S8
Abstracts / Digestive and Liver Disease 41S (2009), S1–S167
in a number of chronic inflammatory disorders. However, the mechanisms regulating its production in inflammatory bowel disease (IBD) are still unclear. Material and methods: Endoscopic biopsies or surgical specimens were taken from inflamed colonic mucosa of 72 IBD patients (38 affected by Crohn’s disease and 34 by ulcerative colitis) and normal colon of 38 control subjects. IL-17 and interferon (IFN)-gamma were detected by ELISA in the supernatants of both unstimulated biopsies cultured ex vivo for 24h, and anti-CD3/CD28-stimulated lamina propria mononuclear cells (LPMCs) incubated for 48h with IL-12, IL-23, IL-1β plus IL-6, trasforming growth factor (TGF)-β1, or anti-IL-21 neutralising antibody. Intracellular flow cytometry was performed to analyse the percentage of IL-17- and IFN-gamma-producing mucosal T cells. Results: IL-17 production by organ culture biopsies was significantly higher in IBD inflamed mucosa than in controls, and was equivalent in amount to IFN-gamma. Anti-CD3/CD28-stimulated IBD LPMCs produced higher amounts of IL-17 and IFN-gamma compared to controls. The percentage of IL-17/IFN-gamma-producing mucosal T cells was increased in IBD patients in comparison to controls, most of them being CD161+. IL-23 and IL-1β plus IL-6 had no effect on IBD LPMC production of IL-17, which conversely was significantly decreased by IL-12. TGF-β1 dose-dependently decreased IFN-gamma, but had no significant inhibitory effect on IL-17. Blocking IL-21 significantly down-regulated IL-17 production. Conclusions: Our findings support a role for TGF-β1, IL-21 and IL-12 in modulating IL-17 production in IBD. Moreover, the abundant IL-17 in inflamed mucosa may help explain the relative lack of efficacy of anti-IFN-gamma antibodies in clinical trials. # L. Inflammatory bowel diseases 1. Basic science
Figure 1. ROC curves of two ANN models for detecting new cases of malignancy in an unselected population referred to UE. Accuracy values corresponding to ASGE guidelines, age and alarm features, and only alarm features.
those patients who need UE. The simple presence of alarm features (i.e. without the age criteria) improves the specificity and it could be used to prioritize the waiting list. # Q. Diagnostic endoscopy 1. Upper GI
CS.3.6 EVALUATION OF BOWEL PREPARATION AND PROCEDURE FOR PILLCAM® COLON CAPSULE. AN INTERIM ANALYSIS C. Spada ∗ , M.E. Riccioni, L. Petruzziello, P. Cesaro, G. Costamagna
CS.3.5 APPROPRIATENESS GUIDELINES AND PREDICTIVE RULES TO SELECT PATIENTS FOR UPPER ENDOSCOPY: A NATIONWIDE, MULTICENTER STUDY ON BEHALF OF THE ITALIAN SOCIETY OF DIGESTIVE ENDOSCOPY (SIED) L. Buri ∗ , G. Bersani, C. Hassan, M. Anti, M.A. Bianco, L. Cipolletta, E. Di Giulio, G. Di Matteo, L. Familiari, L. Ficano, P. Loriga, S. Morini, V. Pietropaolo, A. Zambelli Appropriateness SIED Group, Sied Background and aim: To efficiently select patients referred for upper endoscopy (UE) is of major importance. Material and methods: SIED designed a nationwide, multicentre, prospective study enrolling patients consecutively referred to UE in a period of 1 month. Before UE, the endoscopist assessed the appropriateness of the referral with ASGE guidelines, also collecting clinical and demographic variables. Outcome of the study was the new diagnosis of malignancy at UE. Sensitivity and specificity of five different options were compared: 1) ASGE appropriateness guidelines 2) Age and alarm features (>45 yrs. or alarm feature vs. <45 yrs. without alarm features) 3) Only alarm features 4) Stepwise multiple logistic regression analysis 5) Artificial neural networks (ANN) models. Results: 7,193 patients were enrolled in 57 centres and 215 new cases of gastro-oesophageal cancers were detected. As shown in Figure 1, sensitivity and specificity of ASGE guidelines for cancer (97.1%/20.4%; LR+: 1.2; LR-: 0.14) were very similar to those of age and alarm features (97.2/22.7%; LR+: 1.2; LR-: 0.12). Only alarm features without age criteria were highly specific (73%), but less sensitive (75%). The overall accuracy of only alarm features (74%) was similar to those of multivariate model (77%) and ANN models (79%). Conclusions: A simple predictive rule based on age and alarm feature is equally effective to the more complex ASGE guidelines in excluding the risk of gastro-oesophageal cancer, and it could be used to select
Catholic University, Digestive Endoscopy Unit, Roma Background and aim: Recent studies have shown that PillCam COLON capsule endoscopy (PCCE) is safe and can demonstrate colon polyps and cancers. However, as with other colon imaging techniques, the performance is dependent on the preparation quality. PCCE requires adequate cleansing for as long as 10 hours until capsule excretion. The current prospective blinded pilot study aims to evaluate modified PCCE regimen in comparison to baseline regimen in Italian patients. Material and methods: The study includes 100 patients referred to colonoscopy for suspected or known colonic disease. They are randomized to baseline or modified regimen and ingest the capsule in the morning. Baseline regimen includes clear liquid diet and 3 liters of Polyethylene Glycol (PEG) on day -1, and 1 liter PEG in the morning of the procedure day. Additional small doses of Sodium Phosphate (NaP) boosters are given following capsule ingestion to maintain clean colon and enhance capsule propulsion. The modified regimen included additional low residue diet on days -5 to -2, 4 senna tablets on day -2, clear liquid diet and 3 liters of PEG on day -1, and PEG instead of NaP boosters on the procedure day. Four-point grading scale applied in previous studies was used to objectively describe the cleansing level of the colon: excellent, good, fair, and poor. Following capsule excretion, colonoscopy was performed on same day. Results: This interim analysis includes 40 patients, mean age 59 (27-74) years, enrolled in single center and randomized to the base line (n=20) and modified (n=20) regimens. Severe adverse events: headache in one patient of the baseline regimen, and headache, nausea and abdominal pain in 3 patients of the modified regimen. Colon cleansing level during PCCE and procedure completion rate are provided in the Table. The sensitivity and specificity of PCCE compared to colonoscopy for polyps = 6 mm was 55% (CI 37%-55%) and 100% (CI 79%-100%) for the baseline regimen, and 80% (CI 43%-96%) and 79% (CI 65%-84%) for the modified regimen.
Abstracts / Digestive and Liver Disease 41S (2009), S1–S167 Table 1 % of patients with Good-to% of PCCE procedures Excellent cleansing level during completed within 10 hours PCCE procedure post ingestion Baseline regimen Modified regimen
37% (CI 16–62%) 53% (CI 29–76%)
100% 75% (additional 20% at sigmoid-rectum)
Conclusions: The interim analysis suggests that the modified regimen tends to provide superior PCCE colon cleansing and sensitivity of polyp detection. The finding should be further confirmed with expansion of the study to 3 sites. # Q. Diagnostic endoscopy 6. Video-capsule
CS.3.7 BOWEL ULTRASONOGRAPHY FOR DIAGNOSIS OF INTESTINAL ACUTE GRAFT VERSUS HOST DISEASE AFTER ALLOGENEIC HAEMATOPOIETIC STEM CELL TRANSPLANTATION E. Calabrese ∗ , E. Visconti, M. Postorino, G. Del Vecchio Blanco, L. Franceschini, R. Cerretti, G. De Angelis, C. Petruzziello, S. Onali, L. Biancone, W. Arcese, F. Pallone Università di Roma Tor Vergata, Roma Background and aim: Acute Graft Versus Host Disease (aGVHD) is a major complication during the first 100 days after allogeneic haematopoietic stem cell transplantation (AHSCT), showing a 20-50% incidence. Target organs include the skin, the liver and the gastrointestinal (GI) tract. GI aGVHD shows a progressive worsening of the symptoms, but the prognosis may be improved by an early diagnosis and
S9
timely treatment. Diagnosis is based on clinical assessment, although additional information include histology of the upper and lower GI tract, requiring invasive endoscopic procedures in severely ill patients. Aims. To investigate the sensitivity, specificity and accuracy of Bowel Ultrasonography (BUS) in detecting GI aGVHD, when using clinical assessment as a gold standard. We also aimed to compare, in a subgroup of patients, BUS vs colonoscopy with histology in diagnosing of GI aGvHD. Material and methods: From Aug. 2004 to July 2008, 52 pts underwent AHSCT. All 52 pts developed GI symptoms, including nausea, vomiting, watery diarrhoea, GI bleeding. Symptoms onset began after a median of 30 days from AHSCT. BUS was performed within 48 hours from the onset of GI symptoms. Findings compatible with GI aGvHD were: increased bowel wall thickness (small bowel>3mm, colon>4 mm), dilation (small bowel >25 mm, colon >40 mm). The site of the lesions were also reported (small bowel and/or colon). Results: Clinical assessment lead to a diagnosis of GI aGvHD in 17/52 pts, no GI aGvHD was detected in 20/52 pts, while 15 pts were not able to complete the study (12 deaths, 3 due to severe clinical conditions). BUS detected findings compatible with GI aGvHD in 16/17 pts and showed a 94% sensitivity, 95% specificity, 94% PPV, 95% NPV and 94.5% accuracy. Colonoscopy was performed in 13/52 pts, showing clinical assessment compatible with GI aGvHD in 11/13. In these 11 pts, histology confirmed the diagnosis of GI aGvHD, and BUS detected findings compatible with GI aGvHD in all 11 pts, being negative in the 2 pts with no GI aGvHD. Conclusions: BUS shows a high sensitivity and accuracy in detecting of GI aGvHD and a high concordance with endoscopic and histologic findings. BUS is a feasible and non-invasive diagnostic tool which may provide useful information for a prompt and appropriate treatment of pts with suspected aGVHD. # T. Imaging techniques 1. Ultrasounds