St Ward

The learning curve to achieve satisfactory completion rates in upper GI endoscopy: an analysis of a national training database

Objective The aim of this study was to determine the number of OGDs (oesophago-gastro-duodenoscopies) trainees need to perform to acquire competency in terms of successful unassisted completion to the second part of the duodenum 95% of the time. Design OGD data were retrieved from the trainee e-portfolio developed by the Joint Advisory Group on GI Endoscopy (JAG) in the UK. All trainees were included unless they were known to have a baseline experience of >20 procedures or had submitted data for <20 procedures. The primary outcome measure was OGD completion, defined as passage of the endoscope to the second part of the duodenum without physical assistance. The number of OGDs required to achieve a 95% completion rate was calculated by the moving average method and learning curve cumulative summation (LC-Cusum) analysis. To determine which factors were independently associated with OGD completion, a mixed effects logistic regression model was constructed with OGD completion as the outcome variable. Results Data were analysed for 1255 trainees over 288 centres, representing 243 555 OGDs. By moving average method, trainees attained a 95% completion rate at 187 procedures. By LC-Cusum analysis, after 200 procedures, >90% trainees had attained a 95% completion rate. Total number of OGDs performed, trainee age and experience in lower GI endoscopy were factors independently associated with OGD completion. Conclusions There are limited published data on the OGD learning curve. This is the largest study to date analysing the learning curve for competency acquisition. The JAG competency requirement for 200 procedures appears appropriate.

Selective recruitment and retainment of regulatory T cells in human colorectal cancer

Abstract Background Colorectal cancer (CRC) is the third most common malignancy in the UK, and lymphocytic infiltration is associated with improved survival. Analysis of lymphocytic infiltration yields a prognostic ability rivalling the TNM tumour staging system. Regulatory T cells (Treg) are known to be enriched in CRC, and it is postulated that they promote immunological tumour escape by suppression of effector T-cell responses. Little is known about the signals controlling entry of Treg into CRC. Methods Matched CRC, distal colonic tissue, draining lymph node, and blood were obtained from patients undergoing resection of CRC. Tumour-infiltrating lymphocytes were isolated and phenotyped for chemokine receptors with flow cytometry. The presence of tissue chemokines was analysed with real-time PCR and western blotting. Standard chemotaxis and suppression assays were performed. Peripheral blood lymphocytes were co-cultured with tumour supernatant, and effects on lymphocyte phenotype were assessed. Findings The proportion of T cells with a Treg phenotype was significantly increased in CRC compared with that in colonic tissue (14·8% vs 5·1%, p Interpretation Conditions exist to recruit CCR5+ Treg into human CRC. CCR5 upregulation is promoted by the tumour microenvironment, providing a retention signal. CCR5 inhibition may prove to be a novel immunotherapy for CRC by blocking the recruitment and egress of suppressive Treg, promoting an anti-tumour immune response. Funding UK Medical Research Council.

A novel subset of functional interleukin-10 secreting CD8 regulatory T cells infiltrate human hepatocellular carcinoma

Abstract Background Tumour specific effector T cells can be detected in the blood and tumours of patients with hepatocellular carcinoma but they fail to mount effective immune responses. Attempts to amplify anti-tumour immune responses using immunotherapy show promise, but are hampered by the presence of suppressive regulatory T cells (Tregs) that inhibit anti-tumour immune responses. Tregs are crucial in the maintenance of immune homoeostasis and in the prevention of autoreactive immune response but in the context of cancer they can suppress beneficial anti-tumour immunity leading to tumour progression. A novel subset of CD8 expressing Tregs has recently been described and we now report the presence of such cells in human hepatocellular carcinoma and define their functional and homing properties. Methods Fresh tissue from hepatocellular carcinoma and matched distal non-involved tissue were obtained from patients undergoing liver resection or transplantation at the Queen Elizabeth Hospital, Birmingham, after informed consent. Liver-derived T cells were isolated and phenotyped with multicolour flow cytometry including intracellular cytokine staining. CD8 + Tregs were isolated with a Mo-Flow cell sorter for functional assays. Distribution of CD8 + Tregs was investigated by immunohistochemistry and immunofluorescence. Findings The percentage of CD8 + Tregs (defined as CD8 + CD25 high CD127 low FoxP3 + ) infiltrating hepatocellular carcinoma tumours was significantly greater than matched non-involved liver. Tumour-derived CD8 + Treg isolated by Mo-Flo sorting suppressed allogeneic effectors cells in vitro and secreted interleukin (IL) 10. By contrast, T-cell interferon-γ production was decreased within the tumour compared with matched non-involved liver. The chemokine receptor CXCR3 which is involved in T-cell recruitment to the inflamed liver was highly expressed on tumour-derived CD8 + Tregs. Interpretation A novel subset of functional IL-10 secreting CD8 + Tregs may suppress anti-tumour immunity in hepatocellular carcinoma. Their expression of CXCR3 provides a potential mechanism for recruitment into the tumour environment. Funding Wellcome Trust.

Patients newly diagnosed with ulcerative colitis receive earlier treatment in surgical clinics

The diagnosis and treatment of ulcerative colitis (UC) is traditionally the realm of gastroenterologists. However, the symptoms of UC overlap with those of bowel cancer and patients may be initially referred to colorectal surgery clinics. The aims of this study were to define which specialty most frequently diagnoses UC and to determine if there were differences in management between the two specialities.

PWE-048 The virtual electronic chromoendoscopy score in ulcerative colitis exhibits very good inter-rater agreement in scoring mucosal and vascular changes after computerised module training: a study across academic and community practice

Introduction Mucosal healing is the desired therapeutic endpoint for clinical trials in ulcerative colitis (UC). However, conventional white light endoscopy may fall short of capturing the full spectrum of inflammatory change; and virtual electronic chromoendoscopy (VEC) can show ongoing disease activity even when Mayo scores suggest healing (Iacucci et al. Endoscopy 2015 and 2017). Applicability of VEC scoring requires determination outside the expert setting; thus, our aim was to provide external validation among trainees, consultant gastroenterologists and colorectal surgeons, practicing across six general and specialist centres. Method 15 participants reviewed a computerised training module outlining HD and i-Scan modes. Anchor points for the VEC score indicated mucosal changes (crypt distortion, 0 [A–C]; microerosions, I [1–3]; erosions, II [1–3]; and ulceration, III [1–3]) and vascular alterations (non-dilated vessels, 0 [A–C]; dilated/crowded vessels, I [1–3]; mucosal bleeding, II [1–3]; and intraluminal bleeding, III [1–3]). Performance accuracy was tested using a video library pre-/post-training (n=30). Agreement between raters was tested for the Mayo score, UCEIS and VEC score, and results correlated with histology (New York Mount Sinai system). Results The inter-rater agreement was very good for the Mayo score, UCEIS scoring erosions/ulcers and overall, and for VEC scoring mucosal patterns in both modules (Table 1). For the vascular components of UCEIS agreement was only moderate, and did not improve post-training; unlike the agreement for VEC vascular patterns which improved significantly to very good. Correlation between histology and VEC score was highly significant for mucosal and vascular scoring (Spearman’s ρ: 0.910, p<0.001; and 0.907, p<0.001; respectively, Figure 1). This was superior to the Mayo score (0.876, p<0.001) and UCEIS (0.887, p<0.001). Conclusion The VEC score demonstrates very good inter-observer agreement across all levels of experience and provides excellent correlation with histology. Unlike UCEIS, the VEC score does not have subjective elements (e.g. mucosal erythema, incidental/contact friability) and may better delineate vascular changes due to filter technology. Given the ability to define subtle endoscopic features, VEC may be applied to further stratify treatment paradigms for patients with UC. Disclosure of Interest P. Trivedi Conflict with: Received funding from the National Institute for Health Research (NIHR), Conflict with: This article presents independent research funded by the NIHR. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health, S Ghosh: None Declared, M Iacucci: None Declared Abstract PWE-048 Table 1 Abstract PWE-048 Figure 1

Hepatocellular carcinoma surveillance in hepatitis B virus–infected individuals: Who and how?

We recently reported that, in a large cohort of biopsy-proven patients with nonalcoholic fatty liver disease, severe steatosis, diagnosed by histology or by ultrasound, is independently linked to increased liver stiffness measurement (LSM) values. As a consequence, we reported higher rates of false-positive LSM results for the noninvasive assessment of F2-F4 and F3-F4 fibrosis by transient elastography in patients with severe steatosis compared with their counterparts. Accordingly, we suggested that in nonalcoholic fatty liver disease patients the presence of severe steatosis, per se or evaluated by ultrasound, should always be taken into account in order to avoid overestimations of liver fibrosis. Notably, we reported similar results in the setting of chronic hepatitis C patients as well. We thank Lupsor and colleagues for their interest in our study. According to our hypothesis, they tested whether steatosis noninvasively assessed by the controlled attenuated parameter (CAP), a measure recorded using a different software in the FibroScan, overestimates LSM in a small cohort of 30 patients with F2 fibrosis and with chronic liver disease due to different etiologies. Notably, other than to confirm the expected association between CAP and body mass index, they reported significantly higher LSM values in patients with CAP >280 m/second compared with their counterparts. However, in multivariate analysis they did not confirm CAP as an independent predictor of overestimation of F3-F4 fibrosis, using LSM>9.5 as a threshold. We think that data from Lupsor and colleagues are interesting and can represent a starting point to test, in large cohorts of patients well characterized for liver damage and discriminated for etiology of liver disease—nonalcoholic fatty liver disease and chronic hepatitis C considered separately are the most relevant clinical settings—whether the combination of LSM with CAP could improve the diagnostic accuracy of LSM for staging fibrosis. Concerning the lack of evidence in multivariate analysis of CAP as an independent predictor of fibrosis overestimation reported by the authors, we think that the value of a multivariate analysis in a very small number of patients is too limited to draw conclusions.

PTH-098 Connecting liver and gut in psc: ccl25 expression is upregulated in colitis, correlates with inflammatory activity and facilitates effector ccr9+ lymphocyte recruitment

Introduction Recruitment of mucosal T-cells to the liver in response to aberrantly expressed homing signals drives hepatobiliary inflammation in primary sclerosing cholangitis (PSC). This is exemplified by the ‘gut-homing’ chemokine CCL25, which recruits intestinal CCR9+effector cells to the PSC liver. However, previous studies report CCL25 expression as being confined to the small bowel whereas PSC is typically associated with colonic inflammation. Our aim was to determine whether CCL25 and CCR9 are expressed in the inflamed human colon in patients with colitis. Method Mucosal biopsies were obtained during surveillance colonoscopy (n = 40) and colonic tissue from patients undergoing surgical resection for colitis (n = 6), or non-colitis-associated cancer (n = 5). CCL25 mRNA was evaluated by qRT-PCR (relative to GUSb) and protein expression confirmed using western blotting and tissue-ELISA. Lymphocyte CCR9 expression was quantified by flow cytometry, and the ability of sorted a4b7+CCR9+and a4b7+CCR9-T-cells to transmigrate across hepatic sinusoidal endothelium (HSEC) investigated in flow-based adhesion assays. Results CCL25 mRNA and protein were absent from normal colon but present in colitis. CCL25 mRNA expression correlated with endoscopic Mayo score (p = 0.001) and tissue levels of TNFa (Spearman’s rho 0.811; p < 0.0001) as indices of inflammatory activity (Figure 1). In colitis, >90% of CD3+CD4+and >15% of CD8+T-cells expressed CCR9 (6% and 2% in normal colon; p = 0.01 and 0.03) and were predominantly CD127+effector lymphocytes. a4b7+CCR9+T-cells showed enhanced adhesion and transendothelial migration across HSEC compared with a4b7+CCR9-T-cells (p < 0.05).Abstract PTH-098 Figure 1 Conclusion CCL25 is expressed in the inflamed human colon, correlates with inflammatory activity and is associated with high frequencies of CCR9+tissue-infiltrating lymphocytes. These findings lend further support to the mucosal lymphocyte homing hypothesis of PSC and show how it can be applied to patients with colitis. Disclosure of interest None Declared.

PTH-058 Trainee Colonoscopists Acquire Competency at Different Rates, as Determined by Cusum Analysis Of Colonoscopy Data From The Jets Database

Introduction The number of colonoscopies required to reach competency is not well established. Nevertheless, a minimal number forms part of UK certification criteria. The Cusum technique is a statistical analysis of sequential data to determine if a process is ‘in control’. The Joint Advisory Group on GI Endoscopy have developed an e-portfolio for users to record their endoscopic experience. The primary aim of this study was to determine the range of experience required by individuals to attain a caecal intubation rate (CIR) ≥90%, as defined by Cusum. A secondary aim was to assess which training factors are associated with attaining competence. Methods Inclusion criteria were all e-portfolio users who had performed ≤50 (‘baseline’) colonoscopies prior to submission of data to the e-portfolio; termed ‘trainees’. All colonoscopy records for the trainees were retrieved from the e-portfolio database and learning curve-Cusum analysis was performed. This analysis of colonoscopy completion reports the number of procedures required for CIR performance to reach ≥90%. A colonoscopy was defined complete if the caecum or ileum was reached and was performed without assistance. Trainees who had attained a CIR≥90% were compared to those with a CIR < 90% for differences in previous endoscopic experience, case volume and other trainee factors by univariate (Mann-Whitney, Chi-squared) and multivariate (binomial logistic regression) analysis. Results The e-portfolio contained 169,515 colonoscopy records entered by 1,572 different users. 265 users (‘trainees’) were confirmed to have performed ≤50 baseline colonoscopies and were included in subsequent analyses. By Cusum method, 39 trainees attained a CIR≥90%; 226 achieved a CIR < 90%. For those trainees with over 250 procedures, only 47% attained a CIR≥90%. Factors associated with attaining CIR≥90% were high number of procedures (P < 0.01), high number of colonoscopies per month (P < 0.01), and prior experience of more than 100 sigmoidoscopies (P = 0.017) by univariate and multivariate analysis. Nurse endoscopists attained competency at a higher rate than gastroenterology or surgical trainees by univariate (P = 0.01) but not multivariate analysis. Conclusion This is the largest study to date by both procedure and trainee numbers assessing colonoscopy competency by Cusum method. Trainees achieve competency at different rates. A high proportion of trainees will not attain a CIR > 90% even after 250 procedures. High case volume and prior sigmoidoscopy experience are associated with a CIR > 90%. The potential of both these factors to influence the attainment of competency should be exploited within endoscopy training programmes. Disclosure of Interest None Declared.

PWE-113 Low Risk for Hepatocellular Cancer (HCC) in Hepatitis B Virus (HBV) Infected Asian Migrants: Implications for Cancer Surveillance

Introduction The global incidence of HCC is rising and it is the third most common cause of cancer-related death worldwide. Surveillance of at risk patients has been recommended by expert guidelines to detect early cancers that are amenable to curative treatments. AASLD has recommended HCC surveillance for non-cirrhotic HBV-positive Asian males over 40 and females over 50yrs of age. However, the evidence to support this recommendation is limited and is derived from studies conducted in Asia. Results from such studies may not be applicable in the West, due to differences in environmental risk factors and availability of HBV treatment. Implementation of such a recommendation would place a burden on healthcare resources, and may not be justified if the risk for HCC is substantially lower than previous estimates in this population. Methods A retrospective study was carried out of all Asian patients undergoing follow up for HBV infection from 1990 to 2012. Patients were classified as cirrhotic or non-cirrhotic according to clinical, biochemical, radiological and histological results. Follow-up was until September 2012, and was censored at time of death, development of cirrhosis or loss to follow-up. Results Among 316 identified Asian patients with HBV, 73 non-cirrhotic patients fulfilled the proposed AASLD surveillance criteria, either at time of initial referral or during the period of follow-up. The median at-risk follow up period (as defined by AASLD guidelines for non-cirrhotic Asians) was 57 months (range: 0–354 months). HCC was diagnosed in one non-cirrhotic patient after 77 months of follow up (male, 60yrs), two patients became cirrhotic after 49 and 89 months (male, age 46 and 55yrs) and no deaths occurred. The overall incidence of HCC in the non-cirrhotic cohort meeting the AASLD surveillance criteria was 1 per 429.5 patient-years of follow-up (0.23% per patient-year). Conclusion The incidence of HCC in Asian patients with non-cirrhotic HBV is low in our cohort. This low incidence challenges the rationale for surveillance in this group of patients. More studies are needed to assess the benefit of such approach. Disclosure of Interest None Declared.