Risk Factors for Primary Sclerosing Cholangitis

Kirsten Boonstra; Elisabeth M. G. de Vries; Nan van Geloven; Karel J. van Erpecum; Marcel Spanier; Alexander C. Poen; Carin M. van Nieuwkerk; Ben J. Witteman; Hans A. Tuynman; Anton H. Naber; Paul J. Kingma; Ulrich Beuers; Cyriel Y. Ponsioen


Liver International. 2016;36(1):84-91. 

In This Article

Patients and Methods

For this case–control study, PSC patients were recruited from the Epi PSC PBC project; a large population-based observational longitudinal cohort study of PSC and primary biliary cirrhosis (PBC) in the Netherlands. IBD controls and healthy controls were also included in this cohort.

Study Design and Ethical Consideration

The protocol of the Epi PSC PBC project was approved by the Central Committee for Research Ethics in Utrecht and all 44 local ethics committees of the participating hospitals in the Netherlands (trialregister.nl, NTR2813). This Epi PSC PBC observational longitudinal cohort study was undertaken between 1 January 2008, and 31 December 2011. All PSC patients in 44 hospitals from 2000 onwards were identified in a geographically defined area of 6 adjacent provinces comprising 50% of the Dutch population (2000: 7 342 295 – 2007: 7 758 980) (Fig. S1 http://onlinelibrary.wiley.com/doi/10.1111/liv.12894/suppinfo). This area consists of 169 municipalities subdivided in five categories by degree of urbanization; 1 ≥ 2500, 2; 1500–2500; 3; 1000–1500, 4; 500–1000, 5; <500 residences per km2. The 44 participating hospitals provided care to the entire population of the study area. Healthy controls (HC) and IBD patients were randomly recruited from the outpatient clinics of four participating hospitals, evenly spread throughout the study area (Fig. S1 http://onlinelibrary.wiley.com/doi/10.1111/liv.12894/suppinfo).

Case-finding and Case-ascertainment

The case-finding and case-ascertainment methods have been described previously.[1]

Primary sclerosing cholangitis diagnosis was based on: (i) elevated alkaline phosphatase and gamma-glutamyltransferase, not explained otherwise; (ii) presence of characteristic bile duct changes with multifocal strictures and segmental dilatations on endoscopic retrograde cholangiography (ERC) or magnetic resonance cholangiography (MRC) and/or; (iii) liver histology; and (iv) no evidence for secondary sclerosing cholangitis. When criteria 1, 3 and 4 were fulfilled in the absence of cholangiographic abnormalities on MRC or ERC, cases were diagnosed as small duct PSC.[25] For this case–control study, PSC patients with concomitant IBD were subdivided in PSC-UC and PSC-CD subgroups, in which PSC-IBD unspecified (IBD-U) patients were grouped under PSC-UC.

Subjects visiting an outpatient clinic without a history of liver disease or autoimmune disease were considered to be healthy controls. These volunteers were healthy individuals accompanying PSC, PBC and IBD patients to the outpatient clinic, or visiting the outpatient clinic for regular surveillance of Barrett's oesophagus or colon polyps. IBD diagnosis was based on the Lennard-Jones criteria.[26] Similar to the PSC patients, IBD control patients were subdivided in UC and CD patients, IBD-U patients were grouped under UC. To be able to serve as genuine control group, IBD patients could not have any diagnosis of liver disease and had to have normal routine liver tests.

Data Collection

At inclusion of the Epi PSC PBC project, the following data were collected from hospital patient files of PSC patients: date and type of PSC diagnosis, zip code of residence at time of diagnosis, date of diagnosis and type of IBD, smoking behaviour and date and type of appendectomy. Hospital files of IBD controls were reviewed to verify IBD diagnosis, and exclude patients with history of liver disease or abnormal routine liver tests. If available, hospital files of HC's were checked as well, if not, history of liver disease was excluded anamnestically. PSC patients, IBD patients and HC's were asked to give informed consent and to fill out a questionnaire regarding smoking status (current smoking, defined as smoking at time of inclusion and former smoking, defined as smoking more than 1 year in the past 20 years), history of appendectomy and first-degree family history of IBD and autoimmune liver disease. Autoimmune liver diseases included primary sclerosing cholangitis, primary biliary cirrhosis or autoimmune hepatitis. For PSC patients, data on history of appendectomy reported in questionnaires were completed by data on history of appendectomy from hospital files. Population estimates were retrieved from the Dutch Central Office for Statistics (Centraal Bureau voor Statistiek, Den Haag, the Netherlands; http://statline.cbs.nl).

Statistical Analysis

Potential risk factors (sex, age, smoking behaviour, appendectomy, family history of IBD and autoimmune liver disease) for the development of PSC were analysed by comparing the relative frequencies between PSC patients and healthy controls as well as between PSC-IBD patients and IBD controls using logistic regression models. The few cases, in which response to an item from the questionnaire was missing, were excluded from the pertaining analyses. Only appendectomies performed before PSC diagnosis was established were included in the analysis. Relative frequencies were expressed as odds ratio (OR), with corresponding 95% confidence intervals (CI). To differentiate risk factors for the development of PSC from risk factors for IBD, PSC-IBD patients and IBD controls were matched for type of IBD; PSC-UC patients were compared with UC patients, PSC-CD patients were compared with CD patients. In addition, PSC patients without IBD were analysed separately compared with HC's. After assessing the association between risk factors and disease with univariable logistic regression, multivariable logistic regression modelling was used to find independent risk factors. All relevant risk factors were entered in the multivariable analysis. Differences in age at inclusion and gender between groups were adjusted for in the multivariable analysis. Given the overlap in current smoking and former smoking, only former smoking was considered for multivariable analysis.

For analysis of geographical distribution, all PSC patients from the Epi PSC PBC cohort that were diagnosed between 2000 and 2008 were included. PSC incidence was calculated for the sum of municipalities per degree of urbanization, using the mean population between 2000 and 2008. Standardized incidence ratios were calculated to describe differences between observed and expected incidence of PSC per degree of urbanization. Cochran–Armitage test for trend in proportions was used to determine if there was statistical evidence for increasing or decreasing incidence over degrees of urbanization. Statistical analyses were performed using spss version 20.0 software (SPSS, Inc., Chicago, IL. USA) and r version 2.13.1 (R Foundation for Statistical Computing, Vienna, Austria). P < 0.05 was considered statistically significant.