Search

Introduction We have previously shown how the Allred scoring system may be used to semi-quantify expression of nuclear biomarkers in the Barrett’s (BE) to oesophageal adenocarcinoma (OA) sequence. Recently, a number of digital image analysis (DIA) platforms have been clinically validated for quantification of immunohistochemistry (IHC) in breast tissue. This study aims to compare pathologists scores (PS) with DIA for the quantification of nuclear biomarkers in BE to OA sequence. Methods Paraffin embedded specimens were selected from 34 patients. Pathology grade (PG) was scored as 1 (non-dysplastic BE; n = 5), 2 (low grade dysplasia; n = 5), 3 (high grade dysplasia; n = 11) and 4 (OA; n = 14). Sections were immunostained with antibodies PLK1-M, PLK1-L and Geminin. Intensity (I-PS) (0 to 3+) and extent (E-PS) (0; 66% = 5) of staining were scored by 2 GI pathologists, and mean PS calculated. Intensity and proportions of +ve staining were digitally quantified using Ariol® software. Analysis classifiers were trained to identify thresholds of positive (brown/DAB) and negative (blue/Hx) nuclei (Figure A) in the areas of interest. Background tissue was digitally excluded. Mean intensity (I-DIA) and mean counts for DIA positive (red) and negative (green) nuclei (Figure B) were quantified to calculate percentage positivity (E-DIA) and compared with I-PS, E-PS and total Allred score (A-PS) using Pearson correlation coefficient. Results Significant correlation was seen between E-DIA and A-PS (r = 0.76, p = 0.006; r = 0.73, p = 0.008; r = 0.94, p = 0.0004) with all biomarkers (PLK1-M; PLK1-L; Geminin). PLK1-L showed additional correlation between DIA and PS for intensity (r = 0.985, p = 0.02) and extent (r = 0.95, p = 0.90). Geminin showed additional correlations between DIA and PS for extent (r = 0.99, p = 0.0008) and PG (r = 0.97, p = 0.03). Following training, Ariol® analysis took a mean of 4mins (Range 3–5) per tissue region highlighted. Conclusion This study has demonstrated how DIA may be used to quantify expression of nuclear biomarkers. Significant correlation between DIA with Allred score was seen with all biomarkers, but only PLK1-L correlated with intensity and Geminin with PG. Background staining ignored by pathologists was found to be a confounder for DIA, particularly with PLK1-M. Nonetheless, DIA has great potential to enhance current grading and risk stratification systems for BE, and help select patients for targeted therapies dependent on biomarker expression. Disclosure of Interest None Declared

Introduction Deficiency in the enzyme cytochrome C oxidase (CCO) has proven to be a versatile marker of clonal population in human tissues. CCO is encoded entirely by the mitochondrial DNA (mtDNA) and deficiency in CCO expression is usually attributable to mutation in the mtDNA. CCO-deficient cells are easily detectable by two-colour enzyme histochemistry. This staining provides means to identify clonal patches of CCO-deficient cells. Subsequent sequencing of mtDNA from individual cells within the patch and revealing the same somatic mutation in each cell confirms the clonality of a patch. In the human intestine, crypts are composed of a population of a contiguous CCO-deficient mtDNA-mutated cells, and also non-mutant CCO-proficient cells are frequently observed: from the small clones occupying only a few cell positions on the crypt circumference (partially-mutated crypts), to crypts composed only CCO-deficient cells (wholly-mutated crypts). Patches of adjacent and clonal CCO-deficient crypts are also observed. Larger patches are more common in older patients indicating that CCO-deficient crypts continue to divide in the ageing colon. Methods To study stem cells dynamics within intestinal crypts, we aim to characterise the shape and the size of clones in partially CCO-deficient crypts, by combining the two-colour enzyme histochemical staining with image analysis and computer reconstruction. Multiple serial sections in the transverse plain were taken through frozen samples. Digital images were then taken of each serial section and used to create a ‘crypt map’ using in-house analytical software. A crypt map is a representation of the whole 3D tubular crypt unfurled and laid flat with colour enhancement post-processing. Results Our results in normal and diseased human colon show that clone size can be approximated by the percentage of the crypt circumference measured from crypt transverse sections and occupied by a CCO-deficient clone. Conclusion We envisage that analysis of such clonal distributions in the context of a branching process model could be used to determine the patterns of stem cell division within the human colon. Disclosure of Interest None Declared.

Introduction The origin and development of Barrett’s oesophagus has long been discussed, with three main proposals for the origin of metaplasia: from oesophageal squamous epithelium, from upward migration of cardiac glands, or from submucosal glands. Methods In Barrett’s glands we have studied the distribution of proliferative activity using Ki67 and the migration of cells in Barrett’s glands from patients infused with iododeoxyuridine 7 and 11 days pre-oesophagectomy. We have localised gene expression of mucins using immunohistochemistry (IHC) and trefoil family factor (TFF) peptides using IHC and in situ hybridization (ISH) and the stem cell marker Lgr5 using ISH, combining this with analysis of the clonal architecture of Barrett’s glands using mitochondrial DNA (mtDNA) as a clonal marker. Results In Barrett’s glands proliferation is seen mainly in the middle part of the gland and diminishes towards the surface and the base of the gland. Cells migrate in a bidirectional manner. MUC5AC and TFF1 expression are found superficially, while MUC6 and TFF2 are found at the bases of the glands, similar to the distribution seen in antral gastric glands: MUC2, staining goblet cells and columnar cells, is concentrated superficially. Lgr5 mRNA is also found in the middle part of the glands, indicating the location of the stem cell niche. Barrett’s glands are clonal, indicating derivation from a single cell, and suggesting that Barrett’s stem cells have dual differentiation capacity. Gastric intestinal metaplasia, on the other hand, shows basal Lgr5 mRNA localisation with a distribution of proliferative activity similar to intestinal crypts. Conclusion We conclude that Barrett’s glands show the proliferative and stem cell architecture, and preserve patterns of gene expression of pyloric-type gastric glands, but are maintained by unique stem cells with both gastric and intestinal differentiation capacity: we propose that Barrett’s glands originate as gastric glands and that subsequent intestinal differentiation advances with time, strongly supporting an origin from the proximal cardiac mucosa. Disclosure of Interest None Declared

Introduction Population based studies have shown a small but significant proportion of patients with Barrett’s oesophagus (BE) progress to oesophageal adenocarcinoma (OA). Predicting which patients progress is of vital importance for risk stratification into increased surveillance regimes or early therapeutic intervention. We have previously shown the replication licencing factors (RLF) such as polo-like kinase-1 (PLK1) may act as surrogate markers of DNA ploidy abnormalities (DNA-PA) (1) and subsequent risk of progression to OAC. This study aimed to assess whether the RLF’s are up-regulated in the metaplasia-dysplasia-carcinoma sequence independently of DNA-PA. Methods Paraffin embedded oesophageal specimens were selected from 94 patients, and pathology grade (PG) scored as 1 (squamous = Sq; n = 4), 2 (non-dysplastic BE = NDBE; n = 19), 3 (low grade dysplasia = LGD; n = 20), 4 (carcinoma in-situ = HGD/IMC; n = 36) and 5 (invasive OA = IOA; n = 15). Sections were immunostained with the antibodies PLK1-M, PLK1-L and Gem using the automated BOND-MAX system (Leica). Intensity (I) (0 to 3+) and extent (E) (0; 66% = 5) of staining were scored by 2 expert GI pathologists, and mean scores calculated. DNA-PA was assessed using image cytometry (IC). Pearson r coefficient and two-tailed P values were calculated for correlations and significance. Results A significant correlation with extent (r = 0.87, p = 0.05; r = 0.91, p = 0.03; r = 0.99, p = 0.0008) was seen with all RLFs (PLK1-M; PLK1-L; Gem). Only PLK1-M correlated with intensity (r = 0.92, p = 0.03) and total (I+E) scores (r = 0.91, p = 0.03). Aneuploidy was present in NDBE (6%), LGD (25%), HGD/IMC (61%) and IOA (67%). The proportion of aneuploid cases/group correlated significantly with PG (r = 0.97, p = 0.007) and mean I, E and total scores for PLK1-M (r = 0.97, p = 0.008; r = 0.96, p = 0.008; r = 0.98, p = 0.96). Conclusion All assessed RLFs were found to be significantly up-regulated in the progression to OA, with geminin scoring highest when evaluated by extent scores. Overall PLK1-M appeared the best antibody, showing additional correlation with its intensity and total staining scores with PG and DNA-PA. We further demonstrated that aneuploidy still correlates with PG as we have previously reported. In future, PLK1-M and aneuploidy may provide the basis of a biomarker panel to help guide screening and therapy for those predicted to be at increased risk of progression to OA. Disclosure of Interest None Declared Reference Butt MA et al. Gut 2012; 61:Suppl 2 A302

Introduction Dysplasia in epithelia is an important histological diagnosis although the specific genetic changes which are responsible for this phenotypic change are unknown. Recent reports indicate that the dysplastic phenotype may not be immutable: in basal crypt dysplasia like atypia (BCDA), unequivocal dysplasia is seen in the crypts in Barrett›s oesophagus and other pre-invasive lesions in the gastrointestinal tract, but the upper crypts and surface epithelium associated with these dysplastic crypts show a differentiated epithelium. The genotypic relationship between the BCDA and the differentiated surface epithelium is unclear. Methods We obtained 17 examples of BCDA: the lower crypts and upper crypts and surface epithelium were differentially lazer-microdissected from formalin-fixed, paraffin embedded sections and mutations were sought in tumour suppressor genes frequently associated with progression in Barrett’s oesophagus. Results Two patients showed a c.C238T mutation in the p16 (CDKN2A, p16Ink4A) gene and where the precise microanatomical relationships could be discerned: this clonal p16 mutation was present in both the BCDA at the crypt base and in the upper crypt and surface epithelium. This shows that the surface epithelium is derived from the dysplastic crypt epithelium: the dysplastic phenotype is therefore not fixed and can be reversed. Two patients showed a c.C238T mutation in the p16 (CDKN2A, p16Ink4A) gene and where the precise microanatomical relationships could be discerned: this clonal p16 mutation was present in both the BCDA at the crypt base and in the upper crypt and surface epithelium. This shows that the surface epithelium is derived from the dysplastic crypt epithelium: the dysplastic phenotype is therefore not fixed and can be reversed. Conclusion The mechanism of this change is unclear: dysplastic cells may, probably at an earlystage in their progression, respond to differentiation signals. We are some way from a definition of the genotypic correlates of thedysplastic phenotype, and from an understanding of its plasticity. Disclosure of Interest None Declared.

Introduction Evaluation of HER2 status is standard practise for people with inoperable locally advanced or recurrent and/or metastatic adenocarcinoma of the stomach or gastroesophageal junction (GOJ). HER2 positivity is defined as a histology score of 3+, or 2+ with HER2 gene amplification. This study analysed gastroesophageal HER2 testing in UCL-AD over the past 2.5 years. Methods All referrals for HER2 testing (n = 844) were reviewed. Data was captured for age (n = 367), sex (n = 320), tumour origin (n = 104), referring lab (n = 768), sample adequacy, HER2 grading (0–3+) and tumour type. Sections were immunostained with the Roche Tissue Diagnostics Pathway HER2 assay. For cases scoring 2+, F/DDISH testing was performed and the turn-around-time (TAT) from receipt to result analysed. Results Cases were received from 61 labs, mean age 63yo (26–93yo, SD = 13) and the majority were men (71.5%). Specimens originated from oesophageal (OE) (33.7%), GOJ (13.5%), stomach (48%) and UGI metastases (4.8%). The majority (98.7%, n = 833) were HER2-graded. The tumour types were intestinal (INT) (49.6%, n = 257), diffuse (D) (29%, n = 150), mixed D/INT (14.3%, n = 74), carcinoma-in-situ (HGD/IMC) (6.4%, n = 33), squamous (Sq) and adenosquamous (ASq) (0.4%, n = 2 each). HER2 scores were 0 (44%, n = 366), 1+ (14.5%, n = 121), 2+ (18.8%, n = 156) and 3+ (22.7%, n = 189).Gene amplification of 2+ cases identified 124 (79.5%) negative, 23 (14.7%) positive and 9 (1.1%) unrecordable results. HER2 positivity (3+/2+&F/DDISH+) was noted in 10% (15/150) of D, 30.4% (78/257) of INT, 13.5% (10/74) of mixed D/INT, 54.5% (18/33) of HGD/IMC and in 50% (1/2) of Sq tumours. A TAT of Conclusion Overall HER2-positivity rate was 25.8% in a cohort twice the size of the UK cohort reported in the benchmark ToGA trial. When compared with ToGA, subgroup analysis showed comparable but higher rates in D cancers (10% vs 6.1%), lower in both INT/mixed cancers (30.4/13.5% vs 32.2/20.4%), showed opposite proportions of GOJ/stomach cancers (28.6/31.3% vs 33.1/20.9%), and grouping proximal (OE+GOJ) cancers together, a higher ratio compared to stomach cancers was seen (0.51 vs 0.33). Positivity in Sq cancer is of uncertain significance and warrants further research into new potential applications for HER2 targeting. Disclosure of Interest None Declared

Introduction In Barrett’s oesophagus (BE), Barrett’s dysplasia and esophageal carcinoma (EC) there can be genomic instability. This instability leads to DNA abnormalities, typically aneuploidy, where cells accumulate abnormal amounts of DNA. Tetraploidy is a specific term given to cells with twice the normal (diploid) content of DNA. In our specialist referral centre for BE, ploidy assessment is requested for cases with persistent or recurrent low-grade/indefinite for dysplasia (LGD/IFD), clinical suspicion of progression due for example to a strong family history of EC in long segment BE, or cases suspected to relapse after prior treatment for high-grade dysplasia/intramucosal cancer (HGD/IMC). This study aimed to assess correlation between dysplasia and DNA ploidy abnormalities (DNA-PA) in clinical cases with BE referred for ploidy assessment to the UCLH specialist image cytometry (IC) service. Methods All clinical referrals for IC were retrospectively identified (n = 189) and reports from analysed blocks captured (n = 682). All samples were processed by a specialist IC technician to produce ploidy results with an automated image cytometric analyser. Ploidy histograms were then reviewed to confirm the automated IC result as diploid, aneuploid, tetraploid or equivocal. Histology was then reported or reviewed by an expert GI pathologist and scored as 1 (BE), 2 (IFD), 3 (LGD) or 4 (HGD). Results Case referrals were received from 4 institutions, mean age was 58yo (24–85yo, SD 13) and the majority were men (79.9%). Corresponding histology was available in 69.7% (n = 475) of samples processed. DNA-PA positivity rates (Aneuploidy/Tetraploidy) in the sub-group with histology were 4% in BE (n = 376), 6.9% in IFD (n = 29), 29% in LGD (n = 31) and 51.3% in HGD (n = 39). Pearson product-moment correlation coefficient showed significant correlation between increasing degree of dysplasia and DNA-PA (Pearson r = 0.96, p = 0.039). Conclusion This study has demonstrated that in our large case series DNA-PA’s correlate with increasing degree of dysplasia. DNA-PA is a validated biomarker for cancer progression and has been used in our cohort to guide frequency of surveillance and in some LGD cases minimally invasive endoscopic therapy, where they otherwise would not have received it. Further longitudinal studies on the progression to cancer in our case series will provide further insight into this important biomarker. Disclosure of Interest None Declared

Introduction Abnormalities of the colonic mucosa have been implicated in the pathogenesis of ulcerative colitis (UC). We investigated mRNA profiles of macroscopically non-inflamed mucosal biopsies from the colon in patients with UC, Crohn’s disease (CD) and control subjects without gastrointestinal disease (HC), to identify genes that might be involved in the aetiology of the disease. Methods Paired biopsies were taken for histology and mRNA extraction from macroscopically non-inflamed mucosa in the ascending and descending colon, and the rectum, from 24 patients with UC, 14 with CD and 27 HCs undergoing routine colonoscopy. Patients were in complete clinical remission and were either on no treatment or on 5-aminosalicylates ± azathioprine. cRNA was hybridised to Illumina HumanHT-12 v4 Expression Beadchips. Array expression data were log transformed and normalised. Only probes with a detection p-value Results In group comparisons, of the 26,261 expressed probes, Family with Sequence Similarity 5, member C ( FAM5C ) was the only gene to be significantly under-expressed in UC, both in the rectum (FC = –1.58, p = 0.0008) and the descending colon (FC = –1.64, p = 0.0011). Outlier analysis showed that FAM5C was also grossly under-expressed in the ascending colon in 37.5% of UC patients, demonstrating that its expression is abnormal throughout the colon in a significant proportion of individuals. Expression levels were not abnormal in CD. Expression of FAM5C in UC did not correlate with the known markers of inflammation, IL-8 , S100A8 , DEFA5 and DEFA6 , or with treatment. The under-expression of FAM5C in UC was confirmed in biopsies of non-inflamed rectal mucosa from an independent cohort of patients (FC = –1.68, p = 0.0073) and by qPCR (p Conclusion This is the first description of the under-expression of FAM5C in UC. As these observations were made in non-inflamed mucosa, low levels of this protein might be involved in the pathogenesis of the disease. Indications that FAM5C may function as tumour suppressor [1], could link to the observed predisposition to colonic malignancy in UC. Disclosure of Interest None Declared. Reference Kuriowa T et al. (2009) Oncol Rep, 1005–11.

Introduction We have previously shown that entire fields of dysplasia in the human stomach are derived from a single mutated, metaplastic gland. 1 This suggests that intestinal metaplasia (IM) can be considered a field defect among which dysplasia can arise and this would indicate that adenocarcinomas derived from such dysplasia would also be clonal. Recent work published by our laboratory has indicated that familial adenomatous polyposis–associated colorectal adenomas as well as some sporadic lesions 2 and dysplasia within Barrett9s oesophagus 3 are polyclonal. There is therefore a need to ascertain the clonality of gastric adenocarcinomas (GA). Methods Here we screened a large cohort of GA patients for mutations in genes accounting for nearly 90% of reported mutations in GA in order assess mutation frequencies. The screening was then followed by laser capture microdissection PCR sequencing and loss of heterozygosity (LOH) analysis of the mutated specimens in order to assess clonality of GA from dysplasia and IM. Results From the 51 patients cohort we have found 18 patients (35.3%) presenting mutations, but only one out of the 51 patients (1.9%) presented two independent mutations in a single cancer. We found 3 mutations in APC (6.3%), one in CDKN2A (2.2%), 13 in TP53 (27.7%), one in CTNNB1 (2.2%) and one in K-RAS (2.4%), but none in PIK3CA or PTEN. Our mutation frequencies are comparable to previous reports, however we observed that most functional mutations occurred as a single event despite screening multiple genes. Analysis of the multiple laser capture microdissected areas revealed multiple genotypes within the same cancer in three out of the five patients. Moreover, current LOH data shows LOH of chromosome 17 in IM and throughout the cancers in all different genotypes, suggesting that chromosome 17 LOH might have been the first hit mutation followed by mutations in TP53, and in one patient a subsequent CTNNB1 mutation. Conclusion This suggests that cancer progression may have been initiated by LOH in intestinal metaplasia and that GA cells might become genetically diverse as the cells evolve in the tumour. Therefore, IM is likely to represent a field defect for gastric cancer. Competing interests None declared. References 1. Gutierrez-Gonzalez L , Graham TA, Rodriguez-Justo M, et al. The clonal origins of dysplasia from intestinal metaplasia in the human stomach. Gastroenterology 2011; 140 :1251–60. 2. Thirlwell C , et al. Clonality assessment and clonal ordering of individual neoplastic crypts shows polyclonality of colorectal adenomas. Gastroenterology 2010; 138 :1441–54. 3. Leedham SJ , et al. Individual crypt genetic heterogeneity and the origin of metaplastic glandular epithelium in human Barrett9s oesophagus. Gut 2008; 57 :1041–8.

Introduction Pseudopolyps develop as a result of mucosal ulceration and epithelial regeneration. 1 They appear as islands of relatively normal epithelium in otherwise denuded mucosa and are the likely source of epithelium from which the mucosa repairs. They can be associated with areas of dysplasia but are thought to be benign. We have previously shown that protumourigenic mutations can spread through the entire colon in patients with UC-associated cancer. 2 We hypothesise that pseudopolyps are clonal expansions of crypts that have acquired a protumourigenic survival advantage over surrounding normal epithelium that frequently perishes in the inflammatory milieu. Methods To determine the genetic status of pseudopolyps and frequency of mutated pseudopolyps. DNA extracted from macrodissected UC-associated pseudopolyp tissue sections underwent nested PCR sequencing of common tumour suppressor and oncogenes known to be mutated in colitis associated cancers, using well established published protocols. Results Of 30 pseudopolyp samples analysed from 30 different patients, four patients had identifiable mutations; in CDKN2A (c.C387T p.129Y), tP53 -exon7 (c.C735T p.245G), and K-RAS (c.G37A p.G13S, and c.G35C p.Gly12Ala). Conclusion We have shown that pseudopolyps are a potential source of protumourigenic mutations in UC. Pseudopolyps may possibly be the site within the inflamed epithelium where mutations are harboured, and may be the source for restituted bowel. More numbers are needed to be analysed and this is planned for future work and comparison with normal matched tissue is required. These lesions have been traditionally thought to be benign, genetically inert, incidental findings, characteristic of chronic inflammation. Although this data are preliminary, these findings propose an exciting paradigm shift in the way we consider pseudopolyps and may alter endoscopic management of these lesions in the future. Competing interests None declared. References 1. Jalan KN , Sircus W, Walker RJ, et al. Pseudopolyposis in ulcerative colitis. Lancet 1969; 2 :255. 2. Leedham SJ , Graham TA, Oukrif D, et al. Clonality, founder mutations, and field cancerization in human ulcerative colitis-associated neoplasia. Gastroenterology 2009; 136 :542–50.

Introduction There is increasing attention on the integration of targeted agents for oesophageal adenocarcinoma (OA) therapy. The most notable example of success of oesophagogastric targeted therapy was the addition of a HER2 targeting agent in the Phase III ToGA study. However, there is limited data on the expression of HER2 in the progression from Barrett9s (BE) to OA. This study aimed to clarify expression in this sequence, and to show binding of C6.5, a single chain HER2 targeting human antibody Fv fragment (scFv), to known HER2 expressing OA cell lines in vitro. Efficacy of antibody based therapies can be enhanced by scFv9s which penetrate tumours more quickly and demonstrate better tumour: normal tissue specificity. Methods 33 paraffin embedded oesophageal tissue specimens were selected from patients with squamous (n=4), non-dysplastic BE (NDBE; n=4), low grade dysplasia (LGD; n=6), high grade dysplasia (HGD; n=8) and OA (n=12). Sections were immunostained with the automated Oracle Bond system (Leica, UK) for consistency. Staining was then scored by 2 expert pathologists according to the proportion of cells in the tissue staining positively as negative ( 10%). In phase 2, binding of C6.5 scFv to the cancer cell lines SKOV-3 (ovarian), OE-19, OE-33 (oesophageal) and HT-29 (colon) was identified with flow cytometry using a mouse secondary followed by tertiary anti-mouse FITC. Data were then analysed with FlowJo software. Results Significant expression of HER2 (>10% cells positive according to National Guidelines) was only seen in HGD (25%) and cancer (25%) specimens. Borderline staining (5%–10%) was seen in LGD (17%), HGD (13%) and cancer (8%). All NDBE and the remaining LGD, HGD and cancer samples were negative. Flow cytometry demonstrated C6.5 binding to SKOV-3, OE19 and OE-33 cells but not HT-29 cells. Conclusion This study demonstrated that significant HER2 expression is only seen in roughly a quarter of patients with HGD and OA and not in NDBE or LGD. We also found that the HER2 targeting scFv C6.5 binds to breast cancer and OA cell lines but not colon cancer, the negative control. HER2 targeting therapies could therefore be postulated for patients with BE with HGD and OA, and these may be enhanced with scFv9s such as C6.5. Competing interests None declared.

Introduction Finding an accurate and convenient biomarker for cancer progression in Barrett9s (BE) is of high clinical importance. DNA ploidy abnormalities (DNA-PA) are a reliable predictor of cancer risk in BE, but measurement is expensive and scarcely available. We have shown polo-like kinase-1 (PLK1) may act as a surrogate marker of DNA-PA in oesophageal adenocarcinoma (OA) resection specimens. This study aimed to examine the potential of PLK1 in predicting DNA-PA in the metaplasia-dysplasia-OA sequence. Methods 36 paraffin embedded oesophageal tissue specimens were selected from patients with non-dysplastic BE (NDBE, n=5), low grade dysplasia (LGD, n=5), high grade dysplasia (HGD, n=12), intramucosal cancer (IMC, n=5) and invasive OA (IOA, n=9). Sections were mounted and immunostained with 2 PLK-1 antibodies PLK1-M and PLK1-L using an automated system (BOND-MAX, Leica) for consistency. Results were reported by two independent expert pathologists blinded to patient status with the Allred scoring system, a composite of the percentage and intensity of staining. Results Aneuploidy was present in LGD (20%), HGD (75%), IMC (20%) and IOA (56%) samples. Using linear regression analysis, Pearson coefficient was calculated for the correlation between DNA-PA and mean Allred expression scores for each PLK1 antibody. PLK1-M had a higher degree of correlation (r 2 =0.26, p=0.001) then PLK1-L (r 2 =0.22, p=0.004). Inter-observer analysis with linear κ scores confirmed good correlation of PLK1-M (κ=0.72, 95% CI 0.60 to 0.83) and PLK1-L (κ=0.53, 95% CI 0.38 to 0.68), but a Bland–Altman plot found pathologist 2 had a trend to score PLK-L more highly. However, intra-observer analysis confirmed both PLK1-L scores correlated with ploidy status (r 2 =0.14, p=0.023 and r 2 =0.27, p=0.0016). Pathologists took 90.3 s/slide (mean) to review and score with the Allred method. Using a mean Allred cut off score of 3.5, the sensitivity and specificity for the detection of aneuploidy were 81.3% and 75% for PLK1-M, 93.8% and 45% for PLK1-L. Conclusion This pilot study demonstrated a significant correlation between Allred reporting of PLK1 staining and DNA-PA. PLK1-L appears more sensitive and PLK1-M more specific. PLK1 immunostaining is relatively inexpensive, less work intensive and is more available than current methods to predict DNA-PA. Scoring staining with Allred is rapid and reproducible. In future, PLK1 staining may provide the basis of a more durable biomarker panel to predict DNA ploidy. Competing interests None declared.

Introduction Radiofrequency ablation (RFA) is a relatively new endoscopic method for ablation of high grade dysplasia (HGD) and intramucosal adenocarcinoma (IMC) in Barrett9s oesophagus. Clinical trials have shown enduring eradication of these pathologies. 1 However, some patients develop recurrent dysplasia or cancer. 1 The reason for this is unknown. The aim of the study was to investigate whether this is related to the persistence of known cancer driving mutations after radiofrequency ablation. Methods Patient records were searched for patients with recurrent HGD or IMC after RFA. Biopsies and endoscopic mucosal resection (EMR) specimens were available before and after RFA for each case. Whole biopsies underwent nested polymerase chain reaction (PCR) sequencing for mutations commonly implicated in progression to adenocarcinoma ( TP53, CDKN2A, K-ras ). Results Tissue was obtained for six patients before and after RFA. All patients were male (mean age 67 SD±2). Samples from five patients contained detectable mutations. In 3/5 patients, the same mutation was found in material taken before RFA as after (all TP53 mutations). The indication for RFA in all three patients was HGD; two of the patients developed IMC after RFA. In these two cases laser microdissection was performed on the post-RFA EMR samples. PCR revealed the pre RFA mutation to also be present throughout the IMC specimen. Dysplastic tissue adjacent to the IMC contained a mixture of crypts that were either wild type for the mutation in the IMC or contained the mutation indicating that the IMC was a monoclonal outgrowth and that the persistent mutation was driving the development of the recurrence. Furthermore, in all three cases, PCR of a further specimens (performed at a later time point to the first post-RFA EMR), revealed the same mutation as the initial biopsy before RFA, and the first post-RFA EMR. Conclusion The recurrence of dysplasia and cancer after RFA is likely to be due to failure to remove persistent, cancer driving mutations. Further work will need to be done to assess whether this is because of technical errors, or related to specific problems such as “buried Barrett9s”. Competing interests None declared. Reference 1. Shaheen N , Overholt B, Sampliner RE. Durability of radiofrequency ablation in barrett9s esophagus with dysplasia. Gastroenterology 2011; 141 :460–8.

Introduction Sodium porfimer photodynamic therapy (sp-PDT) is a licensed minimally invasive treatment for Barrett9s Oesophagus (BO) with high grade dysplasia (HGD). Complete reversal (CR)-HGD is 50% at 5 years. sp-PDT is associated with a significant risk of strictures and prolonged photosensitivity up to 3 months. 5 aminolaevulinic acid (ALA) is an alternative treatment with 36 h photosensitivity. Non-randomised data suggests 85% CR-HGD and low risk of side effects. The aim of this study was to compare the side effect profile between the drugs and outcomes of therapy. Methods Single centre randomised controlled trial. Presence of HGD was confirmed on two occasions by two specialist GI pathologists. All patients had Endoscopic Ultrasound and endoscopic resection of visible lesions prior to PDT. Stratification was by length of BO ( 6 cm) and extent of dysplasia. Standard protocols for ALA and sp-PDT were followed. Endoscopic follow-up with 2 cm four-quadrant biopsy was at 6 weeks, 4 months, then annually. All adverse event data were collected. Results Sixty-four patients were randomised, 34 ALA and 30 sp-PDT. Median follow-up is currently 23 months. Strictures and skin photosensitivity were significantly more common with sp-PDT than ALA-PDT (33% vs 9% and 43% vs 6% respectively) p 2 = 4.34, p=0.037. For BO length >6 cm there was no significant difference. Rate of buried glands was significantly higher post PDT (48%) than pre PDT (20%), no significant difference between groups. Conclusion ALA-PDT has a better risk profile than sp-PDT. In patients with BO length ≤6 cm, preliminary results show ALA-PDT is associated with significantly higher CR-HGD. In longer segments of BO neither PDT drug was sufficiently efficacious to warrant use.

Improvement in jaundice and pancreatic enlargement with prednisolone is a reliable and characteristic feature of autoimmune pancreatitis (AIP). The aim of this study was to investigate whether immunosuppression results in long-term preservation of pancreatic structure and function, as this has not been well defined. We report a single-centre cohort study of all patients diagnosed with AIP from 2004-09. Diagnosis was based on the HISORt criteria, whilst assessment of pancreatic morphology was made pre and post-treatment, by CT and/ or MRI, according to clinical need. Exocrine and endocrine function was measured by stool Faecal Elastase 1 (FE1) and plasma glucose/HbA1c. Forty one patients (M:F, 33:8; median age 62 years, range 29-83) were included. Median follow-up to date is 35 months (range 1-71). At diagnosis, a pancreatic mass/diffuse swelling was noted in 36/41 (88%), serum IgG4 was elevated (>1.3g/l) in 22/41 (54%), and 30/41 (73%) had an immunohistological diagnosis, with >10 IgG4-positive plasma cells/high power field (n=15 pancreas, n=8 liver, n=8 ampulla, n=2 salivary gland, n=2 renal). 30/41 (73%) had IgG4-associated cholangitis, 14/41 (34%) had other organ involvement (n=7 renal, n=3 salivary gland, n=2 retroperitoneal fibrosis, n=2 neurological). 7/41 (17%) patients had previously undergone pancreatobiliary surgery for presumed pancreatic cancer (n=3 Whipple resection, n=4 biliary bypass). 38/ 41 (93%) patients were treated with a tapering oral prednisolone regimen, which was ceased at a median of 6 months (range 2.5-70.9 months). An initial clinical, radiological, and biochemical response was seen in all patients. Relapse following disease remission or failure to wean steroids occurred in 19/41 (46%), and of these patients 6/19 were treated with steroids, and 13/19 were treated with prednisolone 30mg and azathioprine 2mg/kg daily. Post-steroid pancreatic atrophy was seen in 23/41 (56%). Low FE1 (

Introduction Improvement in jaundice and pancreatic enlargement with prednisolone is a reliable and characteristic feature of autoimmune pancreatitis (AIP). The aim of this study was to investigate whether immunosuppression results in long-term preservation of pancreatic structure and function, as this has not been well defined. Methods We report a single-centre cohort study of all patients diagnosed with AIP from 2004 to 2009. Diagnosis was based on the HISORt criteria, while assessment of pancreatic morphology was made pre and post-treatment, by CT and/or MRI, according to clinical need. Exocrine and endocrine function was measured by stool faecal elastase 1 (FE1) and plasma glucose/HbA1c. Results Forty-one patients (33M/F8; median age 62 years, range 29–83) were included. Median follow-up to date is 35 months (range 1–71). At diagnosis, a pancreatic mass/diffuse swelling was noted in 36/41 (88%), serum IgG4 was elevated (>1.3 g/l) in 22/41 (54%), and 30/41 (73%) had an immunohistological diagnosis, with >10 IgG4-positive plasma cells/high power field (n=15 pancreas, n=8 liver, n=8 ampulla, n=2 salivary gland, n=2 renal). 30/41 (73%) had IgG4-associated cholangitis, 14/41 (34%) had other organ involvement (n=7 renal, n=3 salivary gland, n=2 retroperitoneal fibrosis, n=2 neurological). Seven (17%) of 41 patients had previously undergone pancreatobiliary surgery for presumed pancreatic cancer (n=3 Whipple resection, n=4 biliary bypass). 38/41 (93%) patients were treated with a tapering oral prednisolone regimen, which was ceased at a median of 6 months (range 2.5–70.9 months). An initial clinical, radiological, and biochemical response was seen in all patients. Relapse following disease remission or failure to wean steroids occurred in 19/41 (46%), and of these patients 6/19 were treated with steroids, and 13/19 were treated with prednisolone 30 mg and azathioprine 2 mg/kg daily. Post-steroid pancreatic atrophy was seen in 23/41 (56%). Low FE1 ( Conclusion While a favourable initial response to steroids is predictable in patients with a definitive diagnosis of AIP, loss of pancreatic structure and function is common, and the disease is associated with long-term morbidity and mortality. It will be difficult to clearly define the impact of treatment on the outcome of pancreatic and extra-pancreatic disease without a placebo controlled trial.

Abstract 124 Introduction: Understanding how multiple myeloma (MM) cells proliferate and self-renew is crucial to treating resistant disease and preventing relapse. The TNF family members APRIL (A proliferation-inducing ligand) and BAFF (B-cell activating factor) are secreted in the MM bone marrow microenvironment and share two common receptors, TACI and BCMA. APRIL and BAFF and have been shown to exert a mainly anti-apoptotic effect in human myeloma cell lines (HMCLs). Little is known, however, about the proliferative effect of APRIL and BAFF on primary MM cells. Aims: We aimed to determine the effect of APRIL and BAFF on cell cycle progression in primary MM cells and on the modulation of D-type cyclins and other cell cycle proteins. Patients and Methods: We studied the effects of APRIL and BAFF on purified MM cells from 26 patients by flow-cytometry using surface CD138 / Ki67 / propidium iodide staining. A culture system was optimised in which a mean (±SEM) of 73.3±10.9% (n=23) of primary CD138+ MM cells survived for up to 3 days in vitro. D-type cyclin expression was assessed by immuno-histochemistry (IHC) ± western blotting and correlated with FISH for IgH translocations. Fourteen patients expressed cyclin D1, 4 in association with t(11;14); while 12 patients expressed cyclin D2, 1 with t(4;14) and 4 with t(14;16). TACI and BCMA expression were determined by flow cytometry. Western blotting was used to determine expression and modulation of cell cycle proteins and activation of signalling pathways in vitro. Membrane-bound APRIL was detected by IHC in MM trephine biopsies. Results: In-vitro culture with APRIL for 72 hours increased the overall percentage of CD138+ cells in S+G2/M phases (S/G2M) from a mean (±SEM) of 5.5 ± 1.1% to 8.1 ± 1.5% (p Disclosures: No relevant conflicts of interest to declare.

Introduction Intestinal metaplasia of the stomach has been heavily implicated as a premalignant condition that can lead to gastric adenocarcinoma (GA). However, while there have been many studies describing the mutation load within GA, there is little evidence to show its genetic origins from metaplasia. Furthermore previous studies from our laboratory have shown that the development of a field metaplasia depends on the division of metaplastic crypts to form a patch. However there is some controversy about the clonality of such crypts. Here we describe experiments where we define the clonal nature of metaplastic crypts within the human stomach and demonstrate the genetic relationship these have with associated dysplasia. Methods Blocks of gastric adenocarcinoma with dysplasia-associated metaplasia were used in this study. Sections were stained with a dual enzyme assay for mitochondrial cytochrome c oxidase (CCO) and succinate dehydrogenase (SDH). It has been established that mitochondrial DNA (mtDNA) mutations that result in CCO-deficiency, are a reliable marker of clonality. We laser-captured individual cells from a single CCO-deficient crypt and PCR sequenced the entire mtDNA genome. To determine the genetic relationship between metaplasia and dysplasia we laser captured crypts along a strip of gastric mucosa that contained patches of metaplastic and dysplastic crypts. We then screened these for TP53 and APC mutations by nested PCR sequencing. Results Every cell within a CCO-deficient metaplastic crypt contained the same mtDNA mutation (G12898A transition in mtCOII, and therefore clona), whereas all cells from neighbouring CCO-positive crypts were wild type. Screening in one patient with GA revealed a truncating APC mutation (c.1450 G insertion). All crypts were laser captured and genotyped for this mutation. We demonstrated that 100% of dysplastic crypts were APC-mutated, interestingly, there were a few metaplastic crypts that also contained the same mutation, indicating the dysplasia did arise from a field of metaplasia. These findings were confirmed in two other patients with APC and TP53 mutations. Conclusion Contrary to previous reports, we have demonstrated that intestinal metaplastic crypts are clonal. Our work shows that there is a founder mutation within each dysplastic lesion. The same founder mutation can be detected within the surrounding metaplasia proving that dysplasia arises from metaplasia in the human stomach.

Background We highlight a chronic inflammatory disease we call 'hyper-IgG4 disease', which has many synonyms depending on the organ involved, the country of origin and the year of the report. It is characterized histologically by a lymphoplasmacytic inflammation with IgG4-positive cells and exuberant fibrosis, which leaves dense fibrosis on resolution. A typical example is idiopathic retroperitoneal fibrosis, but the initial report in 2001 was of sclerosing pancreatitis. Methods We report an index case with fever and severe systemic disease. We have also reviewed the histology of 11 further patients with idiopathic retroperitoneal fibrosis for evidence of IgG4-expressing plasma cells, and examined a wide range of other inflammatory conditions and fibrotic diseases as organ-specific controls. We have reviewed the published literature for disease associations with idiopathic, systemic fibrosing conditions and the synonyms: pseudotumour, myofibroblastic tumour, plasma cell granuloma, systemic fibrosis, xanthofibrogranulomatosis, and multifocal fibrosclerosis. Results Histology from all 12 patients showed, to varying degrees, fibrosis, intense inflammatory cell infiltration with lymphocytes, plasma cells, scattered neutrophils, and sometimes eosinophilic aggregates, with venulitis and obliterative arteritis. The majority of lymphocytes were T cells that expressed CD8 and CD4, with scattered B-cell-rich small lymphoid follicles. In all cases, there was a significant increase in IgG4-positive plasma cells compared with controls. In two cases, biopsies before and after steroid treatment were available, and only scattered plasma cells were seen after treatment, none of them expressing IgG4. Review of the literature shows that although pathology commonly appears confined to one organ, patients can have systemic symptoms and fever. In the active period, there is an acute phase response with a high serum concentration of IgG, and during this phase, there is a rapid clinical response to glucocorticoid steroid treatment. Conclusion We believe that hyper-IgG4 disease is an important condition to recognise, as the diagnosis can be readily verified and the outcome with treatment is very good.

Cells from the great majority of patients with chronic lymphocytic leukemia (CLL) express CD23. A recent histologic study has shown that CD23 is expressed more strongly in the proliferating centers of the lymph nodes, where the large prolymphocytoid cells are located. The aim of our study was to quantify the expression of CD23 and CD21 in small and prolymphocytoid cells from patients with CLL and B-cell lymphomas, and correlate this expression with clinical parameters.Using quantitative flow cyto-metry we analyzed the antigen density of CD23 and CD21 in: 1) 101 cases of chronic lymphocytic leukemia, 84 typical, 14 with increased prolymphocytes (CLL/PL) and 3 atypical, 2) 15 cases of CD23 positive B-cell lymphoma with circulating lymphoma cells and 3) 8 normal subjects. The results were correlated with morphology and clinical staging.Cells from CLL and CLL/PL have a significantly higher number of CD23 molecules than normal and lymphoma B-cells (p0.001 and p0.001, respectively). Differences were not significant for CD21. CLL and CLL/PL cases had similar values of CD23 and CD21 molecules, but analysis at a single level showed that prolymphocytes in typical CLL and CLL/PL expressed significantly higher CD23 (p=0.001, p=0.006) and CD21 (p=0.001, p=0.001) than small lymphocytes. There was no correlation between CD23 or CD21 antigen density and clinical stages although there was a trend for a brighter CD23 in stage C patients.Since interaction between CD23 and CD21 is important for B-cell activation, proliferation and tumor formation, findings that both molecules are upregulated in prolymphocytes suggest that this is the proliferating cell component in CLL and underline the association between progression and increased prolymphocytes in typical CLL and CLL/PL.

To develop and validate histologic diagnostic criteria for autoimmune pancreatitis (AIP) and its types.Thirteen pathologists participated in this 2-phase study to develop diagnostic criteria for AIP types 1 and 2 (phase 1) and validate them (phase 2). A virtual library of 40 resected pancreata with AIP and other forms of chronic pancreatitis (CP) was constructed. Readers reviewed the slides online and filled out a questionnaire for histopathologic findings and diagnosis.Diagnostic criteria for AIP and its types were proposed according to the results from the top 5 reviewers in phase 1. The interobserver agreement was significantly improved in phase 2 by applying the proposed diagnostic criteria. Features distinguishing AIP from alcoholic and obstructive forms of CP were periductal lymphoplasmacytic infiltrate, inflamed cellular stroma with storiform fibrosis, obliterative phlebitis, and granulocytic epithelial lesions. Although there was overlap, 2 types of AIP were recognized. Type 1 had dense lymphoplasmacytic infiltrate with storiform fibrosis and obliterative phlebitis, whereas type 2 was distinguished from type 1 by the presence of granulocytic epithelial lesions.Autoimmune pancreatitis can be distinguished from other forms of CP with substantial interobserver agreement. The 2 types of AIP can be distinguished by the proposed consensus histopathologic diagnostic criteria.

Background Whilst multi-parametric magnetic resonance imaging (mp-MRI) has been a significant advance in the diagnosis of prostate cancer, scanning all patients with elevated prostate specific antigen (PSA) levels is considered too costly for widespread National Health Service (NHS) use, as the predictive value of PSA levels for significant disease is poor. Despite the fact that novel blood and urine tests are available which may predict aggressive disease better than PSA, they are not routinely employed due to a lack of clinical validity studies. Furthermore approximately 40 % of mp-MRI studies are reported as indeterminate, which can lead to repeat examinations or unnecessary biopsy with associated patient anxiety, discomfort, risk and additional costs. Methods/Design We aim to clinically validate a panel of minimally invasive promising blood and urine biomarkers, to better select patients that will benefit from a multiparametric prostate MRI. We will then test whether the performance of the mp-MRI can be improved by the addition of an advanced diffusion-weighted MRI technique, which uses a biophysical model to characterise tissue microstructure called VERDICT; Vascular and Extracellular Restricted Diffusion for Cytometry in Tumours. INNOVATE is a prospective single centre cohort study in 365 patients. mp-MRI will act as the reference standard for the biomarker panel. A clinical outcome based reference standard based on biopsy, mp-MRI and follow-up will be used for VERDICT MRI. Discussion We expect the combined effect of biomarkers and VERDICT MRI will improve care by better detecting aggressive prostate cancer early and make mp-MRI before biopsy economically viable for universal NHS adoption. Trial registration INNOVATE is registered on ClinicalTrials.gov, with reference NCT02689271. Electronic supplementary material The online version of this article (doi:10.1186/s12885-016-2856-2) contains supplementary material, which is available to authorized users.