Charles W. Lees

Genome-wide association study of ulcerative colitis identifies three new susceptibility loci, including the HNF4A region

Ulcerative colitis is a common form of inflammatory bowel disease with a complex etiology. As part of the Wellcome Trust Case Control Consortium 2, we performed a genome-wide association scan for ulcerative colitis in 2,361 cases and 5,417 controls. Loci showing evidence of association at P < 1 × 10−5 were followed up by genotyping in an independent set of 2,321 cases and 4,818 controls. We find genome-wide significant evidence of association at three new loci, each containing at least one biologically relevant candidate gene, on chromosomes 20q13 (HNF4A; P = 3.2 × 10−17), 16q22 (CDH1 and CDH3; P = 2.8 × 10−8) and 7q31 (LAMB1; P = 3.0 × 10−8). Of note, CDH1 has recently been associated with susceptibility to colorectal cancer, an established complication of longstanding ulcerative colitis. The new associations suggest that changes in the integrity of the intestinal epithelial barrier may contribute to the pathogenesis of ulcerative colitis.

Resolution of Non–Small-Cell Lung Cancer after Withdrawal of Anti-TNF Therapy

To the Editor: We report the development of locally advanced non–small-cell lung cancer (tumor–node–metastasis [TNM] stage T4N2M0) in a 69-year-old woman with Crohns colitis. Her condition had been diagnosed in 2000 and had required treatment with methotrexate (from 2003 to the present), together with anti–tumor necrosis factor (TNF) therapies — initially, episodic infliximab (from 2000 to 2004) and thereafter, maintenance adalimumab, until the tumor was diagnosed in June 2006. Remarkably, the tumor expressed TNF receptors and underwent regression and sustained remission on withdrawal of anti-TNF therapy, with no other therapeutic intervention. In 1999, the patient, a former smoker (35 pack-years), .xa0.xa0.

PTU-123 Acute severe ulcerative colitis: the last 12 years in Edinburgh

Introduction Acute severe ulcerative colitis (ASUC) remains a common reason for urgent colectomy, yet there are relatively few large cohort studies exploring prognosis and outcome.1–3 This study aims to examine presentation and management of ASUC in the Western General Hospital, a tertiary referral centre in Edinburgh and to identify prognostic factors. Methods Patients were identified from a large database of participants in genetic studies in Edinburgh, as well as from two previous small cohorts of ASUC studied in Edinburgh. More recent cases were found from minutes of a weekly IBD meeting. Cases were included if they met the standard clinical, radiological and pathological criteria for ulcerative colitis and required an admission for 3u2005days or more requiring intravenous corticosteroids and/or colectomy. Two cohorts were analysed, one with full clinical detail (97 admissions in 86 patients) and one with more basic detail (444 admissions in 323 patients). Results Overall colectomy rate was 31.8%. Haemoglobin, C reactive protein and albumin at days 0 and 3 were significant predictors of colectomy in both cohorts (p<0.05 in each case), while in the detailed cohort day 3 but not day 0 stool frequency was predictive (p<0.001 and 0.81 respectively). A simple score was derived to predict colectomy at admission based on disease extent, albumin and CRP. Scores of 0, 1, 2 and 3 corresponded to risks for colectomy of 10%, 31%, 61% and 75%. For day 3 parameters, both the Edinburgh acute colitis (Ho) score (Abstract PTU-123 figure 1) and Travis criteria performed well.Abstract PTU-123 Figure 1 Conclusion ASUC remains an important cause of colectomy. This study confirms the prognostic value of the Ho score and Travis criteria at day 3, but also indicates that day 0 CRP and albumin are strong predictors of outcome. Competing interests None declared. References 1. Dinesen LC, et al. J Crohns Colitis 2010;4:431–7. 2. Ho GT, et al. Aliment Pharmacol Ther 2004;19:1079–87. 3. Turner D, et al. Clin Gastroenterol Hepatol 2007;5:103–10.

W09-03 - Associations of affective disorder in patients with inflammatory bowel disease

Introduction Inflammatory Bowel Disease (IBD) is associated with co morbid depression and anxiety of up to 42%. Corticosteroids, used commonly in IBD, are known to cause psychiatric side-effects and could be an independent risk factor for affective illness. Recent studies show that depression is also associated with raised CRP and IL6. Aims This study aims to show which demographic, clinical, medication and immunological factors are predictors of anxiety and depression in IBD. Methods The IBD, Steroids and Affective Disorder (ISA) study is a cross-sectional study of IBD patients in Edinburgh, UK. Out patients underwent assessment including Hospital Anxiety Depression Scale (HADS) past psychiatric history, steroid medication history, inflammatory markers, the Medication Adherence Rating Scale (MARS) and Altman Self Rated Mania Scale (ARSM). Results 326 patients with Crohns and 256 with Colitis (72% of clinic attendees) were recruited. 251 (43%) patients scored 12 or above on the HADS questionnaire. 45% of patients had previously suffered from affective illness. Patients on Prednisolone and Budesonide scored significantly higher on HADS Depression (pxa0=xa00.03 and pxa0=xa00.002) as did those who had been on Prednisolone for more than 8 weeks (pxa0=xa00.041). Being on prednisolone was not associated with increased Colitis and Crohns activity indices (pxa0=xa00.2). HADS scores were measured against other disease and demographic variables. Conclusion Affective illness is common in the IBD population and the prescription and duration of systemic corticosteroids are associated with depression. Biological and disease variables may play an important role in co morbid affective illness in IBD

OC-013 Clinical utility of faecal calprotectin in diagnosing inflammatory bowel disease during first presentation to the gastroenterology clinic: a novel investigative algorithm: Abstract OC-013

Introduction Faecal calprotectin (FC) has been shown to distinguish reliably between functional and inflammatory bowel diseases (IBD) in a number of studies of patients with established gastrointestinal disease. However, it is a poor screening test for colorectal cancer. Presently, diagnostic algorithms incorporating FC at first presentation to the gastroenterology clinic are lacking. Methods This study aimed to determine the optimal use of FC, in conjunction with serum markers, in the initial diagnostic workup of patients aged 16–50u2005years. Patients >50u2005years were excluded, as they require colonoscopy to exclude malignancy. Detailed clinical, laboratory, endoscopic and radiological parameters were collected. Patients with a prior GI diagnosis were excluded. All stool samples were analysed in the same laboratory using the same FC assay (Calpro ELISA). ROME III and Lennard-Jones criteria were used. Patients were followed up for a minimum of 12u2005months. Results The Edinburgh FC Register comprises data on 7512 patients (2005–2008) from across Lothian. 1838/7512 were managed at 2 Edinburgh teaching hospitals, and aged 16–50u2005years at time of FC; 851 were excluded, mostly due to pre-established diagnosis of IBD. 987 (mean±SD age 33.5±9.0u2005years; 64.8% female) met the strict inclusion criteria. In patients presenting primarily with bloody diarrhoea (n=68), median FC was 690u2005μg/g (IQR: 72–2323), in contrast to watery diarrhoea (n=325, FC 30u2005μg/g (20–90), p<0.0001), or abdominal pain (n=271; FC 25u2005μg/g (20–85), p<0.0001). 193/987 (19.6%) were ultimately diagnosed with organic disease with median FC of 180u2005μg/g (37–1292), 113/987 (11.4%) with IBD (FC 990.0 (225–2190)) and 605/987 (61.3%) with a functional disorder (FC 20 (20–50)). FC was influenced by NSAID use, but not by age, sex or smoking. ROC analysis of FC in discriminating organic disease and IBD from functional disorders gave an area under the curve of 0.80 (95% CI 0.76 to 0.84) and 0.93 (95% CI 0.90 to 0.96), respectively. Further detailed analysis suggested the following model: FC<70u2005μg/g with normal blood parameters and no alarm symptoms/signs – no further investigation; FC>50u2005μg/g with bloody diarrhoea OR raised CRP/low albumin – combined radiological and endoscopic evaluation. Conclusion In this study, the largest on the clinical utility of FC, we were able to develop clear investigative strategies in adults <50u2005years, dependent on FC and other clinical/laboratory parameters. This simple algorithm could be rolled out into primary care facilitating timely and appropriate triage of new patients and minimising diagnostic delay in IBD. Abstract OC-013 IBD vs functional disease Calprotectin Sensitivity Specificity PPV NPV <20u2005μg/g 0.97 0.49 28.0 99.0 <50u2005μg/g 0.93 0.70 43.3 98.1 <70u2005μg/g 0.92 0.77 50.8 98.1 <100u2005μg/g 0.87 0.84 58.4 97.2

Sa1937 Predictive Factors of Disease Relapse Following Thiopurine Withdrawal for Sustained Clinical Remission of IBD

Introduction Thiopurine therapy is effective in maintaining clinical remission in IBD. However, long-term therapy is associated with an increased risk of lymphoma; therefore in clinical practice we aim to withdraw therapy after 4–5u2005years. Nevertheless, many patients will experience disease relapse within 12u2005months of drug withdrawal. 1 Methods The aim of the present study was to retrospectively determine the relapse rate in ulcerative colitis (UC) and Crohn9s disease (CD) following azathioprine (AZA) or mercaptopurine (MP) withdrawal and to determine factors predictive of relapse. Patients were identified by electronic case note review of an IBD research database in Edinburgh. Major inclusion criteria were AZA and/or MP therapy for a minimum of 3u2005years, AZA/MP withdrawn due to sustained clinical remission no steroid therapy for 6u2005months prior to drug withdrawal, and minimum 12u2005months follow-up. The primary outcome was disease relapse requiring AZA reinitiation, steroids or colectomy within 12u2005months of AZA/MP withdrawal, with secondary outcome assessed at 24u2005months. Clinical/laboratory predictors of relapse were sought. 1826 electronic records were reviewed (865 CD and 961 UC). 634 were treated with a thiopurine (348 CD and 286 UC). 74 met the strict study inclusion criteria (45 CD and 29 UC). Results CD was associated with a significantly higher risk of relapse than UC on Kaplan–Meier analysis (Abstract OC-166 figure 1, p=0.026). The moderate-severe relapse rate for 12u2005months was 44% for CD and 14% for UC. For 24u2005months, relapse rates were 60% for CD and 48% for UC. Elevated platelet count (p=0.03) and elevated white cell count (p=0.03) were predictive of relapse for UC, while no predictive factors were identified for CD. Median (range) duration of thiopurine use was 6.2 (3.4–18.7)u2005years for CD and 6.0 (3.1–18.0)u2005years for UC. Median duration of follow-up was 32u2005months for CD and 45u2005months for UC. Retreatment with a thiopurine after relapse was successful in 7/7 cases for UC and 18/24 for CD. Conclusion Relapse rates after withdrawal of a thiopurine are high, particularly for CD, and predicting this remains difficult. To help increase the power of this study, we are now expanding it across the UK. Competing interests None declared. Reference 1. Treton . Clin Gastroenterol Hepatol 2009; 7 :80.

1128 A Trial With Mercaptopurine Following Azathioprine Intolerance is a Safe Treatment Strategy for the Majority of Patients With IBD

Introduction Azathioprine intolerance (AZA-I) leads to withdrawal of therapy in up to 30% of patients with IBD. Smaller case series demonstrated that mercaptopurine (MP) could be a well tolerated alternative in selected patients. 1 This study aims to further assess its tolerability, in a larger cohort of AZA-I patients, over a longer period of time, and to re-evaluate potential factors predictive of tolerance. Methods A retrospective audit was made of 137 patients with IBD (78 women, median age at diagnosis 32u2005years, 78 with CD, 59 with UC) who had been intolerant of AZA and then subsequently treated with MP. Results MP was tolerated by 58% of AZA-I patients (median follow-up 937u2005days, median dose 0.91u2005mg/kg). Tolerance was highest in patients with AZA related gastrointestinal intolerance (66%) and hepatotoxicity (61%), and lowest in patients with AZA related flu-like illness (40%). The number of patients with AZA induced neutropenia and pancreatitis were too small to draw firm conclusions (see Abstract PTU-122 table 1). Age at diagnosis was significantly associated with tolerability. Patients intolerant of MP were younger (28 vs 33 yro; p=0.024) and of those under the age of 40 only 55% tolerated MP compared with 69% of those aged 40u2005years or over (p=Azathioprine-intolerant patients, subsequently treated with mercaptopurine (AZA-I/MP) grouped by azathioprine intolerance Conclusion Consistent with previous data, this, the largest series to date, with substantial follow- up, has shown that MP is a safe alternative for up to 60% of AZA-I patients, including some with a previous major intolerance. Patients with previous gastrointestinal intolerance or hepatotoxicity may be more likely to tolerate a trial of MP. Competing interests None declared. Reference 1. Lees , et al. Aliment Pharmacol & Ther 2007; 27 :220–7.

Sa1241 Acute Severe Ulcerative Colitis: the Last Twelve Years in Edinburgh

Background Acute severe ulcerative colitis (ASUC) remains a common reason for urgent colectomy, yet there are relatively few large cohort studies exploring prognosis and outcome. This study aims to examine presentation and management of ASUC in the Western General Hospital, a tertiary referral centre in Edinburgh, UK and to identify prognostic factors. Methods Patients were identified from a large database of participants in genetic studies in Edinburgh, as well as from two previous small cohorts of ASUC studied in Edinburgh. More recent cases were found from minutes of a weekly IBD meeting. Cases were included if they met the standard clinical, radiological and pathological criteria for ulcerative colitis and required an admission for 3 days or more requiring intravenous corticosteroids and/or colectomy. Two cohorts were analysed, one with full clinical detail (97 admissions in 86 patients) and one with more basic detail (444 admissions in 323 patients). Results Overall colectomy rate was 31.8%. Haemoglobin, C-reactive protein and albumin at days 0 and 3 were significant predictors of colectomy in both cohorts (p<0.05 in each case), while in the detailed cohort day 3 but not day 0 stool frequency was predictive (p<0.001 and 0.81 respectively). A simple score was derived to predict colectomy at admission based on disease extent, albumin and CRP. Scores of 0, 1, 2 and 3 corresponded to risks for colectomy of 10%, 31%, 61% and 75%. For day 3 parameters, both the Edinburgh acute colitis (Ho) score (figure 1) and Travis criteria performed well. Conclusions ASUC remains an important cause of colectomy. This study confirms the prognostic value of the Ho score and Travis criteria at day 3, but also indicates that day 0 CRP and albumin are strong predictors of outcome.

Depression in IBD: Preliminary Results of the ISA (IBD, Steroids and Affective Disorder) Study

G A A b st ra ct s hospitalized for severe infection compared to IBD patients admitted for other causes (11.8 vs. 7.2 days, p <0.0001) with 12.5% of patients requiring ICU admission, 30-day readmission rate of 12.5% and 4.6% in-hospital mortality rate. The majority (56.6%) of inpatients were on at least 1 form of immunosuppressive therapy, with 13.2% on combination immunosuppression. More patients hospitalized from 2006-2008 were on immunosuppressive therapy compared to 2003-2005 (57.9% vs. 27.3%, p=0.007). Conclusion: While the number of IBD hospitalizations per year has not changed over the past 6 years, the annual incidence rate for severe infections among IBD patients has been increasing, particularly in the older patient, and associated with longer length of hospital stay. The contribution of immunosupppressive therapies towards these observed increased infection rates warrants further investigation. Table 1. Organisms isolated from IBD patients admitted with severe infection

OC-047 Colonic expression profiles of genes identified by genome wide association scan in ulcerative colitis

Introduction Genome wide expression analysis using microarray allows a comprehensive picture of gene expression at the tissue and cellular level. The aim of this study was to investigate the differential expression of 26 genes recently identified by Barrett et al as being associated with ulcerative colitis (UC).1 Methods 67 Ulcerative colitis patients (UC) and 31 controls (HC) were studied. 129 UC and 73 HC paired endoscopic biopsies were taken from specific anatomical locations for RNA extraction and histology assessment. Expression signatures were investigated using 39 probes representing the 26 genes using the Agilent microarray platform. Results When all the UC biopsies were compared to controls MST1 was the most dysregulated gene- fold change (FC) −1.64, p <0.00001 in the UC biopsies. HERC2 was also downregulated in the UC biopsies (FC −1.2, p <0.00001) as was CCNY (FC −1.2, p=0.0001). TNFRSF6B was significantly upregulated (FC +1.2, p=0.005). When inflamed and non-inflamed descending and sigmoid colon UC biopsies were compared TNFRSF6B was upregulated (FC +2.98, p <0.00001) as was MHC (FC +1.88, p=0.027), the novel susceptibility genes LAMB1 (FC +1.23, p=0.0058), HNF4A (FC +1.23, p=0.024) and ARPC2 (FC +1.20, p=0.01). MST1 (FC −1.76, p=0.015) and CCNY (FC −1.42, p=0.001) were downregulated in the inflamed UC biopsies. HERC2 was the only significantly dysregulated gene when non-inflamed UC and control biopsies were compared (FC −1.20, p=0.0028) and was downregulated in UC. Conclusion Disease and inflammation dependent expression changes were observed in a number of the UC genome wide expression scan susceptibility genes further implicating them in the pathogenesis of UC.