 (0.75)

2.13 (1.89)

.076

AC C

EP

Prevalence of colectomy

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT APPENDICES Appendix 1. Technical parameters utilized for MRC sequences acquisition – Philips T2weighted TSE FS

TR (ms) TE (ms) Slice thickness/gap (mm) Orientation Field of view

394-409 40 3/1

496-509 80 3/1

Matrix

Coronal 380 x 332-333 216 x 208

Breath hold

Yes

Axial 298-300 x 375 340 x 234238 Yes

T1weighted in- and out of fase 112-113 2.3/4.6 5/1

MRCP single shot 2D

MRCP radial stack (thick slab)

MRCP 3D

1082-1091 330 3/0.9

8000 800 40

1204-2110 650-722 1,02-1.7 x 1.27-1.5 x 0.8

Axial 375/297

Coronal 350/327

288 x 229

240 x 178

Yes

Yes

RI PT

T2-weighted TSE

Coronal 300 x 300

320 x 256

Yes

AC C

EP

TE D

M AN U

SC

Sequence

Coronal 260-300 x 260-267 256 x 182-205 No/ respiration triggered

ACCEPTED MANUSCRIPT Appendix 2. Demographic and clinical characteristics at baseline for patients still alive at 20-year follow-up. p

One way analysis of variance (ANOVA).

Abbreviations: MRC = Magnetic resonance cholangiography; SD = Standard deviation; IBD = Inflammatory bowel disease; UC = Ulcerative colitis; CD = Crohn´s disease; IC= Indeterminate colitis.

Non MRC examined N=148

MRC examined N=322

Alive, non participating N=129

P-value

Female, n (%) Age (years) Mean (SD) Median (range) IBD type at diagnosis, n (%) UC CD IC Possible IBD UC – location, n (%) Proctitis Left-sided Extensive CD – location, n (%) L1, terminal ileum L2, isolated colon L3, ileocolon L4, upper GI CD – behavior, n (%) B1, non-stricturing non-penetrating B2, stricturing B3, penetrating

82 (55.4)

153 (47.5)

65 (50.4)

0.28

33.9 (14.8) 30.9 (5.4-72.4)

33.0 (11.2) 31.1 (9.8-63.9)

33.4 (14.8) 30.2 (3.8-69.6)

0.81p

91 (61.5) 48 (32.4) 4 (2.7) 5 (3.4)

207 (64.3) 89 (27.6) 13 (4.0) 13 (4.0)

79 (61.2) 40 (31.0) 6 (4.7) 4 (3.1)

0.92

34 (37.4) 25 (27.5) 32 (35.2)

71 (34.3) 65 (31.4) 71 (34.3)

32 (40.5) 22 (27.8) 25 (31.6)

0.87

16 (33.3) 21 (43.8) 10 (20.8) 1 (2.1)

26 (29.2) 35 (39.3) 26 (29.2) 2 (2.2)

13 (33.5) 15 (37.5) 11 (27.5) 1 (2.5)

0.97

31 (64.6) 12 (25.0) 5 (10.4)

52 (58.4) 25 (28.1) 12 (13.5)

18 (45.0) 19 (47.5) 3 (7.5)

0.18

TE D

M AN U

SC

RI PT

Characteristics/Groups N=599

Appendix 3. Distribution of bile duct lesions in 24 IBD patients Stricture

Dilatation

Wall irregularities

Beading

Any lesions

Isolated extrahepatic, n (%)

0 (0)

1 (4.2)

1 (4.2)

0 (0)

1 (4.2)

Isolated intrahepatic, n (%)

7 (29.2)

13 (54.2)

8 (33.3)

5 (20.8)

16 (66.7)

9 (37.5)

5 (20.8)

11 (45.8)

1 (4.2)

13 (54.2)

AC C

EP

Distribution/Findings

Both intra- and

extrahepatic, n (%)

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Appendix 4. Four predefined curves depicting different IBD activity courses (19).

ACCEPTED MANUSCRIPT Appendix 5. Medication use during the follow-up period. Table shows proportions and percentages (in parenthesis) of patients that have used different types of medication during the entire course of follow-up for the PSC group and the non PSC group. The analysis is based on approximately 93.8 %, 96.3 % and 98.8 % of the MRC study population at 5, 10 and 20 years respectively. PSC group, n=26

Non PSC group, n=296

21/24 (87.5) 18/24 (75.0) 3/24 (12.5)

244/289 (84.4) 141/288 (49.0) 18/288 (6.3)

1.00 0.01* 0.21

Accumulated use 5-10 years: 5-ASA Steroids Azathioprine

16/25 (64.0) 7/25 (28.0) 4/25 (16.0)

169/285 (59.3) 78/285 (27.4) 28/285 (9.8)

0.68 1.00 0.31

Accumulated use 10-20: 5-ASA Steroids Azathioprine

9/26 (34.6) 10/26 (38.5) 7/26 (26.9)

152/293 (51.9) 100/294 (34.0) 50/292 (17.1)

Medication use at 20 years: 5-ASA Steroids Azathioprine

7/26 (26.9) 2/26 (7.7) 2/26 (7.7)

AC C

EP

TE D

1/26 (3.8)

SC

M AN U

Accumulated biological treatment 0-20 years

P-value

RI PT

Medication/Groups Accumulated use 0-5 years: 5-ASA Steroids Azathioprine

0.10 0.67 0.28

114/295 (38.6) 20/293 (6.8) 32/294 (10.9)

0.29 0.70 1.00

28/292 (9.6)

0.50

ACCEPTED MANUSCRIPT Appendix 6. Extraintestinal manifestations at different time points of the follow-up. Table shows proportions and percentages (in parenthesis) of patients that have had extraintestinal manifestations like skin, eye and joint symptoms for the PSC group and the non PSC group. The analysis is based on approximately 49.1 %, 95.6 % and 67.7 % of the MRC study population at 5, 10 and 20 years. respectively.

PSC group, n=26

Non PSC group, n=296

P-value

RI PT

Manifestations/Groups Skin disease: -

5 years

1/19 (5.3)

22/135 (16.3)

0.21

-

10 years

0/24 (0)

11/284 (3.9)

0.33

-

20 years

2/15 (13.3)

22/203 (10.8)

0.77

15/138 (10.9)

0.13

16/285 (5.6)

0.23

11/203 (5.4)

0.84

-

5 years

0/19 (0)

-

10 years

0/24 (0)

-

20 years

1/15 (6.7)

M AN U

Rheumatologic disease

SC

Eye disease:

-

5 years

11/20 (55.0)

81/143 (56.6)

0.89

-

10 years

8/26 (30.7)

85/296 (28.7)

0.82

-

20 years

3/15 (20.0)

39/204 (19.1)

0.93

Appendix 7. Mean liver biochemistry levels during the follow-up period.

TE D

Table shows mean levels and standard deviation (in parenthesis) of alanine transaminase (ALT) and alkaline phosphatase (ALP) at baseline, 1, 5, 10 and 20 years of follow-up. A one way analysis of variance (ANOVA) testing with bootstrapping was applied. The analysis is based on approximately 90 %, 74 %, 81 %, 79 % and 95 % of the MRC study population at 0, 1, 5, 10 and 20 years respectively. Reference values: ALT < 45-70 U/L; ALP <105 U/L at 20

EP

years, <230 U/L at 0,1,5,10 years.

AC C

Blood tests/Patient groups (n=322) Mean ALT (U/L): Baseline 1 year 5 years 10 years 20 years Mean ALP (U/L): Baseline 1 year 5 years 10 years 20 years

Known PSC (n=9)

New PSC (n=17)

No PSC (n=296)

Pvalue

122.7 (138.6)* 107.5 (145.6)* 114.5 (157.2)* 117.6 (113.7)* 68.2 (54.2)*

20.6 (14.7) 23.2 (7.4) 23.5 (8.6) 24.0 (11.4) 23.8 (7.5)

24.7 (28.1) 25.6 (16.0) 27.1 (26.4) 26.4 (19.2) 26.8 (15.7)

<0.001* <0.001* <0.001* <0.001* <0.001*

762.6 (970.9)* 534.3 (551.1)* 551.1 (501.6)* 595.1 (462.9)* 165.4 (97.0)*

172.3 (100.4) 164.7 (77.8) 136.7 (39.0) 138.0 (41.0) 70.6 (22.7)

188.2 (89.6) 179.6 (90.8) 165.4 (59.5) 180.7 (241.5) 72.4 (23.3)

<0.001* <0.001* <0.001* <0.001* <0.001*

ACCEPTED MANUSCRIPT Appendix 8. Selected clinical data registered at the 20-year follow-up for for the PSC group and the non PSC group. Abbreviations: PSC=Primary sclerosing cholangitis; HBI= Harvey-Bradshaw Index; SCCAI= Simple Clinical Colitis Activity Index; SD = Standard deviation; IBD = Inflammatory bowel disease; UC = Ulcerative colitis; CD = Crohn´s

RI PT

disease; IC= Indeterminate colitis.

Variable/Group

Non-PSC group, n=296

PSC-group, n=26

P-value

Colorectal cancer

4 (1.4)

1 (3.8)

0.35

HBI (>5)

24 (33.3)

2 (28.6)

SCCAI (>2.5)

42 (20.8)

5 (35.7)

Mayo score (>/=2)

21 (13.7)

2 (15.4)

Calprotectin (>/= 50 mg/kg)

117 (48.1)

Flares in the last year of follow-up

71 (24.1)

Chronic fatigue

59 (20.9)

EP AC C

0.19 1.00

0.38

5 (20.0)

0.80

7 (28.0)

0.45

7 (26.9)

0.29

M AN U

52 (17.6)

TE D

stricturoplasty (except for colectomy)

SC

Prevalence of bowel resection and

13 (59.1)

1.00