Steven Jeuring

Disease Outcome of Ulcerative Colitis in an Era of Changing Treatment Strategies: Results from the Dutch Population-based IBDSL Cohort

BACKGROUND AND AIMS In the past decades, treatment options and strategies for ulcerative colitis [UC] have radically changed. Whether these developments have altered the disease outcome at population level is yet unknown. Therefore, we evaluated the disease outcome of UC over the past two decades in the South-Limburg area of The Netherlands. METHODS In the Dutch population-based IBDSL cohort, three time cohorts were defined: cohort 1991-1997 [cohort A], cohort 1998-2005 [cohort B], and cohort 2006-2010 [cohort C]. The colectomy and hospitalisation rates were compared between cohorts by Kaplan-Meier survival analyses. Hazard ratios [HR] for early colectomy [within 6 months after diagnosis], late colectomy [beyond 6 months after diagnosis], and hospitalisation were calculated using Cox regression models. RESULTS In total, 476 UC patients were included in cohort A, 587 patients in cohort B, and 598 patients in cohort C. Over time, an increase in the use of immunomodulators [8.1%, 22.8% and 21.7%, respectively, p < 0.01] and biological agents [0%, 4.3% and 10.6%, respectively, p < 0.01] was observed. The early colectomy rate decreased from 1.5% in cohort A to 0.5% in cohort B [HR 0.14; 95% confidence interval 0.04-0.47], with no further decrease in cohort C [0.3%, HR 0.98; 95% confidence interval 0.20-4.85]. Late colectomy rate remained unchanged over time [4.0% vs 5.2% vs 3.6%, respectively, p = 0.54]. Hospitalisation rate was also similar among cohorts [22.3% vs 19.5% vs 18.3%, respectively, p = 0.10]. CONCLUSION Over the past two decades, a reduction in early colectomy rate was observed, with no further reduction in the most recent era. Late colectomy rate and hospitalisation rate remained unchanged over time.

Improvements in the Long-Term Outcome of Crohn's Disease Over the Past Two Decades and the Relation to Changes in Medical Management: Results from the Population-Based IBDSL Cohort

Objectives:Medical treatment options and strategies for Crohn’s disease (CD) have changed over the past decades. To assess its impact, we studied the evolution of the long-term disease outcome in the Dutch Inflammatory Bowel Disease South Limburg (IBDSL) cohort.Methods:In total, 1,162 CD patients were included. Three eras were distinguished: 1991–1998 (n=316), 1999–2005 (n=387), and 2006–2011 (n=459), and patients were followed until 2014. Medication exposure and the rates of hospitalization, surgery, and phenotype progression were estimated using Kaplan–Meier survival analyses and compared between eras by multivariable Cox regression models. Second, propensity score matching was used to assess the relation between medication use and the long-term outcome.Results:Over time, the immunomodulator exposure rate increased from 30.6% in the era 1991–1998 to 70.8% in the era 2006–2011 at 5 years. Similar, biological exposure increased from 3.1% (era 1991–1998) to 41.2% (era 2006–2011). In parallel, the hospitalization rate attenuated from 65.9% to 44.2% and the surgery rate from 42.9% to 17.4% at 5 years, respectively (both P<0.01). Progression to a complicated phenotype has not changed over time (21.2% in the era 1991–1998 vs. 21.3% in the era 2006–2011, P=0.93). Immunomodulator users had a similar risk of hospitalization, surgery, or phenotype progression as propensity score-matched nonusers (P>0.05 for all analyses). Similar results were found for biological users (P>0.05 for all analyses).Conclusions:Between 1991 and 2014, the hospitalization and surgery rates decreased, whereas progression to complicated disease is still common in CD. These improvements were not significantly related to the use of immunomodulators and biologicals.

Cohort Profile: The Inflammatory Bowel Disease South Limburg Cohort (IBDSL)

Cohort Profile: The Inflammatory Bowel Disease South Limburg Cohort (IBDSL) Tim RA van den Heuvel,* Daisy M Jonkers, Steven FG Jeuring, Marielle JL Romberg-Camps, Liekele E Oostenbrug, Maurice P Zeegers, Ad A Masclee and Marie J Pierik Department of Internal Medicine, NUTRIM School for Nutrition and Translational Research in Metabolism, Maastricht University Medical Centreþ, Maastricht, The Netherlands, Department of Gastroenterology and Hepatology, Orbis Medical Centre, Sittard-Geleen, The Netherlands and Department of Gastroenterology and Hepatology, Atrium Medical Centre, Heerlen, The Netherlands

Inflammatory bowel disease, cancer and medication: Cancer risk in the Dutch population-based IBDSL cohort.

The management of inflammatory bowel disease (IBD) has changed since the mid‐1990s (e.g., use of thiopurines/anti‐TNFα agents, improved surveillance programs), possibly affecting cancer risk. To establish current cancer risk in IBD, updates are warranted from cohorts covering this time span, and detailed enough to study associations with phenotype and medication. We studied intestinal‐, extra‐intestinal‐ and overall cancer risk in the Dutch population‐based IBDSL cohort. In total, 1,157 Crohns disease (CD) and 1,644 ulcerative colitis (UC) patients were diagnosed between 1991 and 2011, and followed until 2013. Standardized incidence ratios (SIRs) were calculated for CD and UC separately, as well as for gender‐, phenotype‐, disease duration‐, diagnosis era‐ and medication groups. We found an increased risk for colorectal cancer in CD patients with colon involvement (SIR 2.97; 95% CI 1.08–6.46), but not in the total CD or UC population. In addition, CD patients were at increased risk for hematologic‐ (2.41; 1.04–4.76), overall skin‐ (1.55; 1.06–2.19), skin squamous cell‐ (SCC; 3.83; 1.83–7.04) and overall cancer (1.28; 1.01–1.60), whereas UC patients had no increased risk for extra‐intestinal‐ and overall cancer. Finally, in a medication analysis on CD and UC together, long‐term immunosuppression exposure (>12 months) was associated with an increased risk for hematologic cancer, non‐Hodgkin lymphoma, SCC and overall cancer, and this increase was mainly attributed to thiopurines. IBD patients with long‐term immunosuppression exposure can be considered as having a higher cancer risk, and our data support the advice in recent IBD guidelines to consider skin cancer screening in these patients.

Time trends in the epidemiology and outcome of perianal fistulizing Crohn’s disease in a population-based cohort

Objective Perianal disease is a debilitating condition that frequently occurs in Crohn’s disease (CD) patients. It is currently unknown whether its incidence has changed in the era of frequent immunomodulator use and biological availability. We studied the incidence and outcome of perianal and rectovaginal fistulas over the past two decades in our population-based Inflammatory Bowel Disease South-Limburg cohort. Patients and methods All 1162 CD patients registered in the Inflammatory Bowel Disease South-Limburg registry were included. The cumulative probabilities of developing a perianal and rectovaginal fistula were compared between three eras distinguished by the year of CD diagnosis: 1991–1998, 1999–2005 and 2006–2011. Second, clinical risk factors and the risk of fistula recurrence were determined. Results The cumulative 5-year perianal fistula rate was 14.1% in the 1991–1998 era, 10.4% in the 1999–2005 era and 10.3% in the 2006–2011 era, P=0.70. Colonic disease was associated with an increased risk of developing perianal disease, whereas older age was associated with a decreased risk (both P<0.01). Over time, more patients were exposed to immunomodulators or biologicals before fistula diagnosis (18.5 vs. 32.1 vs. 52.1%, respectively, P=0.02) and started biological therapy thereafter (18.6 vs. 34.1 vs. 54.0%, respectively, P<0.01). The cumulative 5-year perianal fistula recurrence rate was not significantly different between eras (19.5 vs. 25.5 vs. 33.1%, P=0.28). In contrast, the cumulative 5-year rectovaginal rate attenuated from 5.7% (the 1991–2005 era) to 1.7% (the 2006–2011 era), P=0.01. Conclusion Over the past two decades, the risk of developing a perianal fistula was stable, as well as its recurrence rate, underlining the lasting need for improving treatment strategies for this invalidating condition.OBJECTIVE Perianal disease is a debilitating condition that frequently occurs in Crohns disease (CD) patients. It is currently unknown whether its incidence has changed in the era of frequent immunomodulator use and biological availability. We studied the incidence and outcome of perianal and rectovaginal fistulas over the past two decades in our population-based Inflammatory Bowel Disease South-Limburg cohort. PATIENTS AND METHODS All 1162 CD patients registered in the Inflammatory Bowel Disease South-Limburg registry were included. The cumulative probabilities of developing a perianal and rectovaginal fistula were compared between three eras distinguished by the year of CD diagnosis: 1991-1998, 1999-2005 and 2006-2011. Second, clinical risk factors and the risk of fistula recurrence were determined. RESULTS The cumulative 5-year perianal fistula rate was 14.1% in the 1991-1998 era, 10.4% in the 1999-2005 era and 10.3% in the 2006-2011 era, P=0.70. Colonic disease was associated with an increased risk of developing perianal disease, whereas older age was associated with a decreased risk (both P<0.01). Over time, more patients were exposed to immunomodulators or biologicals before fistula diagnosis (18.5 vs. 32.1 vs. 52.1%, respectively, P=0.02) and started biological therapy thereafter (18.6 vs. 34.1 vs. 54.0%, respectively, P<0.01). The cumulative 5-year perianal fistula recurrence rate was not significantly different between eras (19.5 vs. 25.5 vs. 33.1%, P=0.28). In contrast, the cumulative 5-year rectovaginal rate attenuated from 5.7% (the 1991-2005 era) to 1.7% (the 2006-2011 era), P=0.01. CONCLUSION Over the past two decades, the risk of developing a perianal fistula was stable, as well as its recurrence rate, underlining the lasting need for improving treatment strategies for this invalidating condition.

A 20-Year Temporal Change Analysis in Incidence, Presenting Phenotype and Mortality, in the Dutch IBDSL Cohort—Can Diagnostic Factors Explain the Increase in IBD Incidence?

Background The aim was to study temporal changes in incidence, disease phenotype at diagnosis, and mortality of adult inflammatory bowel disease [IBD] patients in South Limburg, The Netherlands, diagnosed between 1991 and 2010. In addition, the 2010 IBD prevalence was estimated. Methods A multi-faceted approach including hospital administrations, the national pathology registry [PALGA], and general practitioners led to the identification of 1162 patients with Crohns disease [CD], 1663 with ulcerative colitis [UC], and 84 with unclassified IBD [IBD-U]. Temporal changes in incidence, disease phenotype, and mortality were studied using linear, multinomial regression analyses, and standardised mortality rates [SMR], respectively. Results The annual incidences increased from 17.90/100000 in 1991 to 40.36/100000 in 2010 for IBD, from 5.84/100000 to 17.49/100000 for CD, and from 11.67/100000 to 21.47/100000 for UC [p < 0.01 for all]. A shift towards milder disease at diagnosis was observed over time [eg decrease of complicated disease in CD, increase of proctitis in UC]. IBD mortality was similar to that in the general population (SMR 0.92; 95% confidence interval [CI] 0.81-1.05), and did not change over time. The estimated IBD prevalence was 830/100000. Conclusions The IBD incidence in South Limburg increased significantly between 1991 and 2010. The shift towards milder disease at diagnosis in parallel with the improved diagnostics and ability to detect low-grade inflammation was suggestive of an important role of diagnostic factors in this increase. Environmental factors probably played a role as well. The mortality was low and, together with the increasing incidence, led to the high prevalence of IBD in South Limburg.

Su1218 Concomitant Use of Immunomodulators Increases Long Term Response to Infliximab in Ulcerative Colitis - A Population-Based IBD-Sl Cohort Study

Background Infliximab is an effective treatment for steroid-refractory and steroid-dependent ulcerative colitis (UC). However, a subset of patients does not respond to infliximab and many patients lose response during follow up. Few studies have examined predictors of infliximab failure in UC and all available studies were performed in referral center cohorts. Since anti-TNF treatment is expensive and has potentially severe side effects, it is important to identify markers for response. This study evaluates the loss of response to infliximab in an unselected population-based cohort and aims to identify predictive markers for response. Methods Since 1991, all newly diagnosed IBD patients are included in our populationbased Inflammatory Bowel Disease South-Limburg (IBD-SL) cohort. Patients that ever used infliximab for UC were identified. Their medical records were reviewed with regard to indication for infliximab, reason of infliximab failure, age, gender, smoking status, comedication, disease location at diagnosis and at first infusion, and disease duration. A Cox proportional hazards regression model was used to assess markers for predicting primary and secondary response. Factors with a p-value ,0.1 were analyzed in a multivariate model. Associations were presented as Hazard ratios (HR). Results In our IBD-SL cohort, 109 out of 1464 UC patients (7.4%) were exposed to infliximab. Indication for starting infliximab was steroid-refractory disease (46.8%), steroid-dependent disease (41.3%) or intolerance to thiopurines (11.0%). Median follow-up was 45.5 (IQR 32.9-56.8) months after first infusion. Within 2 years 43.1% of the patients discontinued infliximab. Nine patients (8.3%) discontinued because of no response after 3 infusions. This number was too small to analyze predictive markers for non-response. In the group of primary responders (n=100), 38.0% discontinued infliximab within 2 years: 30 patients (79.0%) because of loss of response, 7 (18.4%) because of side effects and 1 patient (2.6%) for another reason. Predictive marker for loss of response was no concomitant use of immunosuppressive agents (HR 2.36; 95%CI 1.10-5.08). No predictive effects were found for disease duration (HR 0.99; 95%CI 0.99-1.00), and days of prednisone use before first infusion (HR 1.01; 95%CI 1.00-1.01), nor for any of the other factors in the univariate analyses. Conclusion In our population-based IBD cohort, almost half of the UC patients discontinue infliximab within 2 years after the first infusion. The majority of primary responders discontinue the drug because of loss of response. The UCSuccess trial showed higher short term response rates in patients on combination of infliximab and azathioprine. This study provides evidence that concomitant use of immunomodulators with infliximab also improves the long term response to infliximab in UC.

Corticosteroid Sparing in Inflammatory Bowel Disease is More Often Achieved in the Immunomodulator and Biological Era—Results from the Dutch Population-Based IBDSL Cohort

Objectives:Corticosteroid-free remission is an emerging treatment goal in the management of inflammatory bowel disease (IBD). In the population-based Inflammatory Bowel Disease South Limburg cohort, we studied temporal changes in corticosteroid use and assessed the corticosteroid-sparing effects of immunomodulators and biologicals in real life.Methods:In total, 2,823 newly diagnosed patients with Crohn’s disease (CD) or ulcerative colitis (UC) were included. Corticosteroid exposure and cumulative days of use were compared between patients diagnosed in 1991–1998 (CD: n=316, UC: n=539), 1999–2005 (CD: n=387, UC: n=527), and 2006–2011 (CD: n=459, UC: n=595). Second, the corticosteroid-sparing effects of immunomodulators and biologicals were assessed.Results:Over time, the corticosteroid exposure rate was stable (54.0% in CD and 31.4% in UC), even as the cumulative corticosteroid use in the first disease year (CD: 83 days (interquartile range (IQR) 35–189), UC: 62 days (IQR 0–137)). On the long-term, a gradual decrease in cumulative corticosteroid use was seen in CD (era ’91–’98: 366 days (IQR 107–841), era ’06–’11: 120 days (IQR 72–211), P<0.01), whereas in UC an initial decrease was observed (era ’91–’98: 184 days (IQR 86–443), era ’99–’05: 166 days (IQR 74–281), P=0.03), and stabilization thereafter. Immunomodulator and biological users had a lower risk of requiring corticosteroids than matched controls in CD only (33.6% vs. 49.9%, P<0.01, and 25.7% vs. 38.2%, P=0.04, respectively).Conclusions:In a real-world setting, more recently diagnosed IBD patients used lower amounts of corticosteroids as of the second year of disease. For CD, a significant association was found with the use of immunomodulators and biologicals. These conclusions support the increasing use of these treatment modalities.

80 Hospitalization and Surgery Risk in Crohn's Disease in the Biological Era - Results From the Dutch Population-Based IBD-SL Cohort

Background Many patients with Crohns disease (CD) require hospital admission or surgery for CD-related complications or refractory inflammatory disease. As from biological availability, more aggressive treatment strategies (e.g. early use of immunomodulators and biologicals, and combination therapy) are advised for patients with expected poor prognosis to improve long-term outcome. Its effect on disease outcome is yet unknown, mostly due to lack of data before and after biological availability from the same source population. Therefore, we aimed to compare disease outcome of CD patients diagnosed in the prebiological era to CD patients diagnosed in the biological era in a population-based cohort. MethodsSince 1991, incident IBD cases in the South-Limburg (SL) area are included in our populationbased IBD-SL cohort, with over 93% completeness. All CD patients were divided in two time cohorts. The pre-biological cohort comprised patients diagnosed between 1991 and 1998, followed until 1999 (registration of biological therapy for CD). The biological cohort comprised patients diagnosed between 1999 and July 2011, followed until 2014. Disease outcome, in terms of hospitalization and surgery risk, was analyzed with a Kaplan-Meier survival curve, and hazard ratios (HR) were calculated using a Cox regression model. Results In total, 342 patients in the pre-biological and 820 patients in the biological era were included, with a mean follow-up of 4.0 (SD 2.5) and 6.4 (SD 3.6) years, respectively. At diagnosis, patients in the biological era were less often hospitalized (21.1% vs. 36.8%, HR

DOP019 The relevance of population based IBD biobanks: a meta-analysis and introduction of the IBD-SL biobank cohort

DOP018 Low microbial diversity in Crohn’s disease is due to striking depletion of unknown species N. Borruel1 *, H.B. Nielsen2, F. Casellas1, C. Manichanh3, E. Varela3, M. Antolin3, A. Torrejon1, V. Robles1, P. Nos4, X. Calvet5, F. Guarner1, MetaHIT Consortium6. 1Hospital Vall d’Hebron, Crohn’s and Colitis Attention Unit. Digestive System Research Unit, Barcelona, Spain, 2Technical University of Denmark, Center for Biological Sequence Analysis, Department of Systems Biology, Lyngby, Denmark, 3Hospital Vall d’Hebron, Digestive System Research Unit, Barcelona, Spain, 4Hospital Universitari i Politecnic La Fe, Servicio de Medicina Digestiva, Valencia, Spain, 5Corporacio Sanitaria Parc Tauli, Servei d’Aparell Digestiu, Sabadell, Spain, 6MetaHIT Network, Partner Centres, Paris, France