- Email: [email protected]
Sa1850 Should We Continue to Follow Patients With <1 CM Barrett's Esophagus (Irregular Z Line): Results From a Large, Multicenter, Cohort Study
Results: At baseline, for both IHC and FISH the frequency of detected abnormalities increased significantly with histological stage, while combining the techniques detected p53 abnormalities in 100% of patients with HGD and EAC (p < 0.001, Chi2-test). Out of the 91 patients 11 developed HGD or EAC after a median follow period of 67 months (incidence of 2.18 per 100 patient-years). IHC and FISH had HR of 17 (Cox regression; p=0,001) and 7,3 (p= 0,02), respectively (table 1). Combining FISH and IHC was superior to one of the techniques alone, with a sensitivity and specificity of 81.8% and 85.0% to predict progression to HGD/ EAC. Conclusion: FISH and IHC for assessing p53 abnormalities are complementary tools which lead to an increased detection rate of p53 aberrations and an improved accuracy for predicting progression to HGD/EAC in BE. Therefore both techniques should be used for risk stratification of Barrett's esophagus patients. Cox regression analysis of patient characteristics for progression free survival
Global and Gene Specific DNA Methylation Changes in Barrett's Epithelium and Adjacent Normal Mucosa Samuel C. Stevens, Julian A. Abrams, Clare LeFave, Charles J. Lightdale, Tamas A. Gonda Introduction. DNA methylation changes make a major contribution to cancer initiation and progression and can precede genetic alterations. This feature makes such epigenetic alterations attractive markers and targets of intervention. DNA methylation changes may also be observed in cancer associated normal epithelium. Although DNA methylation changes are known to be associated with BE it is less well understood what changes occur in adjacent normal epithelium and at what stage do most epigenetic alterations develop. Our goal was to compare the grade-matched DNA methylation signature of BE, adjacent normal cardia and squamous mucosa and identify (1) the grade of BE when DNA methylation changes occur; (2) the epigenetic similarities between adjacent non-neoplastic epithelium; (3) gene specific methylation changes associated with early neoplastic transformation. Methods. Biopsies were collected from patients with all grades of BE from every 2 cm of BE mucosa, from the squamous epithelium (2 cm above the Z line) and the endoscopic cardia. Global methylation status was assessed by LINE-1 pyrosequencing and the methyl acceptor assay. Gene specific methylation was assessed using the Fluidigm Next Generation Sequencing. 42 Genes were selected from published and validated list of differentially methylated genes in BE/EAC for analysis. Results. 254 biopsies from 56 patients were analyzed. 55% had no dysplasia or low grade dysplasia (ND, LGD) and 45% had high grade dysplasia (HGD) or carcinoma. There was a significant difference between Line-1 and gene specific methylation of cardia, squamous versus BE. A significant decrease in Line-1 methylation was noted between BE with ND/LGD versus HGD or intramucosal carcinoma. A similar trend was noted for the majority of genes that showed change in methylation associated with progression of dysplasia. There was no significant change in Line-1 or gene specific methylation of the squamous epithelium. However, a significant change in Line-1 methylation was noted in the cardia in LGD vs HGD. Of the 15 genes that showed hyper or hypomethylation in BE 12/15 showed similar changes in the cardia while only 3/15 showed similar changes in the squamous epithelium. Conclusions. Both global and gene specific methylation changes are associated with progression of BE and it is between LGD and HGD that most of these changes are detected. Although the DNA methylation status of both adjacent normal cardia and squamous epithelium are different from Barrett's mucosa, with progression methylation changes occur in the cardia and not in squamous epithelium. These results support targeting epigenetic interventions or chemoprevention in patients with LGD and identify an epitgenetic similarity between the cardia and Barrett's epithelium but not the adjacent squamous mucosa.
Sa1849 The Risk of Malignant Progression in Patients With Barrett's Esophagus and a Histologic Diagnosis of Indefinite for Dysplasia: A Dutch Nationwide Cohort Study Christine Kestens, Max Leenders, Johan Offerhaus, Jantine W. van Baal, Peter D. Siersema Introduction and aim: Histological assessment of Barrett's esophagus (BE) is the main diagnostic tool for risk stratification of the progression to esophageal adenocarcinoma (EAC). A diagnosis indefinite for dysplasia (IND) in BE is used when a diagnosis of genuine dysplasia is equivocal. Due to this uncertainty combined with the unknown risk of progression of IND to low-grade dysplasia (LGD), high-grade dysplasia (HGD) or EAC, follow-up includes repeat endoscopy with biopsies. We aimed to assess the risk of progression to high-grade dysplasia (HGD) or esophageal adenocarcinoma (EAC) after a diagnosis of IND in a nationwide cohort of patients with BE in the Netherlands. Methods: Patients with a first diagnosis of IND in BE were selected between 2002-2011 using PALGA, a nationwide registry of histopathology diagnoses in the Netherlands. Exclusion criteria were gastric type BE or intestinal type BE and a diagnosis of dysplasia prior to or simultaneously with the initial IND diagnosis. Patients with an IND diagnosis undergoing surveillance were followed-up until diagnosis of HGD or EAC, or the last pathology report in the database. Results: In total, 1258 patients met the in- and exclusion criteria of whom 842 (67%) had endoscopic follow-up. Of these, 70% were male, 17.6% were diagnosed in an academic center and in 7% the diagnosis was based on histological revision. Patients who had endoscopic followup were significantly younger than those without. Patients were followed-up for a total of 2585 person-years (mean 3.01, SD 2.6), during which 189 (22%) patients progressed to LGD (138), HGD (21) or EAC (30) (incidence 7.3 (95% CI 6.3-8.4)), while the incidence from IND to HGD/EAC was 2.0 (95% CI 1.5-2.6) per 100 person years. After excluding cases with HGD/EAC within 1 year after IND diagnosis (n=16), the incidence rates were 6.7 (95% CI 5.8-7.8) and 1.4 (95% CI 1.0-1.9), respectively. The majority of EACs were detected within 5 years after the first IND diagnosis. Older age was an independent risk factor for neoplastic progression (HR 1.36, 95%CI 1.2-1.6), while gender, follow-up in an academic setting and histological revision were not. Conclusion: In this large populationbased cohort of patients with BE, the incidence rate of HGD/EAC following a diagnosis of IND was 1.4 per 100 person-years, which corresponds with reported progression rates of LGD in the literature. Our results suggest that the interval of the first surveillance endoscopy in patients with IND should not be different from those of patients with LGD.
Sa1847 Adiponectin and Leptin Receptors Activation in Barrett's Esophagus and Esophageal Adenocarcinoma Patients Could Not Be Explained by Obesity Status Anna Mokrowiecka, Dorota Kacprzak, Joanna Wieczfinska, Rafal Pawliczak, Ewa Malecka-Panas Esophageal adenocarcinoma (EAC) incidence is rapidly increasing which may be due to the growing incidence of Barrett's esophagus (BE) and obesity. Adipokines and their receptors may be important in clarifying the relationship between obesity and the progression of EAC. The aim of the study was to evaluate the expression of adipokines receptors in BE and EAC. Methods: 35 patients with BE, 8 - with EAC, 16 morbidly obese controls (BMI>35) and 11 normal weight controls have been included in the study. Expression of adiponectin receptor 1 protein (AdipoR1) and leptin receptor protein (ObR) in esophageal biopsies from BE, EAC mucosa and controls were assessed with Western-blot analysis and the quantitative RTPCR (qRT-PCR). Results: AdipoR1 protein levels in EAC were significantly higher compared to BE, obese controls and normal weight controls (26.7 vs. 13.3, 11.9, and 8.9 OD units, respectively; p<0.01). Similarly, ObR protein levels were significantly higher in EAC compared to BE (55.2 vs. 31.6 OD units; p<0.01), but not to obese controls and normal weight controls (55.2 vs. 35.9 and 42.9 OD units, respectively; NS). There were no significant differences in AdipoR1 and ObR levels in BE, obese and normal weight controls. Those results were confirmed with qRT-PCR analysis. There were no significant correlations between AdipoR1 and ObR protein levels in BE and patients' weight and BMI. Conclusions: Adiponectin and leptin receptors may play a role in the late phase of EAC development. Their activation in BE and EAC may not be dependent on obesity.
Sa1850 Should We Continue to Follow Patients With <1 CM Barrett's Esophagus (Irregular Z Line): Results From a Large, Multicenter, Cohort Study Srinivas Gaddam, Patrick E. Young, Prashanth Vennalaganti, Neil Gupta, Sachin Wani, April D. Higbee, Sharad C. Mathur, Amit Rastogi, Prashanthi N. Thota, Brooks D. Cash, Fouad J. Moawad, Ajay Bansal, John J. Vargo, Gary W. Falk, David A. Lieberman, Richard E. Sampliner, Prateek Sharma
Sa1848 Aberrant P53 by P53 Immunohistochemistry Combined With DNA FISH, Is Excellent for Risk Stratification of Barrett's Esophagus Patients: Results From a Prospective Long Term Follow-Up Study Akueni Davelaar, Silvia Calpe, Chiu Ting Lau, Sybren L. Meijer, Mike Visser, Paul Fockens, Kausilia K. Krishnadath
Background: Clinicians are often faced with patients having a very short length of columnar mucosa (<1 cm) in the distal esophagus (irregular z lines) that on biopsies have esophageal intestinal metaplasia. The risk of high grade dysplasia (HGD) and EAC in such BE patients is not known. Aim: In a large multicenter cohort of NDBE patients: -To define incidence of HGD and EAC in a well-defined cohort of patients with BE length of <1 cm. Methods: This is a multicenter outcomes project (5 centers) of a large cohort of BE patients. BE was defined by columnar metaplasia in the tubular esophagus on endoscopy and intestinal metaplasia on biopsy. Neoplasia was graded as low-grade dysplasia (LGD), high-grade dysplasia (HGD) and EAC. Duration of follow up was calculated from time of BE diagnosis to the most recent endoscopy with biopsy. Patients with non-dysplastic BE (NDBE) on first EGD and with at least 1 year of subsequent endoscopic follow up were included. Patients with visible lesion, dysplasia and EAC developing within 1-year of BE diagnosis were considered to be prevalent cases and excluded. Progression rates between patients with less than one cm of BE length (esophageal biopsies documenting intestinal metaplasia) were compared to those patients with more than one cm using Fisher's exact test. Selection bias was evaluated by comparing the two groups (BE length <1cm vs. >1cm) using Fisher's exact test and Mann-Whitney U test. Results: Of 3635 BE patients, 1447 (39.8%) patients met the inclusion criteria (93.5% Caucasian, 86.9% men) with mean follow up of 6.14 (SD 4.26) years (8889 patient-years). 69 patients were identified to have BE length of less than one cm. These patients underwent a total of 277 (median = 2 (IQR = 1 - 3)] endoscopies
Objective: Barrett's esophagus (BE) is believed to follow a sequence of low and high grade dysplasia eventually leading to esophageal adenocarcinoma (EAC). Histopathological staging, the current gold standard for predicting BE patient outcome, can at times be unreliable. Several studies showed that p53 abnormalities may serve as biomarkers to identify BE patients at risk. Immunohistochemistry (IHC) for detection of p53 protein accumulation indirectly assesses p53 gene mutations. DNA fluorescent in-situ hybridization (FISH) on brush cytology specimens directly evaluates p53 gene locus loss. However, IHC may miss p53 gene losses, while FISH may not detect gene mutations. Here, we aimed to evaluate if IHC and FISH are complementary tools to assess p53 abnormalities and to prospectively test their prognostic value in a long term follow up study of a BE surveillance cohort. Methods: IHC for p53 was performed on tissue sections of 116 BE patients with different stages of dysplasia or EAC. FISH was performed on simultaneously taken brush cytology specimens. Results of the two techniques were compared and evaluated with respect to histology. For 91 patients the predictive value of p53 abnormalities with respect to progression to high grade dysplasia and/or EAC was tested during a long term prospective follow up.
S-311
AGA Abstracts
AGA Abstracts
Sa1846
AGA Abstracts
during surveillance. Univariate analysis comparing the patients with BE length of <1 cm to those >1 cm showed that there was no selection bias with regards to age, gender, race, BMI, smoking history, PPI use, presence of visible lesion on endoscopy or duration of follow up (all p-values > 0.05). None of the 69 patients developed HGD or EAC over the course of a mean follow up of 6 years (SD 4.4). All of the 60 incident HGD/EAC developed in patients with > 1 cm BE length [annual incidence rate = 0.67 (0.52 - 0.87)]. Conclusions: Results of this large, multi-center cohort study show that none of the patients with a BE length of <1 cm (an irregular z-line) developed HGD or EAC over a mean endoscopic follow up duration of 6 years. If these data are confirmed by other studies, these findings could make a case for discontinuation of surveillance in patients with BE length of less than one cm and for not obtaining biopsies from an irregular z line.
and aneuploidy by flow-cytometry on AFI-targeted biopsies. Statistical analyses were performed using chi-square test. Results: AFI-targeted pCLE correctly classified all the HGD/ EC patients and had a sensitivity and specificity for any grade of dysplasia of 93% and 83%, respectively. Optical biopsies had a similar diagnostic accuracy compared to the Seattle protocol, which had a sensitivity for HGD/IMC and any grade of dysplasia of 83% and 89%, respectively. For the per-location analysis, a total of 155 endoscopic areas were analyzed with pCLE and molecular biomarkers. pCLE had a sensitivity and a specificity for HGD/ IMC and any grade of dysplasia of 100%/64% and 78%/75%, respectively. Overall, 40% of pCLE irregular sequences corresponded to non-dysplastic areas (false positive). We found a statistically significant enrichment (p<0.001) of the three molecular biomarkers in pCLE irregular areas (Figure 1). After exclusion of dysplastic areas, a significant correlation between pCLE irregularity and biomarker positivity was retained (p=0.008). The presence of at least 1 positive biomarker significantly correlated with dysplasia both in pCLE irregular (p=0.01) and pCLE regular areas (p=0.05). Conclusions: AFI-targeted pCLE has a high diagnostic accuracy for dysplasia in BE. Tissue biomarkers are a useful adjunct to characterize the field of molecular abnormality associated with optical dysplasia. These results suggest that the presence of pCLE irregularity, even in the absence of histological dysplasia, relates to molecular changes and may warrant close follow up Figure 1. Correlation between outcome of pCLE imaging and molecular biomarkers. On the x-axis, numerical figures indicate the number of positive biomarkers
Sa1851 Index of Social Deprivation in a Barrett's Esophagus Cohort: The Influence of Affluence? Christine Caygill, Santanu Bhattacharjee, Andre Charlett, A. John Fox, Piers A. Gatenby, Anthony Watsom, Christine Royston, Karna Dev Bardhan INTRODUCTION: Squamous cell carcinoma of the esophagus is known to be more prevalent in those with lower socio-economic status. Little is known in this regard for esophageal adenocarcinoma (EAC), although, anecdotally, EAC has been observed in higher socioeconomic groups. Barrett's esophagus (BE) is the only known precursor of EAC. This study investigates the association between Barrett's esophagus and social deprivation using the 2010 Index of Multiple Deprivation (IMD). METHODS Patients: 1076 BE diagnosed in Rotherham, an industrial town in Northern England between April 1978 and August 2012. Industry: Predominantly steel and coal-based at the start of the study; changing over time with expansion of lighter industries. Population: Approximately 250,000 in 2001, with an ethnic minority population of only 3%. IMD: The Office for National Statistics (ONS) divides England into 32,482 geographical areas, each with a similar population size, technically termed Lower Layer Super Output Areas (LSOA). The 2010 IMD comprising a combination of 38 separate indicators is assigned to each LSOA. As every residential postcode can be assigned to a LSOA and thus an IMD, the Rotherham postcodes at the time of BE diagnosis were extracted from medical records and IMD scores assigned. Quintiles of IMD were derived by dividing the distribution of all 32,482 IMD scores in England into 5 equally sized categories. The 6257 residential postcodes for Rotherham including those of the 1076 BE patients were placed in the quintiles relevant to their IMD score. Analysis: Chi square goodness of fit tests were used to compare the observed (O) distribution of BE to that which would be expected (E) using the Rotherham IMD quintile distribution. Analysis was stratified into 2 groups, those diagnosed before and from 2001 onwards, as 2001 was the median year of diagnosis. RESULTS (See Table) Almost 2/3rds of all Rotherham postcodes fell into the 2 most deprived quintiles. The O/E IMD distribution of the BE cohort diagnosed before 2001 was similar to that of the Rotherham population but since then highly significant differences have emerged (p=0.0001): the two least deprived quintiles now had 170 BE patients against only 124 expected, an increase of 37% over expected. DISCUSSION No single factor is likely to explain the change observed in the last decade, however, the highly significant difference observed points to a strong association between lower deprivation and increasing risk of BE. To the best of our knowledge this is the first report showing a quantitative link between BE and socio-economic status, which may form a basis for the apparent socio-economic shift between squamous esophageal cancer and EAC.
Sa1853 Negative Surveillance Endoscopy Occurs Frequently in Patients With Ultrashort Segment Barrett's Esophagus Vishal Desai, Michael Greenspan, Raja Dhanekula, Sohrab M. Mobarhan, John Losurdo, Shriram M. Jakate, Joshua E. Melson Background & Aims: In patients with known Barrett's esophagus (BE), negative surveillance occurs when intestinal metaplasia (IM) is no longer present on surveillance endoscopy. How frequently a negative surveillance occurs due to missed IM (+-+) versus consistently undetected IM at surveillance (+--) is not well described. The goals of this study were to define the frequency in which negative surveillance occurs in patients with BE and to define how frequently negative surveillance occurs in missed identification versus consistently undetected surveillance. Methods: A cohort (n = 236) of newly diagnosed non-dysplastic BE patients (confirmed on endoscopy and histology) were followed. Outcomes of ultrashort segment BE (USBE= < 1cm segment length) were compared to other short segment (SSBE) patients (1-3 cm segment length). Surveillance patients were classified as "missed IM" (++) when IM was re-identified after not being found on initial surveillance endoscopy versus "consistently undetected IM" (+--) when IM was undetected on two surveillance endoscopies. A Fishers exact test was used to determine the significance between groups. Results: 52 of 236 patients newly diagnosed with BE and followed by surveillance endoscopy had long segment BE (LSBE). None of these cases had negative surveillance endoscopy. Of the remaining 184 patients with SSBE (<3 cm) there were 468 endoscopies of which negative surveillance frequency was 19.67% (92/468). 84 out of the 184 had two surveillance endoscopies of which 40.48% (34/84) had a negative surveillance endoscopy. Among these 84 patients, 54% were male, 76% were Caucasian, and 65% had a hiatal hernia. No statistical difference existed in gender, race, or presence of hiatal hernia in USBE versus SSBE. There were 31 of SSBE (1-3 cm) and 53 of USBE cases. In SSBE, 22.58% (7/31) had a negative surveillance endoscopy versus 50.94% (27/53) of USBE patients. Thus, USBE patients were statistically more likely to have a negative surveillance endoscopy (p=0.012). Among negative surveillance endoscopy cases, 19/34 (55.88%) had consistently undetected IM and 15/34 (44.12%) had missed IM. There was a trend toward USBE patients having consistently undetected IM more often than SSBE patients (28.30%, 15/53 versus 12.90%, 4/34; p= 0.17). Conclusions: It is more common to have negative surveillance in USBE than in other SSBE cases. A negative surveillance may occur in SSBE due to a missed IM or consistently undetected IM on surveillance endoscopy. These two outcomes occurred in roughly equal frequency in SSBE patients. Due to the frequent finding of consistently undetected IM in USBE patients, there should be some pause to upfront ablative therapies in this population when initially diagnosed as IM may no longer be present on follow-up endoscopies without any intervention. Holding off on diagnosis of BE should also be considered.
* Each individual postcode includes ~ 20 households Sa1852 Autofluorescence-Targeted Optical Biopsy Accurately Diagnoses Dysplasia in Barrett's Esophagus and Can Detect the Field of Molecular Change Massimiliano di Pietro, Elizabeth L. Bird-Lieberman, Xinxue Liu, Maria ODonovan, Rebecca Fitzgerald Introduction: Probe-based confocal laser endomicroscopy (pCLE) performed in Barrett's esophagus (BE) via imaging of randomly selected areas is time consuming and similar to the Seattle protocol can miss dysplasia due to sampling error. We used autofluorescence imaging (AFI) to highlight areas for an optical biopsy by pCLE and added molecular biomarkers to the histopathological diagnosis. The aims of this study were to assess the diagnostic accuracy for dysplasia of AFI-targeted optical biopsies and to investigate the correlation between pCLE patterns and field of molecular change in BE. Methods: 46 patients with BE (non-dysplastic BE n=20, indefinite for dysplasia n=4, low grade dysplasia n=10, high grade dysplasia (HGD) or intramucosal cancer (IMC) n=12) were recruited at a single center. Patients underwent high-resolution endoscopy followed by AFI and then pCLE was performed on AFI positive (AFI+) areas. Targeted biopsies were taken from AFI+ areas, followed by random biopsies as per Seattle protocol. pCLE sequences were graded according to published criteria. Cyclin A and p53 expression were assessed by immunohistochemistry
AGA Abstracts
Sa1854 Evaluation of Clinical and Endoscopic Parameters As Predictors for Neoplastic Progression in Barrett's Patients With a Diagnosis of Indefinite for Dysplasia Bela Horvath, Prabhdeep Singh, Hao Xie, Prashanthi N. Thota, XiuLi Liu Barrett's esophagus (BE) carries an increased risk for esophageal adenocarcinoma (EAD) and patients may undergo surveillance biopsies. Biopsies are classified as indefinite for dysplasia (IND), when the epithelial abnormalities are not sufficient to diagnose dysplasia or the
S-312
Global and Gene Specific DNA Methylation Changes in Barrett's Epithelium and Adjacent Normal Mucosa Samuel C. Stevens, Julian A. Abrams, Clare LeFave, Charles J. Lightdale, Tamas A. Gonda Introduction. DNA methylation changes make a major contribution to cancer initiation and progression and can precede genetic alterations. This feature makes such epigenetic alterations attractive markers and targets of intervention. DNA methylation changes may also be observed in cancer associated normal epithelium. Although DNA methylation changes are known to be associated with BE it is less well understood what changes occur in adjacent normal epithelium and at what stage do most epigenetic alterations develop. Our goal was to compare the grade-matched DNA methylation signature of BE, adjacent normal cardia and squamous mucosa and identify (1) the grade of BE when DNA methylation changes occur; (2) the epigenetic similarities between adjacent non-neoplastic epithelium; (3) gene specific methylation changes associated with early neoplastic transformation. Methods. Biopsies were collected from patients with all grades of BE from every 2 cm of BE mucosa, from the squamous epithelium (2 cm above the Z line) and the endoscopic cardia. Global methylation status was assessed by LINE-1 pyrosequencing and the methyl acceptor assay. Gene specific methylation was assessed using the Fluidigm Next Generation Sequencing. 42 Genes were selected from published and validated list of differentially methylated genes in BE/EAC for analysis. Results. 254 biopsies from 56 patients were analyzed. 55% had no dysplasia or low grade dysplasia (ND, LGD) and 45% had high grade dysplasia (HGD) or carcinoma. There was a significant difference between Line-1 and gene specific methylation of cardia, squamous versus BE. A significant decrease in Line-1 methylation was noted between BE with ND/LGD versus HGD or intramucosal carcinoma. A similar trend was noted for the majority of genes that showed change in methylation associated with progression of dysplasia. There was no significant change in Line-1 or gene specific methylation of the squamous epithelium. However, a significant change in Line-1 methylation was noted in the cardia in LGD vs HGD. Of the 15 genes that showed hyper or hypomethylation in BE 12/15 showed similar changes in the cardia while only 3/15 showed similar changes in the squamous epithelium. Conclusions. Both global and gene specific methylation changes are associated with progression of BE and it is between LGD and HGD that most of these changes are detected. Although the DNA methylation status of both adjacent normal cardia and squamous epithelium are different from Barrett's mucosa, with progression methylation changes occur in the cardia and not in squamous epithelium. These results support targeting epigenetic interventions or chemoprevention in patients with LGD and identify an epitgenetic similarity between the cardia and Barrett's epithelium but not the adjacent squamous mucosa.
Sa1849 The Risk of Malignant Progression in Patients With Barrett's Esophagus and a Histologic Diagnosis of Indefinite for Dysplasia: A Dutch Nationwide Cohort Study Christine Kestens, Max Leenders, Johan Offerhaus, Jantine W. van Baal, Peter D. Siersema Introduction and aim: Histological assessment of Barrett's esophagus (BE) is the main diagnostic tool for risk stratification of the progression to esophageal adenocarcinoma (EAC). A diagnosis indefinite for dysplasia (IND) in BE is used when a diagnosis of genuine dysplasia is equivocal. Due to this uncertainty combined with the unknown risk of progression of IND to low-grade dysplasia (LGD), high-grade dysplasia (HGD) or EAC, follow-up includes repeat endoscopy with biopsies. We aimed to assess the risk of progression to high-grade dysplasia (HGD) or esophageal adenocarcinoma (EAC) after a diagnosis of IND in a nationwide cohort of patients with BE in the Netherlands. Methods: Patients with a first diagnosis of IND in BE were selected between 2002-2011 using PALGA, a nationwide registry of histopathology diagnoses in the Netherlands. Exclusion criteria were gastric type BE or intestinal type BE and a diagnosis of dysplasia prior to or simultaneously with the initial IND diagnosis. Patients with an IND diagnosis undergoing surveillance were followed-up until diagnosis of HGD or EAC, or the last pathology report in the database. Results: In total, 1258 patients met the in- and exclusion criteria of whom 842 (67%) had endoscopic follow-up. Of these, 70% were male, 17.6% were diagnosed in an academic center and in 7% the diagnosis was based on histological revision. Patients who had endoscopic followup were significantly younger than those without. Patients were followed-up for a total of 2585 person-years (mean 3.01, SD 2.6), during which 189 (22%) patients progressed to LGD (138), HGD (21) or EAC (30) (incidence 7.3 (95% CI 6.3-8.4)), while the incidence from IND to HGD/EAC was 2.0 (95% CI 1.5-2.6) per 100 person years. After excluding cases with HGD/EAC within 1 year after IND diagnosis (n=16), the incidence rates were 6.7 (95% CI 5.8-7.8) and 1.4 (95% CI 1.0-1.9), respectively. The majority of EACs were detected within 5 years after the first IND diagnosis. Older age was an independent risk factor for neoplastic progression (HR 1.36, 95%CI 1.2-1.6), while gender, follow-up in an academic setting and histological revision were not. Conclusion: In this large populationbased cohort of patients with BE, the incidence rate of HGD/EAC following a diagnosis of IND was 1.4 per 100 person-years, which corresponds with reported progression rates of LGD in the literature. Our results suggest that the interval of the first surveillance endoscopy in patients with IND should not be different from those of patients with LGD.
Sa1847 Adiponectin and Leptin Receptors Activation in Barrett's Esophagus and Esophageal Adenocarcinoma Patients Could Not Be Explained by Obesity Status Anna Mokrowiecka, Dorota Kacprzak, Joanna Wieczfinska, Rafal Pawliczak, Ewa Malecka-Panas Esophageal adenocarcinoma (EAC) incidence is rapidly increasing which may be due to the growing incidence of Barrett's esophagus (BE) and obesity. Adipokines and their receptors may be important in clarifying the relationship between obesity and the progression of EAC. The aim of the study was to evaluate the expression of adipokines receptors in BE and EAC. Methods: 35 patients with BE, 8 - with EAC, 16 morbidly obese controls (BMI>35) and 11 normal weight controls have been included in the study. Expression of adiponectin receptor 1 protein (AdipoR1) and leptin receptor protein (ObR) in esophageal biopsies from BE, EAC mucosa and controls were assessed with Western-blot analysis and the quantitative RTPCR (qRT-PCR). Results: AdipoR1 protein levels in EAC were significantly higher compared to BE, obese controls and normal weight controls (26.7 vs. 13.3, 11.9, and 8.9 OD units, respectively; p<0.01). Similarly, ObR protein levels were significantly higher in EAC compared to BE (55.2 vs. 31.6 OD units; p<0.01), but not to obese controls and normal weight controls (55.2 vs. 35.9 and 42.9 OD units, respectively; NS). There were no significant differences in AdipoR1 and ObR levels in BE, obese and normal weight controls. Those results were confirmed with qRT-PCR analysis. There were no significant correlations between AdipoR1 and ObR protein levels in BE and patients' weight and BMI. Conclusions: Adiponectin and leptin receptors may play a role in the late phase of EAC development. Their activation in BE and EAC may not be dependent on obesity.
Sa1850 Should We Continue to Follow Patients With <1 CM Barrett's Esophagus (Irregular Z Line): Results From a Large, Multicenter, Cohort Study Srinivas Gaddam, Patrick E. Young, Prashanth Vennalaganti, Neil Gupta, Sachin Wani, April D. Higbee, Sharad C. Mathur, Amit Rastogi, Prashanthi N. Thota, Brooks D. Cash, Fouad J. Moawad, Ajay Bansal, John J. Vargo, Gary W. Falk, David A. Lieberman, Richard E. Sampliner, Prateek Sharma
Sa1848 Aberrant P53 by P53 Immunohistochemistry Combined With DNA FISH, Is Excellent for Risk Stratification of Barrett's Esophagus Patients: Results From a Prospective Long Term Follow-Up Study Akueni Davelaar, Silvia Calpe, Chiu Ting Lau, Sybren L. Meijer, Mike Visser, Paul Fockens, Kausilia K. Krishnadath
Background: Clinicians are often faced with patients having a very short length of columnar mucosa (<1 cm) in the distal esophagus (irregular z lines) that on biopsies have esophageal intestinal metaplasia. The risk of high grade dysplasia (HGD) and EAC in such BE patients is not known. Aim: In a large multicenter cohort of NDBE patients: -To define incidence of HGD and EAC in a well-defined cohort of patients with BE length of <1 cm. Methods: This is a multicenter outcomes project (5 centers) of a large cohort of BE patients. BE was defined by columnar metaplasia in the tubular esophagus on endoscopy and intestinal metaplasia on biopsy. Neoplasia was graded as low-grade dysplasia (LGD), high-grade dysplasia (HGD) and EAC. Duration of follow up was calculated from time of BE diagnosis to the most recent endoscopy with biopsy. Patients with non-dysplastic BE (NDBE) on first EGD and with at least 1 year of subsequent endoscopic follow up were included. Patients with visible lesion, dysplasia and EAC developing within 1-year of BE diagnosis were considered to be prevalent cases and excluded. Progression rates between patients with less than one cm of BE length (esophageal biopsies documenting intestinal metaplasia) were compared to those patients with more than one cm using Fisher's exact test. Selection bias was evaluated by comparing the two groups (BE length <1cm vs. >1cm) using Fisher's exact test and Mann-Whitney U test. Results: Of 3635 BE patients, 1447 (39.8%) patients met the inclusion criteria (93.5% Caucasian, 86.9% men) with mean follow up of 6.14 (SD 4.26) years (8889 patient-years). 69 patients were identified to have BE length of less than one cm. These patients underwent a total of 277 (median = 2 (IQR = 1 - 3)] endoscopies
Objective: Barrett's esophagus (BE) is believed to follow a sequence of low and high grade dysplasia eventually leading to esophageal adenocarcinoma (EAC). Histopathological staging, the current gold standard for predicting BE patient outcome, can at times be unreliable. Several studies showed that p53 abnormalities may serve as biomarkers to identify BE patients at risk. Immunohistochemistry (IHC) for detection of p53 protein accumulation indirectly assesses p53 gene mutations. DNA fluorescent in-situ hybridization (FISH) on brush cytology specimens directly evaluates p53 gene locus loss. However, IHC may miss p53 gene losses, while FISH may not detect gene mutations. Here, we aimed to evaluate if IHC and FISH are complementary tools to assess p53 abnormalities and to prospectively test their prognostic value in a long term follow up study of a BE surveillance cohort. Methods: IHC for p53 was performed on tissue sections of 116 BE patients with different stages of dysplasia or EAC. FISH was performed on simultaneously taken brush cytology specimens. Results of the two techniques were compared and evaluated with respect to histology. For 91 patients the predictive value of p53 abnormalities with respect to progression to high grade dysplasia and/or EAC was tested during a long term prospective follow up.
S-311
AGA Abstracts
AGA Abstracts
Sa1846
AGA Abstracts
during surveillance. Univariate analysis comparing the patients with BE length of <1 cm to those >1 cm showed that there was no selection bias with regards to age, gender, race, BMI, smoking history, PPI use, presence of visible lesion on endoscopy or duration of follow up (all p-values > 0.05). None of the 69 patients developed HGD or EAC over the course of a mean follow up of 6 years (SD 4.4). All of the 60 incident HGD/EAC developed in patients with > 1 cm BE length [annual incidence rate = 0.67 (0.52 - 0.87)]. Conclusions: Results of this large, multi-center cohort study show that none of the patients with a BE length of <1 cm (an irregular z-line) developed HGD or EAC over a mean endoscopic follow up duration of 6 years. If these data are confirmed by other studies, these findings could make a case for discontinuation of surveillance in patients with BE length of less than one cm and for not obtaining biopsies from an irregular z line.
and aneuploidy by flow-cytometry on AFI-targeted biopsies. Statistical analyses were performed using chi-square test. Results: AFI-targeted pCLE correctly classified all the HGD/ EC patients and had a sensitivity and specificity for any grade of dysplasia of 93% and 83%, respectively. Optical biopsies had a similar diagnostic accuracy compared to the Seattle protocol, which had a sensitivity for HGD/IMC and any grade of dysplasia of 83% and 89%, respectively. For the per-location analysis, a total of 155 endoscopic areas were analyzed with pCLE and molecular biomarkers. pCLE had a sensitivity and a specificity for HGD/ IMC and any grade of dysplasia of 100%/64% and 78%/75%, respectively. Overall, 40% of pCLE irregular sequences corresponded to non-dysplastic areas (false positive). We found a statistically significant enrichment (p<0.001) of the three molecular biomarkers in pCLE irregular areas (Figure 1). After exclusion of dysplastic areas, a significant correlation between pCLE irregularity and biomarker positivity was retained (p=0.008). The presence of at least 1 positive biomarker significantly correlated with dysplasia both in pCLE irregular (p=0.01) and pCLE regular areas (p=0.05). Conclusions: AFI-targeted pCLE has a high diagnostic accuracy for dysplasia in BE. Tissue biomarkers are a useful adjunct to characterize the field of molecular abnormality associated with optical dysplasia. These results suggest that the presence of pCLE irregularity, even in the absence of histological dysplasia, relates to molecular changes and may warrant close follow up Figure 1. Correlation between outcome of pCLE imaging and molecular biomarkers. On the x-axis, numerical figures indicate the number of positive biomarkers
Sa1851 Index of Social Deprivation in a Barrett's Esophagus Cohort: The Influence of Affluence? Christine Caygill, Santanu Bhattacharjee, Andre Charlett, A. John Fox, Piers A. Gatenby, Anthony Watsom, Christine Royston, Karna Dev Bardhan INTRODUCTION: Squamous cell carcinoma of the esophagus is known to be more prevalent in those with lower socio-economic status. Little is known in this regard for esophageal adenocarcinoma (EAC), although, anecdotally, EAC has been observed in higher socioeconomic groups. Barrett's esophagus (BE) is the only known precursor of EAC. This study investigates the association between Barrett's esophagus and social deprivation using the 2010 Index of Multiple Deprivation (IMD). METHODS Patients: 1076 BE diagnosed in Rotherham, an industrial town in Northern England between April 1978 and August 2012. Industry: Predominantly steel and coal-based at the start of the study; changing over time with expansion of lighter industries. Population: Approximately 250,000 in 2001, with an ethnic minority population of only 3%. IMD: The Office for National Statistics (ONS) divides England into 32,482 geographical areas, each with a similar population size, technically termed Lower Layer Super Output Areas (LSOA). The 2010 IMD comprising a combination of 38 separate indicators is assigned to each LSOA. As every residential postcode can be assigned to a LSOA and thus an IMD, the Rotherham postcodes at the time of BE diagnosis were extracted from medical records and IMD scores assigned. Quintiles of IMD were derived by dividing the distribution of all 32,482 IMD scores in England into 5 equally sized categories. The 6257 residential postcodes for Rotherham including those of the 1076 BE patients were placed in the quintiles relevant to their IMD score. Analysis: Chi square goodness of fit tests were used to compare the observed (O) distribution of BE to that which would be expected (E) using the Rotherham IMD quintile distribution. Analysis was stratified into 2 groups, those diagnosed before and from 2001 onwards, as 2001 was the median year of diagnosis. RESULTS (See Table) Almost 2/3rds of all Rotherham postcodes fell into the 2 most deprived quintiles. The O/E IMD distribution of the BE cohort diagnosed before 2001 was similar to that of the Rotherham population but since then highly significant differences have emerged (p=0.0001): the two least deprived quintiles now had 170 BE patients against only 124 expected, an increase of 37% over expected. DISCUSSION No single factor is likely to explain the change observed in the last decade, however, the highly significant difference observed points to a strong association between lower deprivation and increasing risk of BE. To the best of our knowledge this is the first report showing a quantitative link between BE and socio-economic status, which may form a basis for the apparent socio-economic shift between squamous esophageal cancer and EAC.
Sa1853 Negative Surveillance Endoscopy Occurs Frequently in Patients With Ultrashort Segment Barrett's Esophagus Vishal Desai, Michael Greenspan, Raja Dhanekula, Sohrab M. Mobarhan, John Losurdo, Shriram M. Jakate, Joshua E. Melson Background & Aims: In patients with known Barrett's esophagus (BE), negative surveillance occurs when intestinal metaplasia (IM) is no longer present on surveillance endoscopy. How frequently a negative surveillance occurs due to missed IM (+-+) versus consistently undetected IM at surveillance (+--) is not well described. The goals of this study were to define the frequency in which negative surveillance occurs in patients with BE and to define how frequently negative surveillance occurs in missed identification versus consistently undetected surveillance. Methods: A cohort (n = 236) of newly diagnosed non-dysplastic BE patients (confirmed on endoscopy and histology) were followed. Outcomes of ultrashort segment BE (USBE= < 1cm segment length) were compared to other short segment (SSBE) patients (1-3 cm segment length). Surveillance patients were classified as "missed IM" (++) when IM was re-identified after not being found on initial surveillance endoscopy versus "consistently undetected IM" (+--) when IM was undetected on two surveillance endoscopies. A Fishers exact test was used to determine the significance between groups. Results: 52 of 236 patients newly diagnosed with BE and followed by surveillance endoscopy had long segment BE (LSBE). None of these cases had negative surveillance endoscopy. Of the remaining 184 patients with SSBE (<3 cm) there were 468 endoscopies of which negative surveillance frequency was 19.67% (92/468). 84 out of the 184 had two surveillance endoscopies of which 40.48% (34/84) had a negative surveillance endoscopy. Among these 84 patients, 54% were male, 76% were Caucasian, and 65% had a hiatal hernia. No statistical difference existed in gender, race, or presence of hiatal hernia in USBE versus SSBE. There were 31 of SSBE (1-3 cm) and 53 of USBE cases. In SSBE, 22.58% (7/31) had a negative surveillance endoscopy versus 50.94% (27/53) of USBE patients. Thus, USBE patients were statistically more likely to have a negative surveillance endoscopy (p=0.012). Among negative surveillance endoscopy cases, 19/34 (55.88%) had consistently undetected IM and 15/34 (44.12%) had missed IM. There was a trend toward USBE patients having consistently undetected IM more often than SSBE patients (28.30%, 15/53 versus 12.90%, 4/34; p= 0.17). Conclusions: It is more common to have negative surveillance in USBE than in other SSBE cases. A negative surveillance may occur in SSBE due to a missed IM or consistently undetected IM on surveillance endoscopy. These two outcomes occurred in roughly equal frequency in SSBE patients. Due to the frequent finding of consistently undetected IM in USBE patients, there should be some pause to upfront ablative therapies in this population when initially diagnosed as IM may no longer be present on follow-up endoscopies without any intervention. Holding off on diagnosis of BE should also be considered.
* Each individual postcode includes ~ 20 households Sa1852 Autofluorescence-Targeted Optical Biopsy Accurately Diagnoses Dysplasia in Barrett's Esophagus and Can Detect the Field of Molecular Change Massimiliano di Pietro, Elizabeth L. Bird-Lieberman, Xinxue Liu, Maria ODonovan, Rebecca Fitzgerald Introduction: Probe-based confocal laser endomicroscopy (pCLE) performed in Barrett's esophagus (BE) via imaging of randomly selected areas is time consuming and similar to the Seattle protocol can miss dysplasia due to sampling error. We used autofluorescence imaging (AFI) to highlight areas for an optical biopsy by pCLE and added molecular biomarkers to the histopathological diagnosis. The aims of this study were to assess the diagnostic accuracy for dysplasia of AFI-targeted optical biopsies and to investigate the correlation between pCLE patterns and field of molecular change in BE. Methods: 46 patients with BE (non-dysplastic BE n=20, indefinite for dysplasia n=4, low grade dysplasia n=10, high grade dysplasia (HGD) or intramucosal cancer (IMC) n=12) were recruited at a single center. Patients underwent high-resolution endoscopy followed by AFI and then pCLE was performed on AFI positive (AFI+) areas. Targeted biopsies were taken from AFI+ areas, followed by random biopsies as per Seattle protocol. pCLE sequences were graded according to published criteria. Cyclin A and p53 expression were assessed by immunohistochemistry
AGA Abstracts
Sa1854 Evaluation of Clinical and Endoscopic Parameters As Predictors for Neoplastic Progression in Barrett's Patients With a Diagnosis of Indefinite for Dysplasia Bela Horvath, Prabhdeep Singh, Hao Xie, Prashanthi N. Thota, XiuLi Liu Barrett's esophagus (BE) carries an increased risk for esophageal adenocarcinoma (EAD) and patients may undergo surveillance biopsies. Biopsies are classified as indefinite for dysplasia (IND), when the epithelial abnormalities are not sufficient to diagnose dysplasia or the
S-312