- Email: [email protected]
Su1188 Transabdominal Ultrasound Has High Negative Predictive Value for Short-Term Complications of Inflammatory Bowel Disease
Su1188
June 2011 to October 2014. On the basis of discriminant analysis using colonoscopy, laboratory and radiologic parameters, a scoring system for differentiating between two diseases was developed and was validated on additional 40 patients (CD 20, ITB 20) Results: The colonoscopic evaluation was performed by blinded two experienced endoscopists. The accuracy of colonoscopic diagnosis was 81.2% and the positive predictive values for CD and ITB were 100% and 84.7%, respectively. On univariate analysis for laboratory and radiologic parameters, positive IgA or IgG ASCA (anti-saccharomyces cerevisiae antibody), anemia, hypoalbuminemia, elevation of ESR and abnormal lesions on small-bowel followthrough (SBFT) were significantly more common in CD, whereas QuantiFERON®-TB Gold In-Tube Test (QFT-G) and typical findings for pulmonary TB on chest X-ray (CXR) were more common in ITB. On multivariate analysis, positive IgA or IgG ASCA, positive QFTG, abnormal SBFT and abnormal CXR were independent parameters differentiating two diseases. On the basis of multivariate discriminant analysis, a score for CD was 1.784 and for ITB was -1.784 and the cutoff value was zero. The accuracy of discriminant function was 92.5% and the area under the curve for receiver-operating characteristic (AUROC) to assess the ability of these features to discriminate between two diseases was 0.991 (95% CI 0.978-1.000). In a validation model, the accuracy of discriminant function was 95.0% and AUROC was 0.981 (95% CI 0.950-1.000). Conclusion: Simple laboratory and radiologic parameters including ASCA, QFT-G, SBFT and CXR are useful diagnostic aid in combination with colonoscopic evaluation for the differentiation between CD and ITB.
AGA Abstracts
Transabdominal Ultrasound Has High Negative Predictive Value for ShortTerm Complications of Inflammatory Bowel Disease Zuzana Zelinkova, Barbora Kadleckova, Beata Repakova Background&Aims: Transabdominal ultrasound (US) is a non-invasive, easily accessible cross-sectional imaging technique for bowel examination. It has been shown to have high specificity but traditionally lower sensitivity compared with other cross-sectional techniques and/or endoscopy in depiction of inflammatory bowel disease (IBD) activity in patients with established IBD. The data available provide comparison of transabdominal US with endoscopy and other cross-sectional imaging but the predictive value of IBD flare is unknown for US. Therefore, the aim of this study was first, to determine the negative predictive value of US in a setting of IBD patients with clinical suspicion of flare and second, to establish the rate of agreement between US and endoscopic and magnetic resonance enterography (MRE) findings. Patients&Methods: Between July and October 2014, all consecutive patients with clinical suspicion of IBD flare were referred for transabdominal US at one referral IBD centre. The activity of IBD on US was determined according to standard criteria for each small and large bowel segment. Negative predictive value for short-term complications of IBD, i.e. flare or surgery occurring within 1 to 4 months of US, was calculated. Endoscopic and MRE findings were categorized for each bowel segment and the agreement between US and other techniques was considered only if the agreement on presence of disease activity was achieved for all segments. Results: In total, 80 US were performed in 79 patients; 62 (78%) with Crohn`s disease, 16 (20%) with ulcerative colitis, 1 pts with IBD unclassified; 52 (65%) women; mean age 35 years, range 19-79. Overall, there were 44 (55%) cases of US findings of active disease, in 6 patients fistula or abscess were found on US. From 36 cases with negative US findings, 35 were in remission without any changes in therapy at the median follow-up of 3 months (range 1 to 5 month); resulting in negative predictive value of 97% (95%CI 85.42 % to 99.54 %). During the follow-up period, 32 (40%) patients underwent MRE and 21 (26%) ileocolonoscopy, with agreement of US findings in 84% and 73% of cases for MRE and ileocolonoscopy, respectively. Conclusion: Transabdominal ultrasound has high negative predictive value for short-term complications of inflammatory bowel disease and seems to be a promising tool for rapid and non-invasive triage of IBD patients who do not necessitate further diagnostic work-up. Su1189 Genome-Wide Discovery of Inflammatory Bowel Disease-Specific Transcriptional Enhancers and Promoters Identifies Powerful Diagnostic Biomarkers Morana Vitezic, Mette Boyd, Jette Bornholdt, Mehmet Coskun, Malte Thodberg, Robin Andersson, Jesper Troelsen, Jacob T. Bjerrum, Ole H. Nielsen, Albin Sandelin
Su1191 Serum Hepcidin Levels Predict Intestinal Iron Absorption in IBD Patients Miriam Wiesenthal, Franz Hartmann, Tariq Iqbal, Axel U. Dignass, Jürgen Stein Background: Circulating hepcidin is proposed to regulate iron absorption by modulating iron export by ferroportin at the basolateral membrane of the duodenal mucosal cells and/ or uptake into the cells at the apical membrane by DMT1. To date, no data have shown a relationship between plasma hepcidin concentrations and iron absorption in IBD patients. In the present study, we used stored samples from a human iron absorption study to further test the hypothesis that plasma hepcidin may explain interindividual variation in iron absorption in IBD patients. Method: Serum ferritin (SF) and serum markers of inflammation [high-sensitivity C-reactive protein (hsCRP) and IL-6] were measured in stored samples from a human iron absorption study using commercially available immune-assays. Hepcidin25 concentrations were determined in fasting samples from 71 adult subjects with IBD (31 UC, 40 CD) and 26 healthy controls. Hepcidin was measured by LC-MS. Results: There was a positive correlation between hepcidin (mean: 2.3; range: 0.1-7.8nmol/L) and hsCRP (p<0.005), but not between hepcidin and serum ferritin (p>0.05). Whereas iron absorption was negatively correlated with serum ferritin only in patients with inactive disease (hsCRP<5md/dl; p< 0.001), a negative correlation was observed with serum hepcidin in both active and inactive disease (p= 0.006), independent of IBD phenotype. Multiple linear regression models showed that serum hepcidin in isolation significantly predicted the interindividual variation in iron absorption. Conclusions: Concentration of serum hepcidin, but not serum ferritin, was highly correlated with intestinal iron absorption in IBD patients.
Background and aims Inflammatory bowel disease (IBD) is an umbrella term of various chronic disorders of which Crohn's disease (CD) and ulcerative colitis (UC) are the two prevailing subtypes Appropriate management is highly dependent on early and accurate diagnosis, which is challenging. Therefore, new biomarkers for stratifying patients and to improve diagnostics are highly needed - a first step towards personalized medicine in IBD. The search for genetic determinants of IBD has until now identified more than 160 loci containing susceptibility genes. SNPs, however, only explain a minor part of disease phenotypes, and many IBD-associated SNPs fall into non-protein-coding regions. We hypothesized that regulatory regions and novel RNAs may provide better diagnostic tools. Therefore, we aimed to identify enhancers and novel transcription start sites specific for subtypes of IBD. Methods We used a novel RNA sequencing technique, Cap Analysis of Gene Expression, to profile descending colon biopsies from >100 patients (split between CD and UC with quiescent or active disease, and controls). This enabled the identification of transcription start sites of known and novel genes, as well as transcribed enhancer regions. Results With this material, we could for the first time pinpoint IBD-specific regulatory events, including ~50.000 promoter regions and ~40.000 enhancers. This is an important resource for IBD pathology and diagnosis. We revealed that different classes of transcriptional events are indicative of different aspects of IBD. In particular, lowly expressed transcripts, including previously unidentified long non-coding RNAs, alternative transcription start sites of known genes and novel enhancer regions, were far more powerful predictors of disease states (e.g., UC vs. CD) than highly expressed genes. Importantly, these targets cannot be identified using microarray approaches, and may explain the limited classification power of these approaches. Conversely, highly expressed genes were powerful predictors of the degree of inflammation, but not disease subtype. Overall, we could with high accuracy predict the subtype of disease as well as inflammation states using these data. Because RNA levels were measured, the most potent classifier RNAs can be determined with simple and low-cost PCR analyses and can thus become an invaluable tool for establishing a correct and early diagnosis in clinical settings. SNPs associated to UC and CD were highly over-represented within the enhancers identified (odds ratio=41.6 for UC and 33.5 for CD, P < 2.2e-16). Thus, we could identify SNPs with potential regulatory effects in IBD. Conclusions We have produced the first dataset identifying IBD-specific enhancers and promoters. This will be highly valuable for targeted molecular diagnostics, IBD disease biology, and genetics. Su1190 Development of a Score for Differential Diagnosis Between Intestinal Tuberculosis and Crohn's Disease: A Prospective Study Jung Ho Bae, Sang Hyoung Park, Ho-su Lee, Hyo Jeong Lee, Jae Seung Soh, Seohyun Lee, Min Seob Kwak, Dong-Hoon Yang, Kyung-Jo Kim, Jeong-Sik Byeon, Seung-Jae Myung, Suk-Kyun Yang, Jin-Ho Kim, Byong Duk Ye
Fig. 1. Correlation between serum hepcidin and intestinal iron absorption (n=97; p=0.006)
Backgrounds: Although colonoscopy is useful for differentiating between Crohn's disease (CD) and intestinal tuberculosis (ITB), there have still been some limitations. Therefore, we tried to develop clinically useful predicting model in differentiating between CD and ITB using laboratory and radiologic features in addition to colonoscopic diagnosis. Methods: We prospectively enrolled newly diagnosed CD (n = 40) and ITB (n = 40) patients from
AGA Abstracts
S-432
June 2011 to October 2014. On the basis of discriminant analysis using colonoscopy, laboratory and radiologic parameters, a scoring system for differentiating between two diseases was developed and was validated on additional 40 patients (CD 20, ITB 20) Results: The colonoscopic evaluation was performed by blinded two experienced endoscopists. The accuracy of colonoscopic diagnosis was 81.2% and the positive predictive values for CD and ITB were 100% and 84.7%, respectively. On univariate analysis for laboratory and radiologic parameters, positive IgA or IgG ASCA (anti-saccharomyces cerevisiae antibody), anemia, hypoalbuminemia, elevation of ESR and abnormal lesions on small-bowel followthrough (SBFT) were significantly more common in CD, whereas QuantiFERON®-TB Gold In-Tube Test (QFT-G) and typical findings for pulmonary TB on chest X-ray (CXR) were more common in ITB. On multivariate analysis, positive IgA or IgG ASCA, positive QFTG, abnormal SBFT and abnormal CXR were independent parameters differentiating two diseases. On the basis of multivariate discriminant analysis, a score for CD was 1.784 and for ITB was -1.784 and the cutoff value was zero. The accuracy of discriminant function was 92.5% and the area under the curve for receiver-operating characteristic (AUROC) to assess the ability of these features to discriminate between two diseases was 0.991 (95% CI 0.978-1.000). In a validation model, the accuracy of discriminant function was 95.0% and AUROC was 0.981 (95% CI 0.950-1.000). Conclusion: Simple laboratory and radiologic parameters including ASCA, QFT-G, SBFT and CXR are useful diagnostic aid in combination with colonoscopic evaluation for the differentiation between CD and ITB.
AGA Abstracts
Transabdominal Ultrasound Has High Negative Predictive Value for ShortTerm Complications of Inflammatory Bowel Disease Zuzana Zelinkova, Barbora Kadleckova, Beata Repakova Background&Aims: Transabdominal ultrasound (US) is a non-invasive, easily accessible cross-sectional imaging technique for bowel examination. It has been shown to have high specificity but traditionally lower sensitivity compared with other cross-sectional techniques and/or endoscopy in depiction of inflammatory bowel disease (IBD) activity in patients with established IBD. The data available provide comparison of transabdominal US with endoscopy and other cross-sectional imaging but the predictive value of IBD flare is unknown for US. Therefore, the aim of this study was first, to determine the negative predictive value of US in a setting of IBD patients with clinical suspicion of flare and second, to establish the rate of agreement between US and endoscopic and magnetic resonance enterography (MRE) findings. Patients&Methods: Between July and October 2014, all consecutive patients with clinical suspicion of IBD flare were referred for transabdominal US at one referral IBD centre. The activity of IBD on US was determined according to standard criteria for each small and large bowel segment. Negative predictive value for short-term complications of IBD, i.e. flare or surgery occurring within 1 to 4 months of US, was calculated. Endoscopic and MRE findings were categorized for each bowel segment and the agreement between US and other techniques was considered only if the agreement on presence of disease activity was achieved for all segments. Results: In total, 80 US were performed in 79 patients; 62 (78%) with Crohn`s disease, 16 (20%) with ulcerative colitis, 1 pts with IBD unclassified; 52 (65%) women; mean age 35 years, range 19-79. Overall, there were 44 (55%) cases of US findings of active disease, in 6 patients fistula or abscess were found on US. From 36 cases with negative US findings, 35 were in remission without any changes in therapy at the median follow-up of 3 months (range 1 to 5 month); resulting in negative predictive value of 97% (95%CI 85.42 % to 99.54 %). During the follow-up period, 32 (40%) patients underwent MRE and 21 (26%) ileocolonoscopy, with agreement of US findings in 84% and 73% of cases for MRE and ileocolonoscopy, respectively. Conclusion: Transabdominal ultrasound has high negative predictive value for short-term complications of inflammatory bowel disease and seems to be a promising tool for rapid and non-invasive triage of IBD patients who do not necessitate further diagnostic work-up. Su1189 Genome-Wide Discovery of Inflammatory Bowel Disease-Specific Transcriptional Enhancers and Promoters Identifies Powerful Diagnostic Biomarkers Morana Vitezic, Mette Boyd, Jette Bornholdt, Mehmet Coskun, Malte Thodberg, Robin Andersson, Jesper Troelsen, Jacob T. Bjerrum, Ole H. Nielsen, Albin Sandelin
Su1191 Serum Hepcidin Levels Predict Intestinal Iron Absorption in IBD Patients Miriam Wiesenthal, Franz Hartmann, Tariq Iqbal, Axel U. Dignass, Jürgen Stein Background: Circulating hepcidin is proposed to regulate iron absorption by modulating iron export by ferroportin at the basolateral membrane of the duodenal mucosal cells and/ or uptake into the cells at the apical membrane by DMT1. To date, no data have shown a relationship between plasma hepcidin concentrations and iron absorption in IBD patients. In the present study, we used stored samples from a human iron absorption study to further test the hypothesis that plasma hepcidin may explain interindividual variation in iron absorption in IBD patients. Method: Serum ferritin (SF) and serum markers of inflammation [high-sensitivity C-reactive protein (hsCRP) and IL-6] were measured in stored samples from a human iron absorption study using commercially available immune-assays. Hepcidin25 concentrations were determined in fasting samples from 71 adult subjects with IBD (31 UC, 40 CD) and 26 healthy controls. Hepcidin was measured by LC-MS. Results: There was a positive correlation between hepcidin (mean: 2.3; range: 0.1-7.8nmol/L) and hsCRP (p<0.005), but not between hepcidin and serum ferritin (p>0.05). Whereas iron absorption was negatively correlated with serum ferritin only in patients with inactive disease (hsCRP<5md/dl; p< 0.001), a negative correlation was observed with serum hepcidin in both active and inactive disease (p= 0.006), independent of IBD phenotype. Multiple linear regression models showed that serum hepcidin in isolation significantly predicted the interindividual variation in iron absorption. Conclusions: Concentration of serum hepcidin, but not serum ferritin, was highly correlated with intestinal iron absorption in IBD patients.
Background and aims Inflammatory bowel disease (IBD) is an umbrella term of various chronic disorders of which Crohn's disease (CD) and ulcerative colitis (UC) are the two prevailing subtypes Appropriate management is highly dependent on early and accurate diagnosis, which is challenging. Therefore, new biomarkers for stratifying patients and to improve diagnostics are highly needed - a first step towards personalized medicine in IBD. The search for genetic determinants of IBD has until now identified more than 160 loci containing susceptibility genes. SNPs, however, only explain a minor part of disease phenotypes, and many IBD-associated SNPs fall into non-protein-coding regions. We hypothesized that regulatory regions and novel RNAs may provide better diagnostic tools. Therefore, we aimed to identify enhancers and novel transcription start sites specific for subtypes of IBD. Methods We used a novel RNA sequencing technique, Cap Analysis of Gene Expression, to profile descending colon biopsies from >100 patients (split between CD and UC with quiescent or active disease, and controls). This enabled the identification of transcription start sites of known and novel genes, as well as transcribed enhancer regions. Results With this material, we could for the first time pinpoint IBD-specific regulatory events, including ~50.000 promoter regions and ~40.000 enhancers. This is an important resource for IBD pathology and diagnosis. We revealed that different classes of transcriptional events are indicative of different aspects of IBD. In particular, lowly expressed transcripts, including previously unidentified long non-coding RNAs, alternative transcription start sites of known genes and novel enhancer regions, were far more powerful predictors of disease states (e.g., UC vs. CD) than highly expressed genes. Importantly, these targets cannot be identified using microarray approaches, and may explain the limited classification power of these approaches. Conversely, highly expressed genes were powerful predictors of the degree of inflammation, but not disease subtype. Overall, we could with high accuracy predict the subtype of disease as well as inflammation states using these data. Because RNA levels were measured, the most potent classifier RNAs can be determined with simple and low-cost PCR analyses and can thus become an invaluable tool for establishing a correct and early diagnosis in clinical settings. SNPs associated to UC and CD were highly over-represented within the enhancers identified (odds ratio=41.6 for UC and 33.5 for CD, P < 2.2e-16). Thus, we could identify SNPs with potential regulatory effects in IBD. Conclusions We have produced the first dataset identifying IBD-specific enhancers and promoters. This will be highly valuable for targeted molecular diagnostics, IBD disease biology, and genetics. Su1190 Development of a Score for Differential Diagnosis Between Intestinal Tuberculosis and Crohn's Disease: A Prospective Study Jung Ho Bae, Sang Hyoung Park, Ho-su Lee, Hyo Jeong Lee, Jae Seung Soh, Seohyun Lee, Min Seob Kwak, Dong-Hoon Yang, Kyung-Jo Kim, Jeong-Sik Byeon, Seung-Jae Myung, Suk-Kyun Yang, Jin-Ho Kim, Byong Duk Ye
Fig. 1. Correlation between serum hepcidin and intestinal iron absorption (n=97; p=0.006)
Backgrounds: Although colonoscopy is useful for differentiating between Crohn's disease (CD) and intestinal tuberculosis (ITB), there have still been some limitations. Therefore, we tried to develop clinically useful predicting model in differentiating between CD and ITB using laboratory and radiologic features in addition to colonoscopic diagnosis. Methods: We prospectively enrolled newly diagnosed CD (n = 40) and ITB (n = 40) patients from
AGA Abstracts
S-432