Expression Data From Crohn’s Disease Small Bowel Resections Indicate Distinct Clinical Sub-Groups

Expression Data From Crohn’s Disease Small Bowel Resections Indicate Distinct Clinical Sub-Groups

AGA Abstracts Spiking recovery presented by each assay Su1949 EXPRESSION DATA FROM CROHN'S DISEASE SMALL BOWEL RESECTIONS INDICATE DISTINCT CLINICAL...

1MB Sizes 0 Downloads 32 Views

AGA Abstracts

Spiking recovery presented by each assay

Su1949 EXPRESSION DATA FROM CROHN'S DISEASE SMALL BOWEL RESECTIONS INDICATE DISTINCT CLINICAL SUB-GROUPS Alka A. Potdar, Dalin Li, Talin Haritunians, Dermot McGovern, Stephan R. Targan, Janine Bilsborough Background: Crohn's disease (CD) is a chronic Inflammatory bowel disease most commonly affecting the small bowel (SB). CD behavior sub-phenotypes include stricturing, penetrating and perianal CD (pCD). Such heterogeneity provides a challenge for the development of effective therapies. Thus, defining a unique gene expression signature for these clinical phenotypes would ultimately aid in better definition of patient sub-groups and identification of pathways associated with the pathogenesis of these subsets. Here, we focus on identifying clinically relevant sub-groups using expression data from uninvolved tissue taken from small bowel resections. Methods: A total of 139 Caucasian CD patients were included in this study. Expression datasets using Agilent microarray platform were generated in two batches (SB85, n=85 and SB54, n=54). Hierarchical and kmeans clustering using the first three principal components in normalized expression data with SB85 indicated the presence of three sample clusters (CD1, CD2 and CD3). Combining SB85 with SB54 after removal of batch effects preserved these three sub-groups (Figure1). Clinical phenotype and genotype data (generated using ImmunoChip) were analyzed for associations. Results: We identified the most distant sub-groups (CD1, n=12 and CD3, n=20) of SB85 to be clinically different. CD3 was associated with disease recurrence after first surgery (OR= 6.78, P=0.04) and presence of pCD (OR=5.0, P=0.04). Survival analysis using time from first surgery to recurrence or last follow-up indicated the proportion of recurrence free survival in CD3 was smaller than CD1 (p=0.02, median survival time (days), CD1=315 and CD3=189). These data suggest CD3 represents individuals with more severe disease. No significant differences were found when comparing the three sub-groups simultaneously for differences in clinical phenotypes. Pathway analysis using the fold changes of differentially expressed genes in the three sub-groups indicated EIF2 signaling was activated in CD1 compared to CD3. Similar results were obtained using combined SB85 and SB54 datasets (n=139), with the CD3 sub-group showing an increase in second surgeries and higher pCD occurrence (Table1). Gender was also a predictor of clustering (Table1). The pCD associated CD3 subgroup had significant genetic associations with SNPs in DAPK1 locus, a known pCD gene locus (Kaur M., et al. in IBD journal) compared to CD1 (intronic SNP rs888338, OR=3.5, P= 6.5 x 10-4). Conclusions: We identified two clinically distinct CD sub-groups (CD1 and CD3) using expression data from uninflamed ileal tissue from SB resections. CD3 represented a severe patient sub-group, as defined by an association with disease recurrence, second surgery and pCD. These data potentially identify patient subsets that may need aggressive treatment after first resection. Table1: Characterizing the CD1 and CD3 subgroups using the combined dataset (n=139).

Su1948 THERAPEUTIC DRUG MONITORING OF BIOSIMILARS OF INFLIXIMAB CAN BE ASSESSED BY THE NEW INFLIXIMAB POINT-OF-CARE QUANTITATIVE TEST Joana Afonso, Helena T. Sousa, Isadora Rosa, Jooa Carvalho, Camila Dias, Fernando Magro Background and Aims Therapeutic Drug Monitoring (TDM) is an effective strategy in the management of inflammatory bowel disease (IBD) patients and is widely used in the adjustment of the originator infliximab therapy. CT-P13, a biosimilar of the originator infliximab (IFX), has been recently approved for the treatment of IBD. The aim of this study was to validate a point-of-care IFX (POC IFX) device, already available in the market, for quantification of IFX biosimilar CT-P13 by comparing it with three validated ELISA assays. Methods POC IFX assay and three ELISA-based established assays were used to analyse 184 serum samples of IBD patients treated with CT-P13. The results were statistically compared both in quantitative and qualitative terms. Intraclass Correlation Coefficient (ICC) was assessed for quantitative comparison and both accuracy and kappa (95% CI) statistics were used for qualitative analyse. Exogenous CT-P13 was spiked in donors' serum samples to assess spiking recovery. Results Both quantitative and qualitative comparison showed an excellent agreement between POC IFX assay and the three ELISA-based established methods. ICC was 0.907 and 0.935 for POC IFX/in-house and POC IFX/r-biopharm, respectively. For qualitative comparison, accuracy and kappa (95% CI) statistics were determined after stratification of results by therapeutic interval (<3, 3-7 and >7). Results are presented in table 1. Figure 1 depicts the spiking recovery presented by each assay. POC IFX assay revealed an average spiking recovery percentage of 102%. Conclusion POC IFX assay is a new methodology validated and available in the market to assess IFX originator concentration. Our results showed a good agreement with ELISA-based established assays when used to assess IFX biosimilar and an excellent spiking recovery. TDM of CT-P13 can be assessed using this new methodology that delivers results in only 15 min. Accuracy and kappa (95% CI) after stratification of results by therapeutic interval (<3, 3-7 and >7).

AGA Abstracts

S-608

Figure1: 3-D scatter plot of first three principal components using combined dataset (n= 139). The colors indicate the three clusters identified by kmeans clustering.

Su1950 THE SENSITIVITY AND SPECIFICITY OF NON-INVASIVE BLOOD-BASED TESTS AND HISTOPATHOLOGY FOR DETECTING CYTOMEGALOVIRUS REACTIVATION IN INFLAMMATORY BOWEL DISEASE: A SYSTEMATIC REVIEW AND META-ANALYSIS Parul Tandon, Tushar Shukla, Paul D. James, Ranjeeta Mallick, Jeffrey D. McCurdy

Figure 1: Scatter dot plot with median (wide line) and interquartile range (narrow line) of fecal calprotectin values on each day after 300ml blood ingestion.

Background and aims: Cytomegalovirus (CMV) reactivation in inflammatory bowel disease (IBD) is associated with an increased risk of colectomy, health-care utilization, and in some instances, mortality. Several diagnostic modalities, including histopathology (H&E) and noninvasive blood-based tests, are available to assist in the diagnosis of CMV reactivation, though it remains unclear if they are sufficiently accurate compared to immunohistochemistry (IHC) or tissue PCR (tPCR). As such, we assessed the diagnostic accuracy of blood-based tests and H&E in detecting colonic CMV reactivation. Methods: A systematic search of electronic databases was performed from inception through July 2015. Observational studies comparing two or more diagnostic tests for CMV reactivation in IBD were identified. The primary outcome was the diagnostic accuracy of blood-based tests (pp65 antigenemia or wholeblood polymerase chain reaction (bPCR)) and H&E for predicting colonic CMV reactivation compared with IHC or tPCR as reference standards. Weighted summary estimates with 95% confidence intervals were calculated using bivariate analysis. Study quality was assessed by the QUADAS-2 tool. Results: Nine studies with 459 patients assessed the diagnostic accuracy of blood-based tests: 5 studies by pp65 antigenemia (272 patients) and 4 studies by bPCR (187 patients). The pooled sensitivity and specificity of blood-based tests was 50.8% (95% CI, 19.9-81.6%) and 99.9% (95% CI, 99-100%) respectively. The sensitivities of pp65 antigenemia and bPCR were 39.7% (95% CI, 27.4-52.1%) and 60.0% (95% CI, 46.5-73.5%) respectively. These remained stable in subgroup analysis (by study type and geographical location). Nine studies (297 patients) examined the diagnostic accuracy of H&E compared with IHC or tPCR. The sensitivity and specificity of H&E was 12.5% (95%CI, 3.6-21.4%) and 99.1% (95% CI, 97.9-100%) respectively. Study quality was low or unclear for at least one domain for all studies when assessed by the QUADSAS-2 tool. Conclusion: Noninvasive blood-based tests and histopathology appear to be insensitive diagnostic tests and should not replace IHC or tPCR for determining CMV reactivation in IBD.

Figure 2: Cumulative fraction of patients with at least one elevated fecal calprotectin test (>50ug/g) after ingestion of 100ml (black dots) and 300ml (grey squares) blood.

Su1952 DIGITAL HOLOGRAPHIC MICROSCOPY RELIABLY DETECTS FIBROSIS IN STRICTURING CROHN'S DISEASE AND A PILOT STUDY Arne Bokemeyer, Philipp Lenz, Emile Rijcken, Florian Rieder, Björn Kemper, Dominik Bettenworth OBJECTIVE: Intestinal strictures are a frequent complication of Crohn's Disease (CD). Differentiation of the inflammatory from the fibrotic component of strictures is crucial for the right choice of therapy. Current diagnostic modalities have a limited capability to assess the presence and degree of fibrotic alterations within CD-associated strictures. Digital holographic microscopy (DHM) enables stain-free quantitative phase contrast imaging and provides determination of the refractive index (RI), which is directly related to tissue density. The aim of this study was to evaluate DHM for assessing the grade of fibrosis in surgical specimens from patients with stricturing CD. METHODS: Full thickness surgical resection specimens (n=14) were obtained from patients with symptomatic CD strictures. Clinical characteristics were extracted from medical records. Cryostat sections from stenotic and non-stenotic bowel segments for each patient were evaluated separately by conventional H&E staining and were simultaneously analyzed by DHM. To quantify tissue density, RI measurements were performed in the epithelium (e), submucosa (sm) or muscularis propria (mp). For every tissue layer, up to 10 quantitative DHM phase images of adjacent areas were captured in which in 10 representative regions of interest (ROI) refractive index measurements were performed. RESULTS: Clinical disease activity scores as well as laboratory examinations documented a moderate disease activity (CD activity index [CDAI]: 162 ± 103, white blood cell count: 10.5 ± 3.5 x109/l, C-reactive protein: 3.8 ± 4.4 mg/dl). The majority of included patients underwent ileocecal resection (5/7) while two patients were treated by right hemicolectomy. The average length of stenosis was 11.1 ± 2.0 cm. Histopathological evaluation by an expert pathologist confirmed the presence of fibrosis in stenotic segments. Non-stenotic segments did not show any fibrotic alterations. Employing DHM, 166 digital holograms were generated in 14 surgical specimens and ultimately 1660 measurements within defined ROIs were performed. Determination of average RI allowed for successful differentiation between the three layers of the intestinal wall. The average RI determined for the muscularis propria of stenotic and non-stenotic samples significantly differed as assessed by DHM (mp: P < 0.001). No significant changes were detected regarding the

Su1951 INGESTION OF 100ML AND 300ML BLOOD MIMICKING UPPER GI BLEEDING LEADS TO SIGNIFICANT CALPROTECTIN ELEVATION IN HEALTHY VOLUNTEERS AND RESULTS FROM THE VAMPIRE STUDY™ Stephan R. Vavricka, Henriette Heinrich, Benjamin Misselwitz, Simon Buetikofer, Emanuel Burri, Gerhard Rogler, flavia breitenmoser, Xiaoye Schneider-Yin, Jonas Zeitz, Luc Biedermann, Matthias Sauter INTRODUCTION AND AIM: Fecal Calprotectin (fC), a calcium binding protein abundant in neutrophiles, is increasingly gaining importance as a noninvasive biomarker for intestinal inflammation. It correlates with both histological and clinical activity in inflammatory bowel disorders (IBD) and can predict IBD relapse [1,2]. However, gastrointestinal (GI) bleeding might induce elevated fC levels, erroneously suggesting intestinal inflammation. To the best of our knowledge, the interference of present intraluminal blood in the GI tract on fC values has not yet been systematically assessed. METHODS AND ANALYSIS: 15 healthy volunteers (HV) (mean age 25 years, range 23-33 years) without GI symptoms or known GI disease and normal fC baseline values ingested 100 and 300ml of their own blood in a randomized order by drinking or via nasogastric tube, with a 28 day wash out period. fC and fecal occult blood test (FOBT) as well as the occurrence of visible melena were assessed at baseline, on day 0-7 and day 14. fC was measured by a commercially available particle-enhanced turbidimetric immunoassay (fCALTM turbo; Buehlmann Laboratories Ltd, Switzerland). fC > 50 ug/g was defined elevated. RESULTS: Ingestion of blood was tolerated well by all HV with only slight symptoms such as nausea, bloating or heartburn in 7 of 15 HV (46%), lasting <24 hours. A total of 304 stool samples were analyzed. Melena was reported by all 15 HV after 300ml and by 10 out of 14 HV after 100ml blood ingestion (71%). Upon

S-609

AGA Abstracts

AGA Abstracts

ingestion of 300ml blood mean fC levels rose significantly within 5 days compared to baseline (p=0.0025) and at least one fC value above 50ug/g was observed in 9/15 (60%) of HV (Fig 1). Increase in fC levels after ingestion of 100ml was also significant over baseline (p=0.028); calprotectin levels above 50ug/g were observed in 7/15 (46%) of HV. Pronounced increases in fC levels above 200ug/g were rarely observed (1 individual (6%) after 100ml, none after 300ml ingestion). FOBT testing became positive in 14 (93%) and 6 (40%) of 15 HV, after ingesting 300ml and 100ml of blood, respectively. FOBT and fC levels showed a positive correlation, fC levels being significantly higher in FOBT-positive samples than FOBTnegative samples (p<0.0001). CONCLUSION: Ingestion of 100ml and 300ml blood led to significant fC rise in the majority of participants. Very high fC levels (>200ug/g) were rarely observed (6% of participants). We conclude that upper GI-Bleeding should be considered as a potential reason for otherwise unexplained mild fC elevation. 1. Smith LA,Gaya DR. Utility of faecal calprotectin analysis in adult inflammatory bowel disease. World J Gastroenterol. 2012 Dec 14;18(46):6782-9. 2. Tibble JA,Sigthorsson G, Bridger S et al. Surrogate markers of intestinal inflammation are predictive of relapse in patients with inflammatory bowel disease. Gastroenterology. 2000 Jul;119(1):15-22.